0% found this document useful (0 votes)
12 views26 pages

Silicon Chip June 2017 p28-53

The document describes an Arduino-based digital inductance and capacitance meter that provides accurate measurements of passive components, outperforming typical digital multimeters. It features a compact design, operates on 5V DC, and can measure inductance and capacitance with a resolution of five digits and an accuracy of ±1%. The meter is easy to build and can be constructed for under $100, making it accessible for both professionals and hobbyists.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
12 views26 pages

Silicon Chip June 2017 p28-53

The document describes an Arduino-based digital inductance and capacitance meter that provides accurate measurements of passive components, outperforming typical digital multimeters. It features a compact design, operates on 5V DC, and can measure inductance and capacitance with a resolution of five digits and an accuracy of ±1%. The meter is easy to build and can be constructed for under $100, making it accessible for both professionals and hobbyists.
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 26

Arduino-based Digital

Inductance/Capacitance Meter
Do you ever need to check or confirm the values of inductors or
capacitors? This Arduino-based LC meter will give you a digital
readout and can even measure parasitic inductance or
capacitance present in a circuit. It’s much more
accurate than most DMM-based LC meters.

M
any digital multimeters most any passive component quickly Essentially it’s an improved version
(DMMs) have capacitance and automatically, often measuring of the PIC-based Digital LC Meter we
ranges but they are not nor- not just their primary parameter (like described in the May 2008 issue of
mally accurate for values below about inductance or capacitance) but one or SILICON CHIP. This time, we’re basing
50pF. And those few DMMs that can more secondary parameters as well. it around an Arduino Uno or equiva-
measure inductance are often not very However, many of these you-beaut lent module.
good at measuring inductance in the instruments also carry a hefty price
range of 1-100H – those that are typ- tag, keeping them well out of reach Main features
ically used in audio and RF circuits. for many of us. Our new Digital LC Meter is com-
An inductance meter with a 10H Fortunately, thanks to microcon- pact and easy to build, since the Ar-
resolution (typical for DMMs) isn’t troller technology, much more afford- duino board comes pre-assembled. It
very helpful if you want to wind a able digital instruments are becoming also has a better LCD readout than the
choke of say 6.8H, for an amplifier available. These include both commer- previous version. It fits snugly inside
output filter. cial and DIY instruments like the low- a UB3 utility box and you should be
Professionals tend to rely on digital cost unit described here. able to build it for under $100.
LCR meters for these types of measure- It offers automatic digital measure-
ments. They allow you to measure al- By JIM ROWE ment of both inductance (L) and ca-

28  Silicon Chip siliconchip.com.au


pacitance (C) over a wide range and
Fig.1: operating principle of the LC Meter. L1 and C1 form a tuned circuit
with 5-digit resolution. Measurement in combination with an external capacitance/inductance connected via S1.
accuracy is better than ±1% of read- Feedback for oscillation is provided by a comparator and the frequency of
+5V
ing over most of the ranges. oscillation depends
C on the known values +5V
It operates from 5V DC, drawing an of L1/C1/C2 and the unknown external 100k 4 .7k
average current of about 62mA, so it component. The unknown value can be 100k
100k 4 .7k
can run from a 5V USB supply (either computed from the frequency of FOUT, 100k
mains or battery) or from a spare USB as described by
L the accompanying 1 0F
port on your PC. equations.
COMP Fout
How it works L1
L1 C2
1 0F COMP Fout

S1
The meter’s impressive performance Cx/Lx
S1
C2
Cx/Lx
relies on an ingenious measurement RLY1
100k 47k
100k 47k
technique developed almost 20 years C1 RLY1 TO AND
C1 TO AND
FROM
ago by the late Neil Heckt in the USA. C 1 0F FROMO
ARDUIN
It uses a wide-range test oscillator 1 0F ARDUIN O
and its frequency is varied by con-
necting the unknown inductance or CAL
CAL
capacitance you’re measuring. The re- L
C/L
sulting change in frequency is meas- C/L
ured by the microcontroller and used
to calculate the component’s value, HOW IT WORKS: THE EQUATIONS
which is displayed directly on a small (A) In calibration mode (B) In measurement mode
LCD panel. 1 (5) When Cx is connected: F3 = ————————
1
(1) With just L1 and C1: F1 = ——————
To achieve reliable oscillation over 2.  L1.C1 2.  L1.(C1+Cx)
a wide frequency range, the test oscil- 1
so Cx = C1  ( F1
2

—– –1 )
(2) With C2 added to C1: F2 = ————————
lator is based on an analog compara- 2.  L1.(C1+C2) F3 2

tor with positive feedback around it, (3) From (1) and (2), we can find C1: (6) Or when Lx is connected:
as shown in Fig.1. This configuration C1 = C2
F2
· —————
2
F3 = ———————
1
has a natural inclination to oscillate, 2
(F1 – F2 )
2
2.  (L1+Lx).C1
because of the very high gain between
( F1 )
2
(4) Also from (1) and (2), we can find L1: so Lx = L1 . —– –1
the comparator’s input and output. 1
2
F3
L1 = —————
When power (+5V) is first applied, 4.2 F12 .C1 NOTE: F2 & F3 should always be lower than F1
the comparator’s positive input is held
at +3.3V by the divider formed by the negative resistance across the tuned When power is first applied to the
two 100kΩ resistors and the 100kΩ and circuit, to cancel its losses and main- circuit, the microcontroller measures
4.7kΩ resistors. Initially, the voltage tain oscillation. Once this oscillation this frequency (F1) and stores it in
at the negative input is zero because is established, a square wave of the memory. It then energises reed relay
the 10µF capacitor at this input needs same frequency is present at the com- RLY1, which switches capacitor C2
time to charge via the 47kΩ resistor. parator’s output and it is this frequen- in parallel with C1 and thus lowers
So with its positive input much cy (FOUT) that is measured by the mi- the oscillator frequency. The micro
more positive than the negative input, crocontroller. then measures and stores this new
the comparator initially switches its In practice, before anything else is frequency (F2).
output high, to near +5V. connected to the circuit, FOUT, will Next, the micro uses these two fre-
Once it does so, the 10μF capacitor simply correspond to the resonant fre- quencies plus the known value of C2
connected to the negative input begins quency of the tuned circuit compris- to accurately calculate the values of
charging up via the 47kΩ resistor and ing L1, C1 and any stray inductance both C1 and L1. The equations it uses
the voltage at this input rises. As soon and capacitance that may be associ- to do this are shown in Fig.1. Follow-
as it goes above +3.3V, the comparator ated with them. ing these calculations, the micro turns
output switches low and the positive
input is brought to 1.67V due to the
100kΩ feedback resistor pulling the
Features & specifications
100kΩ divider low. Inductance range: ........................... 10nH
to 100mH+
The low comparator output voltage Capacitance range: ......................... 0.1p
F to 2.7µF+ (non-polarised only)
is also coupled through the 10µF input Measurement resolution: ............... five
capacitor to the tuned circuit formed digits in either mode
by inductor L1 and capacitor C1. This Range selection: ............................. auto
matic
makes the tuned circuit “ring” at its Sampling rate: ................................ appr
oximately one measurement per seco
resonant frequency. Accuracy (when calibrated): ........... ±1% nd
As a result, the comparator and the of reading, ±0.1pF or ±10nH
Supply voltage: ............................... 5V
tuned circuit now function as an os- DC @ <65mA (including backlit LCD)
cillator at that resonant frequency. In Supply type: ................................... USB
charger or the USB port on a PC
effect, the comparator functions as a

siliconchip.com.au June 2017  29


RLY1 off again to disconnect C2, allow- the value of Lx or Cx. As a result, we are able to use the
ing the oscillator frequency to return Each of these values needs to be Arduino to easily measure the oscil-
to F1. The unit is now ready to meas- calculated to a high degree of reso- lator’s frequency.
ure the unknown inductor or capaci- lution and accuracy, using floating- The results of the Arduino’s meas-
tor (Lx or Cx). point maths. urements and calculations are dis-
As shown in Fig.1, the unknown played on a blue back-lit 16x2 alpha-
component is wired to the test termi- Circuit details numeric LCD module.
nals at far left. It is then connected The full circuit diagram is shown This has a serial I2C module fit-
to the oscillator’s tuned circuit via in Fig.2. It mainly consists of the Ar- ted, so it can be controlled from the
switch S1. duino microcontroller module and the Arduino via its I2C port lines (SCL
When measuring an unknown ca- serial I2C LCD module together with and SDA).
pacitor, S1 is switched to the “C” posi- the oscillator circuit we’ve already in- Its features were fully described in
tion so that the capacitor is connected troduced, built using an LM311 high- SILICON CHIP March 2017 issue.
in parallel with C1. Alternatively, for speed comparator (IC1).
an unknown inductor, S1 is switched The Arduino controls RLY1 to Calibration functions
to the “L” position so that the inductor switch calibrating capacitor C2 (1nF) The firmware sketch running in the
is connected in series with L1. in and out of circuit, via its IO3 pin. Arduino is designed to perform its
In both cases, the added Lx or Cx Diode D1 is connected across the relay “zero calibration” adjustment just af-
again causes the oscillator frequency coil to prevent the Arduino’s internal ter initial startup.
to change to a new frequency (F3). circuitry from being damaged by in- But pushbutton switch S2 is also
As with F2, this will always be lower ductive spikes. provided to allow zero calibration to
than F1. The Arduino senses which position be performed at any other desired time
So by measuring F3 as before and L-C switch S1 is in using its IO2 pin, as well (to allow for temperature drift,
monitoring the position of S1 (which which is pulled high internally when for example).
is done via the C/L line), the micro can it’s not pulled low by S1b (in the L S2 pulls the Arduino’s RESET pin
calculate the value of Lx or Cx using position). (pin 4) down, so that it is forced to re-
one of the equations shown in the right The output of the oscillator at pin set and start up again, readjusting its
section of the equations box in Fig.1. 7 of IC1 is taken to pin IO5 of the Ar- zero setting in the process.
From these equations, you can see duino via a series 6.8kΩ resistor. It LK1 and switch S3 can be used to
that the micro has some fairly solid needs to be taken to this pin because nudge or tweak the calibration in small
number-crunching to do, both in the this is also the external input pin for increments or decrements, if you have
calibration mode when it works out the the 16-bit timer/counter inside the access to an accurate reference capaci-
values of L1 and C1 and in the meas- ATmega328P micro which forms the tor. When LK1 is fitted, pulling input
urement mode when it must work out heart of the Arduino Uno. pin IO7 low, the micro will increase

Fig.2: complete circuit of the LC Meter. The oscillator circuitry is as LCD WITH I C SERIAL BACKPACK
2

GND
shown in Fig.1; most of the remaining work is done by the Arduino
module. The result is displayed on a serial (I2C) LCD while VCC

additional switches and a link are provided for calibration SDA 16 x 2 LCD
and zeroing of the Meter. SCL
Diode D1 protects the IO3
pin which drives the reed +5V +5V SET ZERO
relay from back-EMF
spikes when the relay 100nF 100k 4.7k
S2
switches off.
100k
ADC1
ADC 5/SCL

ADC 4/SDA

ADC2

+3.3V
ADC3

GND
GND
ADC0

+5V

RESET

+5V

DC VOLTS
VIN

INPUT
1 0F 8
2 5
IC1 6
7 6 5
TANT
ACTIVE 3 LM311
LM 3 11 GND RST ARDUINO UNO,
UNO ,
L1 100 H 1 4 3
4 FREETRONICS ELEVEN,
ELEVEN , USB TYPE B
MOSI SCK MICRO
2 1 DUINOTECH CLASSIC,
CLASSIC , ETC
Cx/Lx S1 100k
+5V MISO
ICSP
1nF 1%
IO 11/MOSI

1nF 1%
IO 12/MISO

GND
IO 13/SCK
IO 9/PWM
IO 6/PWM
IO 3/PWM

IO 4/PWM

IO 5/PWM
IO 2/PWM

10F 47k
IO 0/RXD

IO 10/SS
IO 1/TXD

(C1) (C2)
TANT
AREF
GND

SDA

6.8k
IO8

SCL
IO7

C
C/L

Fout
NUDGE
1, 14 2 READING CONNECT TO PC
CAL
L RLY1 K INCR S3 CALIBRATE TO PROGRAM
JAYCAR ARDUINO , OR TO
D1 5V/1A PLUGPACK
SY-4030 1N4148
DECR LK1
(5V/10mA) TO RUN SWEEPER
A LK1 SHOULD BE OUT FOR NORMAL OPERATION,
7, 8 6
IN ONLY FOR CALIBRATING METER VIA S3

1N4148
SC
20 1 7 ARDUINO - BASED DIGITAL LC METER A K

30  Silicon Chip siliconchip.com.au


the capacitance reading by about
TO SERIAL LCD ARDUINO “PROTO SHIELD” NOTE: PIN HEADER
0.5% each time S3 (a centre-off rocker STRIPS TO CONNECT
switch) is pushed to the upper “INCR” SHIELD TO ARDUINO

GND
VCC
SDA
SCL
SDA A GND D12 D10 D8 D6 D4 D2 D0
100nF 100nF MOUNTED ON THE
position, or alternatively decrease the SCL REF D13 D11 D9 D7 D5 D3 D1
UNDERSIDE
reading by the same amount if S3 is

5V

5V
pushed to the lower “DECR” position. SCL

6.8k
100k

4.7k
SDA
So the idea is to push S3 in one di- VCC
+ ICSP
(C2) 1nF
rection or the other until the reading GND 10 F
(FIT LK1 JUMPER
ONLY FOR
is correct. IC1
CALIBRATION)
NOTE: PIN HEADER LM 311

RLY1 SY4030
Each time a change is made, the ad-

C1
GND
C/L
STRIPS TO CONNECT

4148
justment factor is stored in the Ardui- SHIELD TO ARDUINO 100k IO6/INCR
MOUNTED ON THE 47k + GND
no’s EEPROM memory, so it’s remem-

10 F
UNDERSIDE 100k IO4/DECR
bered for future sessions. When link

GND

GND
LK1 is not fitted, pressing S3 in either 3mm
LED
560R

IO 3V3 GND VIN A1 A3 A5

POWER
direction has no effect at all.

RESET
REF RST 5V GND A0 A2 A4

This is a safety feature, to prevent S2


unintended changes to the meter’s cal- L1 S3
ibration during normal use. Although TOP (=)
this calibration is normally done us-
ing a reference capacitor, it also im- S1
C1
proves the accuracy of inductance 1nF
measurements. BOTTOM (–)

Construction L C
There is no custom PCB used for the
LC meter’s circuitry; instead, most of
the added circuitry is fitted on a proto-
type shield board which simply plugs
into the top of the Arduino PCB. TEST
There aren’t that many components TERMINAL
BINDING
involved, so it’s a straightforward job POSTS
to wire it up as shown in the wiring
diagram, Fig.3.
The only components which are not
mounted on the ProtoShield are the se-
rial LCD module, switches S1-S3, the
test terminal binding posts and refer-
Fig.3: follow these diagrams to fit the components to the ProtoShield and
ence components L1 and C1. also to wire up all the external connections. Connections made between
As shown in Fig.3 and the photos, component pads on the underside are shown below, significantly larger
these are all mounted on the lid of the than the 1:1 diagram above, for clarity. These should be made with
UB3 box, which forms the meter’s front insulated wire to avoid short circuits.
panel. These off-board components are
all linked to the ProtoShield board via
short multi-wire interconnection leads 0D 2D 4D 6D 8D 0 1D 2 1D D N G A AD S
and SIL connector plugs and sockets, 1D 3D 5D 7D 9D 1 1D 3 1D FER L CS Fn 0 0 1

which are also shown in Fig.3.


You can get an idea of how every-
V5

V5

thing fits together from the internal


cutaway diagram of Fig.4, along with
the internal photos.
The Arduino module mounts in the
P S CI
+
bottom of the box via four 9mm long
M2.5 machine screws and four M2.5
nuts, with another four M3 or M2.5
Nylon nuts used as spacers.
The ProtoShield is plugged into the +
top of it. The rest of the meter circuit-
ry connects via the 90° pin headers on
DNG

DNG

the ProtoShield. R065


Follow the wiring diagram (Fig.3) 5A 3A 1A NIV DNG 3V 3 OI
mm 3
DEL
R E W OP

and internal photos to build the Pro-


TESER

4A 2A 0A DNG V5 TSR FER

toShield. Start by soldering the com-


ponents into place where shown in

siliconchip.com.au June 2017  31


2. Fit a four-pin vertical header for
switch S2.
3. Fit the four SIL pin headers to the
underside, along the upper and low-
er edges of the ProtoShield, which
connect it to the Arduino. These
comprise a 10-pin header at upper
left, two 8-pin headers (one at upper
right and the other at lower centre)
and a 6-pin header at lower right.
Do not fit a 3x2 DIL pin header in
the ICSP position at centre right on
the ProtoShield board.
4. Fit the 8-pin DIL socket for IC1, with
its notched end to the left, then re-
lay RLY1, with its notched end to-
wards the top.
5. Mount the six resistors, the 100nF
capacitor and the two 10µF tanta-
lum caps. Note that the last two are
polarised, so make sure you fit them
with the orientation shown.
6. Fit the 12 insulated wires on the top
“Larger than life” photo of the wiring on the top side of the Freetronics Arduino
of the board and any insulated wires
ProtoShield board (actual size is shown in Fig.3, below). This board “plugs in”
to the Arduino Uno (etc) board via the rows of pin headers on the underside; the required to complete the wiring on
I2C LCD board plugs into the ProtoShield board. the underside. This will require you
to strip the insulation from each end
Fig.3, ensuring you use the correct leads while fitting them and trim them by about 5mm or so.
orientation for polarised components: so that they reach the adjacent pads. 7. Fit diode D1, making sure its end
IC1, diode D1, RLY1 and the two 10µF For longer connections, use compo- with the cathode band is uppermost
tantalum capacitors. nent lead off-cuts, routed carefully to and adjacent to pin 2 of RLY1, then
Next, add the wiring on the under- avoid the possibility of shorting any- plug IC1 into its socket.
side, as shown in the underside wiring thing else, or short lengths of light-
diagram of Fig.3. Use insulated wire duty hookup wire (eg, stripped from Box and lid preparation
because several of these wires cross a piece of ribbon cable) or bell wire. There are four holes to drill in the
over each other. Here’s our suggested order of fitting bottom of the box for mounting the Ar-
In cases where adjacent pads are the components and wiring the Pro- duino module and two larger holes to
connected, you can simply place a toShield board; check Fig.3 for the cut in the left-hand end for the USB
solder bridge between the two pads exact placement in each case: plug and alternative DC power plug.
or alternatively, bend the component 1. Fit the four 90° SIL headers. The locations and dimensions of

BINDING POSTS

GND TEST M2 x 6mm SCREWS


SECURING SLIDE SWITCH
M3 x 15mm
SCREWS

9mm LONG
UNTAPPED NYLON S1
SPACERS 16x2 LCD MODULE
S2 L1 S3

M3 NYLON
HEX NUTS
I 2C SERIAL INTERFACE
MODULE

ARDUINO “PROTO SHIELD”


WITH L-C METER CIRCUITRY
M3 OR M2.5 M2.5 NUT
M2.5 NUT
NYLON NUTS
AS SPACERS ARDUINO UNO
OR EQUIVALENT
SELF-ADHESIVE
SELF-ADHESIVE M2.5 x 9mm
UB3 BOX (CUTAWAY) RUBBER FOOT
RUBBER FOOT

Fig.4: this shows how everything fits together inside the UB3 “Jiffy” box. The Arduino module is attached to the bottom
of the case with the proto-board hosting most of the remaining circuitry plugged on top. The three switches, two binding
posts and the I2C LCD module are mounted on the lid and connected to the ProtoShield via flying leads.
32  Silicon Chip siliconchip.com.au
Parts list – Arduino-
based LC Meter
1 Arduino Uno R3, Duinotech Classic,
Freetronics Eleven or equivalent
microcontroller module
1 Serial I2C 16x2 LCD module with
back-lighting (SILICON CHIP online
shop Cat SC4198)
1 Arduino Uno Prototyping Shield
(eg, Freetronics SH-Proto-Basic)
1 UB3 “Jiffy” box, 130 x 68 x 44mm
1 laser-cut clear acrylic lid for UB3
box [optional but recommended]
(SILICON CHIP online shop Cat
SC4274)
The Freetronics Eleven (Uno equivalent) board, mounted in the bottom of the 4 self-adhesive rubber feet
case (see drilling template on pages 35 and 36).
1 5V/10mA DIL reed relay
(RLY1; Jaycar SY4030)
all of these holes are shown in Fig.5, the box lid using a sharp hobby knife. 1 100µH axial RF inductor
the drilling template, while the cor- Or, for longest life and an even more (L1; Jaycar LF1534)
responding information for the holes professional finish, consider fitting the 1 DPDT subminiature slide switch (S1;
to be drilled and cut in the lid/front label to the underside of the lid – it’s Jaycar SS0821)
panel are shown in Fig.6. more fiddly to fit but doesn’t require 1 panel-mount SPST NO momentary
For best results, start the larger holes laminating, nor double-sided tape to pushbutton switch (S2; SP0710)
with a smaller pilot drill and enlarge hold it in place (the switches and ter- 1 panel-mount SPDT on-off-on
with a stepped drill bit, series of larg- minals hold it in position; a very light momentary rocker switch
er drill bits or a tapered reamer. Rec- mist of clear spray adhesive will also (S3; Jaycar SK0987)
tangular or other non-round holes can ensure it stays tight against the lid). 1 8-pin DIL IC socket
be made by drilling a series of holes, Perhaps it’s gilding the lily some- 1 40-pin header, 2.54mm pitch
knocking out the centre section and what but if you can print the label onto 1 40-pin right-angle header, 2.54mm
then filing the hole to shape. clear film, you can see the “works” pitch
We fixed four self-adhesive rubber through the label, as we did with the 1 150mm socket-to-socket jumper
feet to the underside of the box to pro- photo on page 28. ribbon cable (Jaycar WC6026)
tect any surface it’s placed on. Just make sure you get the right film 1 jumper shunt
Making all the required holes in the to suit your type of printer (eg, laser 2 binding posts with integral banana
lid is rather tedious as there are twelve, printer or inkjet printer, etc). socket (1 red, 1 black)
including three rectangular cut-outs Once the lid/front panel is finished, 4 9mm Nylon untapped spacers, 3mm
and two holes with flat edges. fit switches S1-S3 to it, along with the inner diameter
To save time and guarantee a neat two test terminal binding posts and the 4 15mm M3 machine screws
result, you can purchase a laser-cut serial LCD module. 8 M3 Nylon hex nuts
clear acrylic lid (which replaces the lid Slide switch S1 attaches to the front 4 9mm pan head M2.5 machine
supplied with the box) from the SILI- centre of the lid via two 6mm long screws
CON CHIP online shop (see parts list). M2 machine screws, while switch S2 4 M2.5 hex nuts
As the acrylic panel is transparent mounts using the spring washer and 2 6mm M2 machine screws (for S1)
the lid doesn’t need a cut-out to view nut supplied with it and S3 simply Semiconductors
the LCD. Note that since the 3mm pushes into its rectangular mounting 1 LM311 DIP high-speed comparator
acrylic is slightly thicker than the lid hole until its two barbs spring out- (IC1; Jaycar ZL3311)
supplied with the UB3 box, depending wards to hold it in place. 1 1N4148 small signal diode (D1)
on the length of the screws that came Just make sure that you fit it with
with it, you may need to use slightly the “=” sign on its rocker actuator up- Capacitors
longer self-tapping screws to attach it. permost (see photos). 2 10µF 16V through-hole tantalum
We have also prepared artwork for The two binding posts are mount- 1 100nF multilayer ceramic
the front panel, to give it a professional ed using the mounting nuts and lock 2 1nF 1% NP0 ceramic, mica, MKT,
look. You can download this as a PDF washers provided. polypropylene or polystyrene
file from the SILICON CHIP website. Take care doing so, however, as the (SILICON CHIP online shop Cat
There are two ways to go here: after upper and lower mounting bushes SC4273)
you print it, it can be hot laminated, have D-shaped sections which should Resistors (all 0.25W, 1%, through-
then attach it to the box lid using dou- mate with the matching holes in the hole mounting)
ble-sided adhesive tape or spray glue. lid/front panel. 3 100kΩ 1 47kΩ 1 6.8kΩ 1 4.7kΩ
After that, you can cut out the holes The serial LCD module mounts un-
in the front panel to match those in der the lid in the top centre position,
siliconchip.com.au June 2017  33
The underside of the lid, showing the LCD
module, I2C module, the three switches and two
terminals attached.

using four 15mm long M3 machine ble for the LCD, you can just plug it
screws passing down through four in at both ends.
9mm long untapped Nylon spacers In each case, make sure each wire
and fastened using four Nylon M3 nuts goes to the correct pin as with indi-
(under the module PCB). In contrast, the lead which connects vidual sockets, it’s easy to get them
With the LCD module in position, to the serial LCD module has a four- out of order.
your front panel assembly is ready way SIL header socket at each end, Having made up all the required
to be wired up and provided with its while the lead to connect zero/reset leads, complete the LC Meter assem-
various leads to connect to the Pro- switch S2 (although of only two wires) bly with the following steps:
toShield board. ends in a four-way SIL header socket, 1. Mount the Arduino module inside
Refer back to Fig.3 and the internal with the wires connecting only to the the bottom of the box using four
photo, following them carefully to pins on each end. 9mm M2.5 screws and nuts, using
make the correct connections between The two pins in the centre of the four Nylon M3 nuts as spacers.
S1, the test terminal binding posts and socket can be either cut short or pulled 2. Plug the LC Meter ProtoShield into
L1 and C1 in particular. out, since they are not used. the Arduino, making sure you have
Note that the leads of L1 and C1 Rather than using SIL sockets like all four SIL pin headers lined up
should be kept as short as possible, to we did on the prototype, we suggest correctly.
keep stray capacitance low (and sta- you simply split a 40-way ribbon 3. Holding the front panel assembly
ble). You can then make up the vari- jumper cable with individual “Du- close to the top of the box and ori-
ous short leads which will connect Pont” sockets on each wire. entated correctly, plug the various
the front panel components to the Pro- This makes the job really easy; you connection leads into their match-
toShield board. simply pull off the required number of ing pin headers on the ProtoShield.
Note that the lead which connects wires and then cut the cable to length Be especially careful to get the cor-
S1, L1, C1 and the test terminals to the and strip the free end, to solder to the rect connections between the Pro-
ProtoShield ends in a three-way SIL switch or connector. toShield and the LCD module, as
header socket, as does the lead from S3. You don’t even need to cut the ca- shown in Fig.3.

Resistor Colour Codes


No. Value 4-Band Code (1%) 5-Band Code (1%)
 3 100kΩ brown black yellow brown brown black black orange brown
 1 47kΩ yellow purple orange brown yellow purple black red brown
 1 6.8kΩ blue grey red brown blue grey black brown brown
 1 4.7kΩ yellow purple red brown yellow purple black brown brown

34  Silicon Chip siliconchip.com.au


Here’s the alternative finish using a paper-printed label fixed to the outside of the UB3 Jiffy box lid, after it has been
drilled and cut to suit. (You can, of course, glue a paper label to the laser-cut lid purchased from the SILICON CHIP online
store). In this case the meter is measuring a nominal 100µH inductor and showing it is slightly high at 103µH.

4. Lower the lid assembly down into to suit different operating systems: If you’re using a Freetronics Elev-
the box and fix it into place. Windows (32-bit or 64-bit), macOS en module, you can download the
5. Program the Arduino, as described and Linux (32-bit, 64-bit and ARM). appropriate driver from their website
below. After the IDE has been installed, (www.freetronics.com.au). All of their
download our firmware sketch for the drivers are zipped up in a file called
Uploading the firmware LC Meter from the SILICON CHIP website “FreetronicsUSBDrivers_V2.2.zip”,
In order to do this, you need to have (www.siliconchip.com.au). It’s called and there’s also a document which
the Arduino IDE installed on your PC. “Arduino_LC_meter_sketch.ino”. explains how to install it.
The latest version of the IDE can Now plug your LC Meter into one Once the USB driver has been in-
always be downloaded from the Ar- of your PC’s USB ports, using a suit- stalled and your operating system con-
duino website (www.arduino.cc/en/ able USB cable (usually with a Type firms that it can communicate with the
Main/Software). A plug on one end, and a micro Type Arduino in your LC Meter, use Control
At the time of writing, the latest B plug on the other). You may need to Panel to find out which COM port the
version is V1.8.2, dated 22/03/2017. install the correct USB VCP driver for Meter’s Arduino has been allocated (eg,
There are various versions available it if this is not already installed. COM5, COM7, or whatever).

CL
29 A
B HOLES A: 2.5mm DIAMETER
A HOLE B: 12mm DIAMETER

ALL DIMENSIONS IN MILLIMETRES


24 39
18 19

Fig.5: the drilling CL


templates for the 11 9
four Arduino
A
mounting holes in 24
13

the bottom of the 38

box along with the 24


14
A
12.6

USB and DC power


access holes in the LEFT-HAND END OF UB3 BOX UNDERSIDE OF UB3 BOX
left-hand end.
siliconchip.com.au June 2017  35
Set the port for communication at This means that the Meter has de-
115,200 baud with the 8N1 “no hand- tected that S1 is set to the L position, Credit where it’s due
shaking” protocol. The COM port and is assuming that you want to do As mentioned earlier, this Digital LC
number should be entered into the the zero calibration in this mode. As Meter, like our earlier May 2008 design,
Arduino IDE’s Tools->Port pull-down a result, it’s advising you to fit a very is based on a 1998 design by the late Neil
menu after you start it up. low inductance shorting bar between Heckt, of Washington, USA.
Now open the LC Meter firmware the test terminals. This can be in the Since then, various people have pro-
sketch in the Arduino IDE, verify and form of a 40mm long piece of 1.66mm duced modified versions of the design,
compile it, and then upload it into diameter copper or brass rod between including Australian radio amateur Phil
the LC Meter’s Arduino flash memory. the terminals, or (better still) a 40 x Rice VK3BHR, of Bendigo in Victoria. Mr
Soon after it has been uploaded, your 30mm rectangular piece of 1mm thick Rice and others also modified the firm-
meter should spring into life, flashing copper or brass sheet between them. ware and adapted it to use the PIC16F628
this message on the LCD screen: In either case, the rod or sheet must micro with its internal comparator. They
be shiny rather than oxidised. also added a firmware calibration facility.
Silicon Chip So a significant amount of credit for
Note that if you have set S1 to the L
Digital LC Meter this latest version of the design must go
position accidentally and don’t have a
This should remain visible for two shorting bar available, there’s no harm to these earlier designers. The author is
seconds, after which the screen should done. Simply flick S1 to the C position happy to acknowledge their earlier work.
go black, before the Meter begins its and then press switch S2 to get the
initial zero calibration. Meter to reset and begin over again. This will display for one second,
If you don’t see this initial message, Or if you do want to calibrate in in- after which the Meter will begin mak-
this may be because the contrast trim- ductance mode, simply fit the short- ing measurements. If you have done
pot on your LCD display module’s se- ing bar between the terminals (if you the initial calibration in C mode and
rial interface PCB is not set to the cor- haven’t already done so) and then S1 is still in this position but no un-
rect position. press S2 to reset and begin over again. known capacitor is as yet connected
The remedy is to swing open the lid In either case, there will be a brief to the test terminals, you should now
of the box just enough to fit a very small pause after which the meter will show get a display like this:
screwdriver or alignment tool into the the values for C1 and L1 it has found
trimpot’s adjustment slot, turn it and from the initial calibration. This will Cx = 0.004 pF
then press switch S2 to force the Ar- be something like: (F3 = 515838 Hz)
duino to reset and start again. where the value shown for Cx is very
C1 = 1084.2 pF
Try changing the pot setting in one close to zero, while the frequency F3
L1 = 91.24 uH
direction or the other until the mes- shown on the second line is for the cur-
sage becomes clearly visible, pressing The actual values displayed will de- rent oscillator frequency; essentially
S2 after each adjustment. pend on the components in your unit, the same as F1 at the current ambient
as well as the stray capacitance and in- temperature.
Startup & calibration ductance. They’re shown at this stage The Meter’s oscillator frequency
At start-up, the Meter normally ex- mainly as reassurance that the Meter does drift a little with temperature.
pects slider switch S1 to be set in the is working correctly. The measured This means that after a while, the value
Capacitance (C) position, and no ex- values of C1 and L1 will be displayed shown for Cx with no external capaci-
ternal capacitor to be connected to the for three seconds, after which this mes- tor connected may creep up from the
test terminals. If you have done this it sage will appear: almost-zero reading you get initially.
will now display the message: At the same time, the reading for F3
Calibration done
S1 set for C: OK Ready to measure would slowly decrease.
Now calibrating If you find the value shown for Cx

But if you have set S1 in the CL


A 37.5 37.5 A
Inductance (L) position instead,
you’ll see a different message:
Fit shorting bar “WINDOW”
CUTOUT OR CUTOUT
FOR LCD VIEWING
Now calibrating 26.5 FOR LCD VIEWING

CUTOUT
10 FOR
C 65 x 16mm
S3
CL
Fig.6: you can either drill and 4.5 A A
cut the twelve cut-outs required S2
11.5 9.5 9.5
in the lid supplied with the B B
13 x 20mm

UB3 “Jiffy” box, as shown in


this diagram, or (much easier!) 49 49
11.5 D D
purchase a laser-cut acrylic lid HOLES A: 3mm DIAMETER;
CUTOUT FOR S1
from the SILICON CHIP online HOLES B: 2.5mm DIAMETER
9.5 x 4.5mm
store and use that instead of HOLE C: 7mm DIAMETER; BINDING
HOLES D: 9mm DIAMETER WITH FLAT POSTS ALL DIMENSIONS IN MILLIMETRES
the lid that came with the box.

36  Silicon Chip siliconchip.com.au


DIGITAL LC METER
NUDGE
CALIBRATION
INCR
ZERO

INDUCTANCE CAPACITANCE
DECR

(ONLY WHEN
LK1 FITTED)
Lx Cx
SILICON
OR

CHIP GND TEST

www.siliconchip.com.au

Fig.7: same-size front panel artwork designed to fit a UB3 Jiffy Box. It will also
fit the laser-cut acrylic front panel from the SILICON CHIP online store. This,
along with the two cutting/drilling diagrams, can also be downloaded (as a PDF)
from www.siliconchip.com.au

with no external capacitor has crept zero calibration.


up to 0.1pF or more, simply press S2
again to get the Arduino to perform a Optimising accuracy
new zero calibration. If all is well so far, your Digital LC
On the other hand, if you’ve done Meter should be operating correctly
the initial calibration in L mode and S1 and ready for use. If you have been
is still in this position but the shorting able to procure a couple of 1% toler-
bar is still connected across the termi- ance (or better) capacitors for C1 and
nals, you should get a display like this: C2, it should also be able to deliver
that order of accuracy without any ex-
Lx = 0.002 uH
tra calibration.
(F3 = 516615 Hz)
But as mentioned earlier, it is pos-
The value shown for Lx is again very sible to achieve even better accuracy
close to zero, and the frequency F3 with the meter providing you have
shows the current oscillator frequen- access to a reference capacitor whose zero figure. If it does, press switch S2
cy, again very close to F1 at the cur- value is accurately known (because to force a reset and bring the reading
rent ambient temperature. Now if you you’ve been able to measure it with a back to less than 0.01pF.
remove the shorting bar in this mode, high-accuracy LCR meter). 6. Connect your known-value capaci-
you’ll find the display will change to Ideally, this reference capacitor tor to the test terminals and note
something like this: should have a value of between 10nF the Meter’s measurement reading.
and 100nF, but even one with a value It should be fairly close to the ca-
Over Range!
between 1nF and 10nF would be OK. pacitor’s known value, but may be
(F3 = 2 Hz)
This is achieved by tweaking or a little higher or lower.
This simply shows that in this “nudging” the Meter’s reading for the 7. If the reading is too low, press the
mode, an open circuit between the reference capacitor using switch S3. rocker of switch S3 at the upper
terminals is equivalent to a very high Here’s how you do it: (“=”) end for a second or so; if it’s
inductance, because it causes the os- 1. Remove the 5V supply from the too high, press the lower end (“-”)
cillator frequency to drop to near zero. Meter instead. The reading should change
When you connect a real inductance 2. Lift the lid/front panel up from the by about 0.5%. Continue until the
between the test terminals, the Meter box and carefully fit the jumper reading is as close as possible.
will measure its inductance and dis- shunt over the pins for LK1, down 8. Remove power, open the lid and re-
play it (assuming its value is within on the ProtoShield. move the jumper from LK1.
the Meter’s range, which is from 10nH 3. Close the box and slide S1 to the 9. Re-attach the lid.
to 150mH). C position but don’t connect your Note that since the Arduino always
As before, drift in the Meter’s oscil- reference capacitor to the test ter- saves the revised calibration factor in
lator may cause the Lx reading for the minals. its EEPROM after every measurement
shorting bar to creep up gradually. So 4. Re-apply the 5V power and let the during this nudging procedure, so you
before making a particularly critical Meter go through its initial zero only have to do the calibration once.
measurement, it’s a good idea to fit calibration. Also, when you calibrate the meter
the shorting bar between the test ter- 5. Wait a couple of minutes, watching in this way using a known value ca-
minals and press S2 again to force the the reading for Cx to see if it drifts pacitor, it’s also calibrated for induct-
Arduino to reset and perform a new up appreciably from the initial near- ance measurements too. SC

siliconchip.com.au June 2017  37


LTspice Part 1: by Nicholas Vinen

simulating and
testing circuits
SPICE is a powerful tool which allows you to use a computer to simulate
how a simple or complex circuit will behave without actually having to
build it. This allows you to experiment with different configurations and
examine the internal operation of a circuit before building it, saving you
time and effort.

I n this series of articles, we’ll take


you through installing and using LT-
spice, a free, easy-to-use and yet very
simulations can match the real-world
behaviour of a circuit. Note that
accurate simulation does rely upon
dows in this tutorial series because it’s
free, easy to install and use and most
importantly, is supplied with a fairly
powerful circuit simulation package. accurate component models and these large and mostly complete library of
Once you’re familiar with LT- are not always available. component models so that you can get
spice, you can draw up a circuit and Simulating a circuit starts with up and running right away.
start simulating it. Testing circuits in drawing it. During this process you A component model defines its char-
LTspice is a lot cheaper and safer will place component symbols on a acteristics. For example, each type of
than building them – if you blow up sheet and “wire them up”. transistor has a different curves for
components in LTspice you don’t You will then need to tell the simula- Vbe, Vce, hfe, maximum voltage and
have to buy new ones! Just modify the tor the type code of each component so current and so on. The model provides
circuit and try again. that it can select an appropriate mod- coefficients so that the simulated com-
Besides just figuring out whether a el. In many cases, for components like ponent behaves similarly in these re-
given circuit will do what you expect, resistors, capacitors and inductors, spects to an average, real component.
you can also use SPICE (which stands totally realistic behaviour is not To start off, download the latest
for Simulation Program with Integrated terribly important and you can sim- version of LTspice from www.linear.
Circuit Emphasis) to determine cer- ply use a default “ideal” component. com/designtools/software/ It's avail-
tain performance parameters such as To get accurate results with devices able for 32-bit or 64-bit Windows 7,
stability, efficiency, distortion, noise, like transistors and diodes, you would 8 or 10; there is also an older version
reaction time, overshoot, frequency re- be better off picking one of the avail- available for Mac OS X 10.7+. Sim-
sponse, power consumption and dis- able component models which ex- ply download the executable file, run
sipation, and so on. Throughout this actly matches the part you intend to it and follow the prompts to install it.
series we’ll show you examples of how use, or at least has similar characteris- It’s a straightforward process. Once
to calculate all of these parameters. tics. We’ll discuss this aspect in more installed, run the program and you
While SPICE isn’t perfect and may detail later. will see a blank window like in Fig.1.
sometimes fail to simulate some com- Now select the “New Schematic”
plex analog circuits reliably, it is quite Installing LTspice option from the “File” menu. Not
surprising how close the results of We’re going to use LTspice for Win- much will appear to have changed

38  Silicon Chip siliconchip.com.au


Toolbar Icons
Concentrate your gaze on the right-most section of the toolbar, blown up in Fig.1. From left-to-right, the buttons are:
Wire – connects two or more component pins
Ground – place a ground (0V) symbol on the circuit
Label Net – assign a name to a “net” (more on that later)
Resistor – place a resistor in the circuit
Capacitor – place a capacitor in the circuit
Inductor – place an inductor in the circuit
Diode – place a diode in the circuit
Component – place something else in the circuit, such as a transistor, IC, regulator, voltage or current source, etc
Move – move something around in the circuit diagram
Drag – same as Move, but keeps any wire connections to the selected component(s) intact
Undo/redo – revert the last change to the circuit, or reinstate it
Rotate component – rotates the selected component/components by 90°
Mirror component – flips the selected component horizontally
Text – add text to the circuit diagram
SPICE Directive – add an instruction to the circuit diagram which tells the simulator how to behave

Fig.1: how the LTspice window looks just after creating a blank simulation. The toolbar at top has been blown up to
show the important buttons, which are (from left-to-right) Wire, Ground, Label Net, Resistor, Capacitor, Inductor,
Diode, Component, Move, Drag, Undo, Redo, Rotate, Mirror, Text and Spice Directive.

but you are now ready to start drawing in square brackets). Scroll to the right Having read that, right-click on the
your circuit. First though, it’s best to and click on “voltage”, then “OK” (or V1 element you have placed and then
give it a name. Select “Save As” under just double-click “voltage”). click Advanced. You can now select
the “File” menu, type in “tutorial1” Click somewhere in the blank circuit SINE from the list on the left, and
and press Enter. to place your first voltage source. This enter 0V for DC offset, 325V for
Chances are that it will say that you will be simply shown as a circle with Amplitude (this is the peak value; not
don’t have permission to save the file positive and negative symbols inside, RMS), 50Hz for the frequency and
into the “C:\Program Files” directory corresponding to the two output ter- leave the rest blank.
and it will ask if you want to save it in minals. Note that a voltage source will Units in these values are optional,
the User folder instead. That’s a good always take the same form, whether it however, for clarity it’s usually best to
idea, so say Yes and then press Enter is intended to produce AC or DC. include them.
again to save your file. Now right click your mouse or press Click OK and the circuit updates to
We’ll now draw up a simple mains escape, since we only want one volt- include these parameters. Now click
power supply circuit. But first, let’s age source for now. on the Ground button in the toolbar
look at the toolbar at the top of the This is one of the most fundamen- and place a ground symbol directly
window. This is important since you tal parts of a circuit to simulate; the below the "negative" end of your volt-
will be using these buttons a lot. The voltage source can generate AC, DC, age source.
description of each icon in the tool- both AC and DC or a function such as You need to define 0V somewhere
bar is under “Toolbar Icons” at the top a sinewave or pulse train and is used in the circuit if you want to simulate it
right of the next page. to feed other components in the cir- and this (effectively, the incoming Neu-
You’ll find that you will need to cuit. Voltage sources can be combined tral line), is as good as anywhere. As
use nearly all these icons when draw- in various ways. before, right-click your mouse or press
ing up the circuit you want to simu- escape to stop placing components.
late. We’re going to start by creating a Voltage source mode setting Now use the Wire tool (left-
source of 230VAC. There are three different kinds of most on the section of the toolbar
Click on the Component button voltage sources and we need to use the described above) to draw a wire
(which looks like a logic gate). You right one to simulate 230VAC mains; between the negative end of the voltage
will then be presented with a list of refer to the panel titled “Simulation source and the ground symbol. Click
components and folders (which are Types” for an explanation. at one end, then the other, then right-

siliconchip.com.au June 2017  39


Simulation Types
There are two common types of sim-
ulation you can perform, plus several
other less common types. The two
most common types are “transient”
and “AC”.
A transient simulation is essen-
tially equivalent to hooking an oscil-
loscope up to various points in the
circuit and then freezing its display
to examine how the voltages and cur-
rents vary over time. An AC analysis
is more like connecting a spectrum
analyser with tracking generator up
to a circuit.
AC voltage sources in SPICE are
primarily useful for AC analysis. For
transient analysis, you need a com-
bination of DC voltages or “function”-
based voltage sources which are gen-
erally one of the following: PULSE,
SINE, EXP (exponential), SFFM (sin-
gle frequency FM) or PWL (piece-
wise linear).
Basically, if you want an AC voltage Fig.2: the Edit Simulation Command dialog comes up the first time you select
source in a transient analysis, you use the Run option from the Simulate menu. Select the simulation type from the
the SINE function. If you try to use an tabs at the top and then fill in the details below. For a Transient analysis,
AC voltage source in this situation, the most important ones are: Stop Time; Time to Start Saving Data; and Skip
you’ll find it won’t do anything useful. Initial operating point solution.

click or press escape to stop drawing wave is nowhere near as clean as this following instructions, refer to Fig.4
wires. Note that if at any point you but it’s a good start! to see how the finished circuit looks.
make a mistake, you can press F9 or Note the text reading “V(nc_01)” at Let’s start by adding a capacitor con-
click the Undo button on the toolbar the top. This indicates that the green nected to the 230VAC “positive” ter-
to revert to the previous state. trace is showing the voltage at the node minal (effectively mains Active). Click
Now we can run the simulation for labelled “nc_01” which is a name au- somewhere inside the circuit diagram,
the first time. Select the “Run” option tomatically generated for this part of then click the Capacitor button in the
under the “Simulate” menu. As this is the circuit, as we have not provided toolbar and place the capacitor above
the first time, you will need to set up our own name yet. the voltage source. Right-click the
the simulation conditions, using the Hold down the CTRL key on your capacitor to set its Capacitance value
dialog which appears (see Fig.2). keyboard and click on this label. to 470nF. Set the voltage rating to 400V
“Transient” is the default simu- You will get a dialog box showing (peak) at the same time and the RMS
lation mode (tab) selected so all information about the “trace” in- Current Rating to 250mA.
you need to do is enter a Stop Time cluding the start and end times, the Use a similar process to add a resis-
(let’s use 100ms) and then click OK. average (which is very close to zero, tor to the right of that capacitor and
A SPICE Directive automatically as it should be) and the RMS val- set its value to “10Meg”. Note that one
appears on the circuit, which reads ue which is just under 230VAC; of the traps when using SPICE is that
“.tran 100ms”, and you will find a black exactly what we wanted. You can now “10M” would be interpreted as “10m”
box appears at the top half of the screen, dismiss this dialog. (ie, 10 milliohms) so you need to write
with the circuit shrinking below. By the way, if you want to change it with “Meg” on the end. You can set
This is our virtual scope dis- the parameters later, you can right- the tolerance to 5% and power rating
play. Move your mouse cursor click on the “.tran” directive to to 1W at the same time.
down to hover just over the little re-open the simulation dialog. Now use the Wire tool to wire the two
square box at the positive end of the components up in parallel and connect
voltage source in the circuit diagram Building the circuitry the common bottom end to the voltage
below and the mouse cursor should Note that if you already know how source. Add a second resistor, in series
change to look like a probe. Click to build a circuit in LTspice, you can with the capacitor/resistor combina-
there and you should get a display download the tutorial1.asc file from tion, and set its value to 470 (ohms),
like Fig.3. the Silicon Chip website and skip to tolerance to 5% and power rating to 1W.
This shows our simulated mains the next cross-heading. The next step is to add two diodes
voltage. Of course, the real mains sine- If you find yourself confused by the to form a half-wave rectifier. Click

40  Silicon Chip siliconchip.com.au


on the Diode tool in the toolbar, then
move the mouse down into the circuit.
You will notice that if you place it, its
cathode will face towards the bottom
of the circuit but we want it at the top.
So before placing it, move the mouse
back up to the toolbar and click the ro-
tate button twice (note that this but-
ton will be disabled before moving the
mouse down into the circuit area, so af-
ter clicking the diode button, you need
to move it down and then back up).
Now place the diode above and to
the right of the existing components.
Right-click the diode symbol, which
is currently configured as a generic
(ideal) diode, and click the “Pick New
Diode” button. You will now get a list
of the diode models built into LTSpice,
which includes silicon/switching/
Rectifier (standard) diodes, fast recov-
ery diodes, schottky diodes, zener di-
odes, LEDs and transient voltage sup-
pressors (TVS/varactor). Fig.3: the result of our first Transient simulation, showing the voltage at the top
Scroll down to where the “silicon” of voltage source V1 over a 100ms period. Note that the 325V figure selected
type diodes are listed and click on the defines the peak voltage, not RMS and that several parameters have been left
blank and so default to zero, including the DC offset and phase values.
MURS120 which is roughly equivalent
to the 1N4002, then click OK. If the
placement of the diode is not ideal,
click the “Move” button in the tool-
bar (or press F7 on the keyboard) and
click on D1 to move it to a better spot.
Now we need a second, identical di-
ode so the easiest solution is to clone
the one we have. Press F6 on the key-
board, then click on D1 and place the
new diode (D2) directly above it. Join
the adjacent anode and cathode pins,
then connect the free end of the 470W
resistor to this junction, all using the
Wire tool. Connect the free anode at
the bottom to ground, as we did with
voltage source V1.
Now we need a zener diode. You can
clone one of the two existing diodes,
placing it immediately to the right of
voltage source V1, then right-click on
and select “Pick New Diode” to change
its type. Scroll down to the zeners
and you will find multiple 15V zener
diodes in the list (look for 15 in the Fig.4: now we’ve built up a basic mains power supply with a simple resistive
Vbrkdn(V) column). Pick the KDZ15B load and can observe how the main 220µF filter capacitor charge increases
every 20ms during the peak of each mains cycle. We can see that D3 (a 15V
as this is a 1W type, then click OK.
zener) begins to conduct after around 350ms, but some ripple remains.
Move D3 if necessary, to avoid labels
from overlapping.
Now connect the zener’s anode (bot- lating a power supply load, in paral- to 500ms, then re-run the simulation
tom end) to ground and the cathode lel with both. When finished, your cir- (“Run” option under the “Simulate”
(top end) to the free cathode of the rec- cuit should look similar to that shown menu). Click on the “wire” at the cath-
tifier diode above. Having done that, in Fig.4. ode of D3 to view the resulting volt-
add a 220µF 25V capacitor in parallel age. Your result should be the same as
with D3, with a 500mA ripple current Making some measurements shown in Fig.4.
rating and ESR of 0.1 (ohms). Also Right-click on the “.tran 100ms” As you can see, it takes around
add a 1.5kW 10% 5W resistor, simu- directive and change the Stop Time 370ms from the application of mains

siliconchip.com.au June 2017  41


in “V(VIN) * -I(V1)” and click OK. A
new trace will appear showing the in-
stantaneous power being drawn from
V1. V(VIN) refers to the voltage at the
node labelled VIN and I(V1) refers to
the current through voltage source V1.
“*” is the multiplication operator so
giving us the product of the two.
The minus sign before I(V1) just sets
the polarity of the result and is some-
thing you’d normally need to deter-
mine experimentally. You will see that
the instantaneous power goes positive
and negative at different times in the
mains cycle.
This is because sometimes, current
flow into C1 is in-phase with the mains
voltage and sometimes it is out-of-
phase. In other words, there are times
when power is flowing from the mains
into C1, and times when it is flowing
out of C1 and back into the mains.
If you CTRL-click the expression
Fig.5: not only can we see the voltage across C2 but now we can also observe at the top of the window, you will
the current drawn from the mains as it charges – all without having to wire up see that the average is 712.21mW and
a single component and without any test equipment! One of the benefits of using its integral (ie, total energy consumed
SPICE is how easy it is to make multiple voltage and current measurements.
in the 500ms window) is 356.11mJ.
But note that this includes the time
power before the 220µF capacitor is on the right-hand side of the plot, al- that C2 is charging. So to get an ac-
fully charged to 15V. You can drag a lowing you to see that the peak cur- curate result, right-click on the “.tran
box around the waveform at the top of rent through C1 is just below 50mA. 500ms” directive and change the
the screen to zoom in and examine it You can CTRL-click the label at the “Time to Start Saving Data” to 400ms,
in more detail (right-click and select top of the display to read off the RMS then re-run the simulation. The aver-
“Zoom to Fit” or press CTRL+E to go current which is 33.5mA (see Fig.5). age is now 783.93mW, which repre-
back to the normal view). sents a steady-state value, and you
Once zoomed in, you can see that Efficiency calculations will notice that the waveform is con-
the peak voltage across the capacitor The efficiency of this circuit is the sistent across the five mains cycles
is clamped to around 15.35V and with power delivered to the load (R3) di- (100ms) shown.
the 1.5kW load, the minimum voltage vided by the power drawn from the By the way, if you want to change
is around 14.85V, giving a ripple of mains (V1). In both cases, we can the expression used to plot the power,
around 0.5V. compute power as V × I. We could you can do this by right-clicking where
You can make reasonably accurate use V2 ÷ R for R3 but then we could it’s shown at the top of the window.
measurements by placing the mouse need to change the calculation if we Now, to compute the power con-
cursor over the trace and then reading changed the value of R3, and it would sumed by R3, right-click in the top
the time and voltage values shown in also make it harder to change the cir- window (or press CTRL+A) and en-
the bottom-left corner of the LTspice cuit to a more realistic load. ter the similar expression “V(VOUT)
window. Also, once you’ve zoomed in, To make it easier to calculate both * I(R3)”. If you CTRL+click the new
if you CTRL-click the V(n001) text at power figures, let’s label the two volt- expression which appears at the top
the top of the screen, it will calculate ages. Click the “Label Net” button in of the window, you will see that the
average and RMS values for the time the toolbar and type in “VIN”, then average power is 152.69mW (see
period displayed, in this case, both press OK. Place the label at the junc- Fig.6). This is in line with what you’d
around 15.124V. tion of V1, C1 and R1. expect from 15V across a 1.5kW resistor
Now click the mouse in the circuit Similarly, label the junction of D1, (V2 ÷ R = 15 x 15 ÷ 1500 = 150mW).
window at bottom and move the cur- D3, C2 and R3 as “VOUT”. Press the So we can calculate the efficiency
sor over capacitor C1. You will note DEL key on your keyboard and click as 152.69mW ÷ 783.93mW = 19.5%.
that the cursor changes to what looks on the labels at the top of the simula- That’s pretty lousy! That means that
like a clamp meter. Click here and the tion output to delete the traces, then 80.5% of the energy drawn from the
current through this capacitor will also re-run the simulation. mains (630mW or so) is being dissi-
be shown in the top window. Note Now right-click on the (now blank) pated elsewhere in the circuit, just
that it is essentially symmetrical and top half of the window and select “Add turned into useless heat. Luckily, we
looks like a sinewave with zero-cross- Trace” (or, having clicked in this sub- can use LTspice to figure out where
ing artefacts. window, press CTRL-A). It will prompt and improve the situation.
Note also that a new y-axis appears you for “Expression(s) to add”. Type First, let’s see how much power is

42  Silicon Chip siliconchip.com.au


Helping to put you in Control
Capacitive Oil Level Sensor
1000mm 4-20mA out.
Level Sensor for non conductive
liquids such as oil and diesel.
The 1000mm probe can be cut
to suit tank depth and easily
calibrated.
SKU: FSS-232
Price: $449.00 ea + GST

60W Ultra Slim DIN Rail Supply


Meanwell HDR-60-12
measures only 53W x 90D
x 55Hmm it supplies 12VDC
45A.
SKU: PSM-0181
Price: $45.00 ea + GST

H685 Series 4G Cellular Router


H685 4G router is a 4G
cellular serial server and
Ethernet and Wi-Fi gateway.
It can act as an RS-232
Fig.6: plot of the product of the input voltage and current; LTspice automatically serial cable replacement over
shows the result in watts and changes the Y-axis to suit. The area enclosed by the the mobile phone network
or as a serial server on the
power curve below the horizontal axis is smaller than that above, with the net internet. It also shares the cellular internet
power consumption shown in the average (in the box to the right of the circuit). connection out over an RJ45 port and Wi-Fi.
SKU: OCO-002
Price: $495.00 ea + GST
dissipated in D3, the zener clamp di- reduce its dissipation and improve the
Waterproof Digital Temperature
ode. We can simply plot the expression overall efficiency but we need to check Sensor
“V(VOUT) * I(D3)” and integrate it as that this won’t cause any problems and DS18B20 digital
before, to yield a figure of 73.282mW. also quantify just how much of an im- thermometer comes with
waterproof 6 × 30 mm probe with 3 metre
Well, that’s barely more than 10% of provement we can achieve. cable. -55 to 125 °C range with ±0.5 °C
accuracy from -10 to 85 °C.
the energy being wasted, so that isn’t So let’s simulate the (almost) worst SKU: GJS-003
the culprit; we may still be able to case, where the circuit is connected Price: $16.00 ea + GST
make some tweaks to reduce this fig- to the mains at the peak of 325V and Pressure Transducer 0 to 25 Bar
ure and improve efficiency but let’s C1 is discharged, and see how low we Firstrate FST800-211
pressure sensor features
figure that out later. can make the value of R2 before we IP67, 3 wire connection,
What about R2? To calculate the risk damaging something. 0-5VDC output ¼” BSP
process connection. ±0.3%
voltage across that, we need to label To do this, rightclick on the body F.S. accuracy. 0 to 25 Bar.
the wires (nets) at both ends. Let’s label of V1 and enter 90 for “Phi(deg)”. We SKU: FSS-1530
Price: $159.00 ea + GST
the one junction of C1/R1/R2 as “VA” also need to make two changes to the
and the junction of R2/D1/D2 as “VB”. simulation directive, which we can Heating/Cooling Self Adaptive PID
We can then plot the expression access by right-clicking on the “.tran Controller
1/16 DIN Panel mount Heating
“(V(VB) − V(VA)) * I(R2)”, in other 400ms 500ms” text. and Cooling self adaptive PID
words, the difference between the volt- First, change the “Time to Start Sav- controller. Features universal
input 2 Relays, 2 Digital
age at points VB and VA (ie, the voltage ing Data” back to 0ms so that we can Input/Output and 24 VAC/DC
across R2) times the current through R2. see the initial conditions, then also powered.
SKU: PID-048
Integrating this gives us a figure of tick the “Skip Initial operating point Price: $299.00 ea + GST
529.33mW. Adding this to the pow- solution” box towards the bottom.
er dissipated in D3 gives a result of This tells the simulator to start with Eight 12VDC Relay Card
Eight-way relay card on DIN rail mount
602.6mW, explaining over 95% of the all capacitors and inductors fully dis- allows driver direct connection to many
power lost in the circuit (the other ~5% charged (although you can specify an logic families, industrial sensors (NPN or
PNP) dry contacts or
is probably in R1). So to improve the initial charge on a case-by-case basis voltage outputs. Relay
efficiency we need to do something if necessary; we’ll explain how to do output load 10A(240AC)
SKU: RLD-128
about R2. this in a future instalment). Price: $109.95 ea +
GST
Re-run the simulation, clear all the
Improving the efficiency traces and plot the current through C1;
For Wholesale prices
R2’s purpose is to reduce the inrush you can achieve the latter two sim-
Contact Ocean Controls
current into C1 when the circuit is first ply by moving the mouse cursor over
Ph: (03) 9782 5882
connected to the mains, especially if C1 until it turns into the clamp sym-
oceancontrols.com.au
that happens to be in the middle of a bol and then clicking twice. The first
Prices are subjected to change without notice.
cycle. If we reduce R2’s value, that will time it will show the current plot for

siliconchip.com.au June 2017  43


Consider that in a real circuit, this
would be an X2 capacitor which is
designed for direct connection across
the mains supply with no real cur-
rent limiting whatsoever so it should
be able to tolerate a high inrush
current. So on that basis, let’s reduce
R2 to 68W, giving an inrush current
of just under 5A.
The only other components which
need to handle this current are R2
(which should be OK given how brief
the spike is) and D1/D2 (which will
handle much larger spikes as long as
they’re short or non-repetitive).
At the load end, how much of the
initial spike will be borne by D3 and
C2 depends on the polarity of the ap-
plied mains voltage (ie, whether D1
or D2 conducts) and C2’s ESL (equiv-
alent series inductance). Typical ESL
of a moderately-sized electrolytic ca-
pacitor appears to be pretty low at
Fig.7: by zooming into the early part of the current trace for C1, we see the around 1nH so C2 should safely ab-
inrush current is around 700mA for a fraction of a millisecond. The “uic” on the sorb the brief initial spike, but even
end of the “.tran” directive is critical; it stands for “use initial conditions” and
if it doesn’t, it should not pose much
without it, capacitors and inductors start in a “steady state” condition.
difficulty for D3.
We can now re-run the simula-
tion, adjusting the time to start sav-
ing data back to 400ms and calculate
the steady-state figures as input pow-
er: 327mW, output power: 152.7mW,
efficiency: 46.7%. That’s a lot better
but still not great.
Let’s look again at the power con-
sumed in D3, the zener diode. It’s virtu-
ally identical to before at 73.75mW but
now this is around 50% of the power
loss. We can reduce this by lowering the
value of C1, so that it doesn’t deliver
more current than the load requires and
D3 will then only conduct rarely (eg, if
the mains voltage is higher than nomi-
nal or the load is lighter than expected).

Parameter stepping
Now we consider whether changing
the value of C1 will affect efficiency. It
will because if the value is too high, D3
will shunt more of the current coupled
Fig.8: parameter stepping is a valuable method for optimising component values. through it, effectively wasting power
Here we can see how varying the value of C1 between 220nF and 470nF affects whereas if the value of C1 is too low,
circuit operation. You can also use this method to vary the simulated ambient
the voltage across D3 will not rise to
temperature or to see how component tolerance affects circuit operation.
the desired value of ~15V.
What we really want to do to
C1 and the second time, it will erase figure simply by assuming that C1 is figure out the ideal value is look at the
all the other traces except for that plot. initially a short circuit and doing the effect of changing the value of C1 with
If you zoom into the first few milli- calculation 325V ÷ 470W = 0.7A. This everything else the same. We can
seconds you can see that the peak cur- suggests that whatever we do to reduce do this by stepping its capacitance
rent is around 700mA but this drops the value of R2 is inevitably going to through different values. To do this,
very rapidly, to just a few milliamps af- increase the inrush current but the click on the SPICE Directive (“op”)
ter 1ms or so (see Fig.7). In retrospect, simulation shows that this is really button in the toolbar and then type
we could have calculated the 700mA very brief as C1 rapidly charges up. in “.step param CV list 220nF 330nF

44  Silicon Chip siliconchip.com.au


470nF”. This creates a parameter the ease of simulating such a circuit.
called “CV” which steps through three Compare this to the difficulty of meas- Next Month
different capacitance values. uring it, especially when you consid-
Now change the value of C1 from er it would be directly connected to Modelling relays in SPICE is a
470nF to {CV}. Re-run the simulation, the mains! little tricky but it can be done, as
with a start time of 0ms and finish time we will demonstrate by building
of 1500ms and plot VOUT. The result Apparent power consumption a fairly realistic relay model next
is shown in Fig.8. There are a couple of final issues to month. We’ll also get into some
Unfortunately, LTspice doesn’t pro- discuss regarding simulating this cir- more advanced techniques that
vide a colour-coding legend but it’s cuit. Firstly, our method of integrating are possible with LTspice.
fairly obvious that the green curve is the instantaneous power gives us the
for C1=220nF, blue for C1=330nF and real power consumption of this circuit,
red for C1=470nF. 220nF is too low as as would be measured by your power Secondly, there’s nothing to stop
VOUT doesn’t even reach 10V, while meter (and which would be used to you from taking the simulation fur-
with both 330nF and 470nF it reach- charge you for electricity). ther and actually drawing up a real
es the same final voltage, albeit after But note that the RMS current drawn load instead of using resistor R3. This
a different time delay. from the “mains” (V1) is now 23.65mA would give a more realistic depiction
So it seems that 330nF is probably with an RMS voltage of 230VAC. That of the voltage regulation of this power
close to the ideal value. Let’s set the gives an apparent power consumption supply circuit in the face of changing
capacitance value of C1 back to 330nF, of 0.02365A x 230VAC = 5.44W. load demands.
delete the step directive (press DEL on That tells us that this circuit has a For example, this sort of circuit is
the keyboard, then click on the direc- very low power factor. In fact, we can commonly used to power a relay, ei-
tive) and then re-run the efficiency calculate it, it’s simply the real input ther to act as a mains timer or some
calculations. power of 219mW divided by the ap- sort of load-detecting switch.
parent input power of 5.44W, giving a Actually, if you look at our SoftStart-
Final results power factor of 0.04 or 4%. Note that er in the April 2012 issue, Soft Starter
After changing the “Time to Start because this is so low, many domestic for Power Tools in the July 2012 issue
Saving Data” back to 1400ms and using power meters would have trouble giv- and Mains Timer for Fans and Lights
the same steps as before, we can now ing any kind of reading at all and the project in the August 2012 issue, you
compute the input power as 219mW power reading could range from zero will see just this type of circuit.
and the power consumed by the load all the way up to several watts. In those cases, the load current de-
at 151.94mW, only a tiny bit lower The low power factor is due to the pends heavily on whether the relay
than before, giving an efficiency figure fact that so much of the energy drawn is energised and it’s acceptable for
of 69.4%. That’s pretty reasonable for from the mains goes into simply charg- the supply voltage to drop once the
such a simple circuit, and with a vir- ing up C1 and this is returned later in relay has latched, as a lower voltage
tually identical load voltage. the cycle, so the power moving into is required to hold the relay than to
So we’ve barely sacrificed any per- and out of the unit via the mains sock- switch it initially. So further simula-
formance for what is a pretty large im- et is much higher than the actual net tion would definitely help optimise
provement in efficiency, all thanks to consumption. such a circuit. SC

Radio, Television & Hobbies: the COMPLETE archive on DVD


This remarkable collection of PDFs covers every issue of R & H, as it was known from the beginning (April
YES!
A 1939 – price sixpence!) right through to the final edition of R, TV & H in March 1965, before it disappeared
MORE THAN URY
QU ARTE R CE NT forever with the change of name to EA.
ONICS
OF ELECTR
HISTORY! For the first time ever, complete and in one handy DVD, every article and every issue is covered.
If you’re an old timer (or even young timer!) into vintage radio, it doesn’t get much more vintage than this.
If you’re a student of history, this archive gives an extraordinary insight into the amazing breakthroughs made
in radio and electronics technology following the war years. And speaking of the war years, R & H had some
of the best propaganda imaginable!
Even if you’re just an electronics dabbler, there’s something here to interest you. ONLY
Please note: this archive is in PDF format on DVD for PC. Your computer will need a DVD-ROM
or DVD-recorder (not a CD!) and Acrobat Reader 6 or above (free download) to enable you to
view this archive. This DVD is NOT playable through a standard A/V-type DVD player.
62
$

+$10.00 P&P
00

Exclusive to:
SILICON Order now from www.siliconchip.com.au/Shop/3 or call
(02) 9939 3295 and quote your credit card number.
CHIP
siliconchip.com.au June 2017  45
SERVICEMAN'S LOG
Fixing the food processor that wouldn't Dave Thompson*

I’m not sure whether it is Sunspots, Murphy’s Items Covered This Month
Law or just plain bad luck, but it appears • A Serviceman’s kitchen
there is a conspiracy among our household • Brownout protection for a TV
appliances to all fail around the same time. • TEAC HDR PVR
Sitting among those appliances that fail is one *Dave Thompson runs PC Anytime
of our food processors which is only a few years in Christchurch, NZ.
Website: www.pcanytime.co.nz
old. Email: [email protected]

Last month it was the vacuum clean- road; not impossible, as demonstrated plastic lid then clamps securely onto
er. Then a fan heater decided to stop by many car-crazy middle-eastern the glass bowl and sitting on that lid
fanning and heating. Even my own YouTubers, but for the rest of us it is is a chrome-and-black plastic housing
computer has been increasingly throw- definitely pretty tough. which contains the motor and drive
ing up those cloying “your PC has en- Now our brand-new dishwasher, assembly.
countered a problem so we have shut purchased when we renovated the This spins a centrally-mounted twin
it down” screens. kitchen nine months ago, has started blade system down through a hole in
Between crashes, it often displays making odd noises and during the last the lid via a splined drive shaft. This
weird on-screen artefacts, such as col- few nights, the LED display has been machine is so powerful, you could
oured blocks of pixels and very thick randomly flickering between showing stuff the bowl full of ice cubes and
black lines appearing randomly. As I all 8's or nothing, to the time remain- with a few short bursts of the motor,
write this article, it is as if someone is ing and back again. turn them into slush.
trying to redact what I am writing in I can hurry things up with a light The four curved and razor-sharp
real time. Maybe the CIA really can tap on the door, so it looks as if some- blades make small work of anything
hack into our homes via our smart thing is not quite making contact some- in the bowl, and while Mrs Service-
fridges (or is it ASIO?)! Hmm, we don't where in the electronics behind the man and I aren’t great foodies (or cooks
have a smart fridge... door panel. for that matter), when the urge takes
Obviously, the computer has a prob- That’ll mean a trip to the repair us, we like to have decent tools with
lem and I suspect my graphics card’s agent; I have to resist the temptation which to do the job.
VRAM is faulty, though this is a wasted to go searching for the fault myself. For The problem with this thing was
diagnosis because I can’t actually do starters, I haven’t the room to pull it to that it no longer powered up. There
anything about it except replace the bits in my workshop and besides, Mrs is no obvious on/off switch; instead,
card anyway. I’m not about to start Serviceman wouldn’t be too keen on it has an interesting push-switch
stripping memory chips off the thing me voiding the warranty! arrangement. On the very top of the
and replacing them. motor housing is a black plastic cap
Considering that it cost a small for- The curse takes another victim that activates a switch when pushed
tune and was the single most expen- Now to top it off, one of our food with the palm of one’s hand.
sive component in this system, I’m not processors has decided it wants to However, for (I assume) safety rea-
that happy it has decided to fail just stop processing. This appliance is sons, there is another switch activated
outside of warranty. I would expect one of the better and most-used of when the whole assembly is pushed
more than 16 months out of a high-end our kitchen tools so having it give up against the lid of the bowl. This way,
graphics card, but that’s how it often is a bit of a curse, as it is a few years the motor will only run when sig-
goes with high-performance hardware. old now and I’m reasonably sure they nificant pressure is put on the whole
The irony is that I haven’t had time don’t make them any longer, so we motor assembly and both switches
to play the games I originally obtained can’t just go out and buy another one are pushed.
it for as I am too busy doing far more (which is I’m sure what the manufac- It sounds more awkward than it
mundane things, such as renovating turer intended). Then I remembered; really is; in use, it is quite natural to
workshops, which is proving surpris- I’m a serviceman! This shouldn’t be a push down to start operation. It also
ingly difficult to do whilst actually problem for the likes of me! makes sense because the motor part
working out of them. This particular mixer has a solid of the appliance isn’t held down by
It’s a bit like changing the tyre on Pyrex glass mixing bowl forming the anything other than the operator’s
your car while driving it down the bottom half of the appliance. A tough hand pressure.

46  Silicon Chip siliconchip.com.au


The splined output shaft of the thing else). However, this is a classic going. That’s not only hugely wasteful,
motor assembly locates into a circu- example of the way everything is going it’s bordering on criminal.
lar moulding in the thick plastic lid, these days. We have become a consum- People these days buy new printers
which is well-fixed to the glass bowl able, throw-away society, and this is rather than buy hyper-expensive car-
underneath, but other than that hand very apparent in the kitchen. tridges for their old one. Tablet owners
pressure, the motor assembly is free chuck a tablet with a broken digitiser
to move, so it makes sense to hold it Start of Serviceman’s rant rather than repair it, even though in
tightly when it is running and makes By comparison, my mother still uses most cases it costs far less than a new
even more sense to require downwards a mixer she bought in the seventies. one. And people chuck food processors
pressure before power is switched to Sure, it’s had the motor reconditioned because a part that would cost just a
the motor. a few times and there are a few minor few bucks to manufacture gives out and
Except for the fact that no mat- cracked mouldings here and there, because the parts aren’t there to repair
ter how hard I pushed on it, nothing but the point is that it was designed them, and who can blame any of them?
happened. to be repaired and there are still parts Most manufacturers today are only
So I had to assume that either some- available for it. concerned with moving as many units
thing was wrong with the switch ar- Most of the high-end appliances you as they can and don’t give a toss what
rangement or the motor had burned buy today don’t have anything like happens to their products once they
out. I was hoping it wasn’t a dead mo- the parts backup these older brands break down; that becomes someone
tor, because then we’d likely have to have and in 40 years they’ll be part else’s problem.
bin the whole thing (although I sup- of a landfill somewhere while those Increasingly, I cannot source parts
pose we could use the bowl for some- like Mum’s will probably be still for even newer models of computers,

siliconchip.com.au June 2017  47


Serviceman’s Log – continued
ing down on the plastic cap by way of
a captive plastic rod. When the cap was
pressed, the rod was forced down onto
the spring and that flexed and pressed
onto one toggle of the power switch.
The bottom spring actuated the sec-
ond pole of the switch via a clever little
plunger arrangement; when the motor
body was pressed down onto the lid of
the bowl, a corresponding solid piece
of plastic in the lid pressed onto the
plunger, forcing it upwards into the
second spring and causing that to push
the switch’s second toggle. Only when
the two toggles were actuated would
power flow to the motor.
This was actually quite a clever
“manual” safety system, as it meant
the motor could neither be run with-
out the load of the blades to govern
its speed nor without holding it down
forcing me to look to the second-hand From the outside, there wasn’t any- quite tightly. Pressing down also en-
or refurbished market. Most punters thing much to suggest how it was held gaged the blade assembly down in the
these days accept that having to buy a together. There was a hole where the bowl and unless this was all connected
whole new anything is inevitable, the power cable entered the case but that properly, the bottom plastic plunger
collateral damage of technological pro- was about it. would not be actuated. Simple and
gression. And maybe you can’t really After having a closer look at the effective; and broken.
blame manufacturers for not wanting bottom of the motor assembly, I could The motor wouldn’t power on
to have capital tied up in parts sitting see a black plug about 5mm in diam- because the bottom spring had some-
around gathering dust on a warehouse eter that could be hiding something. how broken away from its mount,
shelf somewhere. It’s easier to just sell I used one of my favourite tools, an meaning that it simply flexed out of
more computers and let someone else old dentist’s pick, to pry the bung out the way when the plunger touched it,
create a second-hand or refurbished and sure enough, it was covering one instead of transferring that pressure
market. of those annoying safety-type screws. to the switch.
And that’s exactly what has hap- I found the right bit for my driver and I hadn’t noticed any plastic shrap-
pened; these days many companies are removed the screw. nel floating around in the case but
buying up old appliances – whether After a few attempts to pull the then again, any fragments might have
computers, food processors or wash- cover off, I concluded that something been small enough to work their way
ing machines – just to strip them down else must be holding it on. However, through the small gap for the blade
for spare parts. They know what car I couldn’t find anything, and so after shaft in the lid and fall down into the
wreckers have known for decades; that sitting and contemplating it for a few bowl below. I suppose we’ll find them
they can make good money from sell- minutes, I decided to give it a gen- when we defrost and eat the Ćevapi
ing parts rather than selling complete tle twist. my wife made when she last used the
appliances. Sure enough, the cover slowly gave blender!
This is exactly why I am so keen to way to the sideways pressure and after Now though, I would have to rebuild
get stuck into trying to repair anything turning ninety degrees, the whole mo- the support for the bottom spring so I
that breaks down. I consider it a chal- tor mechanism began to pull out from made a rough-and-ready former using
lenge to buck that wasteful philosophy the plastic case. It only came out part strips of gaffer tape stuck to the remain-
and try to keep things going for at least way as the power cable held every- der of the plastic mount. It was origi-
a reasonable lifespan. If something thing in, so I forced the cable through nally a rectangular shape and thus easy
wears out to the extent it cannot func- the stress-relieving sleeve bit by bit to recreate. To build it up, I utilised a
tion any longer, then that’s fair enough, and pulled the motor far enough out two-part compound that comes in a
but when an entire printer is junked until I could see the power switching tube and one simply slices off a small
simply because a 20-cent proprietary arrangement. amount and rolls it to mix it.
part is not made available, that is not The series-wound universal It sets rock-hard in about five to 10
OK in my book. (brushed) motor was switched by minutes, depending on the ambient
a large two-stage microswitch that temperature, so you can’t faff about
Normal service(man) resumes was actuated by two formed, plastic- once mixed. Another smaller piece of
Getting into this blender motor capped copper-coloured springs. One tape held the contact in place while
looked like it could be a challenge. of these springs was actuated by press- the pressed-in compound set and I

48  Silicon Chip siliconchip.com.au


left it overnight to be certain it was wiring and was too smart (or afraid) real head scratcher. This law is closely
properly cured. to mess about with mains wiring. Fair related to the one which decrees that
The next afternoon, I removed the enough; sometimes power tools can when a screw is dropped, it will al-
tape and tested the contact’s travel, be wired up in weird ways and just ways roll to the darkest, dustiest, most
making sure it reached and actuated randomly re-connecting spade con- inaccessible corner of the workshop!
the microswitch. It did, so I reassem- nectors and plugging it in isn’t the best A local retailer phoned me with an
bled the motor back into the case and way forward. appeal to talk with one of their cus-
with power applied, tried it a few times The first thing I did was look for a tomers who had purchased a 55-inch
to ensure it started consistently with circuit diagram for the saw. I found TV three months earlier and he was
the usual amount of pressure. Hope- plenty, all American and not much now convinced the TV was faulty.
fully, that fix will last and we’ll get a use for our configuration. I Googled The farming owner lived near the
bit more use before we have to con- the paddle-style safety switch’s part end of a long peninsula, a 60km one-
sider chucking it away. number and found a suitable usage hour drive over reasonable but quite
diagram in the datasheet, and with twisty roads.
DIY circular saw it all wired up correctly, I plugged it I was very busy with other work so
upgrade goes awry into my Variac and wound on 50VAC I was not keen to waste half a day un-
And in one of the funnier service before pushing the green button. The productively driving back and forth all
moments this month, a customer (and motor spooled up but there wasn’t that way, despite the very scenic vis-
neighbour) brought in one of those enough juice to hold the button in. tas I would enjoy on the journey. So I
bright green, plastic-bodied benchtop I added another 100 volts and away was eager to solve the problem by any
table saws that he’d stripped down in it went. other means, hoping for an installation
order to mount the guts into another Pushing the red “Stop” button or operation issue which I could talk
more solid benchtop. He’d made a switched it off as expected so I called the owner through.
nice job of swapping everything over my neighbour to tell him it was ready. I phoned the owner and he ex-
but had needed to remove the power The next day, he brought it back, claim- plained his problem. The TV would
switch connections and associated ing the motor was running backwards. completely cut out for two to three
wiring in order to get it out of the orig- Puzzled, I drew a quick wiring diagram minutes and then come on again with
inal housing. and looked at how changing the con- no sound or picture. It would repeat
After throwing away all the old nections on the switch would make this several times an evening and it
parts, he’d fixed the saw and motor the motor run in reverse. was happening virtually every night.
assembly into a recessed pocket in There wasn’t any configuration Soon after cutting out, the screen
a sheet of 32mm custom-wood. That that would do that; to swap rotation, would show the word “SAT” in the
would certainly make a far sturdier I’d have to change either the arma- top left corner. When the picture and
base for it than the original flimsy ture polarity or the field polarity with sound eventually came back on every-
plastic body and thin, cast aluminium respect to each other and as it had run thing worked normally until the next
table top. properly before, I couldn’t see how that time. He had only experienced these
The problem was that he’d chucked was all necessary. problems since purchasing the new
away the diagram he’d made of the After pondering it for a while, it hit TV. The old plasma TV had worked
me; he’d put the blade on backwards, just fine.
which of course made it look as if the I asked some obvious questions
motor was running the wrong way. such as whether the problem was oc-
After flipping it, all was well. At least curring on all sources which was re-
it was an easy fix! sponded to with a considerable pause.
So I clarified my question by asking if
Browned off by brownout protection there was anything else connected to
G. M., from Pukekohe in New the TV such as a set-top box or DVD
Zealand, recently had a call-out to a player. Yes, there was a Sky box (pause
remote location to a malfunction in a again); not sure about a DVD player.
new TV installation which proved to Now we're getting somewhere – or are
be a case of “too much protection is we? I gained the impression that he
too much”. was not sure what a DVD player was.
I am the local service agent for a He only watched Sky broadcasts.
well-known and popular range of The owner explained that he had
home entertainment products. I nor- had the Sky people out twice attempt-
mally restrict my call-out radius to ing to solve the issue and they had
about twenty-five kilometers from eventually replaced the decoder, to
base but this was one of those occa- no avail. Having eliminated that as a
sions when Murphy's Law kicked in, cause, he surmised that the problem
and a faulty TV as far away from base now had to be with the TV. He even
as I could be enticed to go became a took a photo of the malfunctioning TV

siliconchip.com.au June 2017  49


Serviceman’s Log – continued
with his phone and sent it to me so he Up until this point I had not known ble cause of the original complaint.
was obviously not completely techno- there was a sound bar. It was now I If the owner wants this feature, it can
phobic! I talked the owner through also learned that the TV and sound be readily restored, once our present
the procedure to reset the TV to fac- bar had been delivered and installed issues were behind us.
tory condition and left him to test it by the retailer at the time of purchase. Back at base I connected the new TV
that night. I knew there was power to the wall to an HDMI source and left it running.
The next morning, I phoned a cou- so I peered over the back of the TV It didn't miss a beat. That evening I
ple of satellite installers whom I knew and saw a nice new multi-outlet pow- received a text from the exasperated
and discussed the possibility of there er box there, no doubt sold by the re- owner to say the same thing was hap-
still being a decoder or satellite sig- tailer as an add-on to give protection pening with the loan TV except that
nal issue since this was not my area to the new equipment in the event of this TV displayed the "HDMI1" in the
of expertise, but neither of them were a power surge. In my opinion this is top left corner, not “SAT” which the
convinced that the problems were Sky a duplication of the sort of protection new TV had displayed.
related. However, since they had both which is now already built into most It was now that he offered a little
been in the trade nearly as long as me, equipment, but having been a retailer more information which in hindsight,
neither was prepared to lay odds that myself in a past life, I was not about I probably should have asked about
it was not a Sky reception fault. to cast aspersions on the practice of earlier. The problem only seemed to
Meanwhile, I forwarded the photo to up-selling. occur at around dinner time each even-
the TV brand’s technical manager and The box had an on/off switch on the ing and it did not happen so often later
had several discussions with him by top so I reached for it and toggled it to in the evening.
email and phone regarding the fault. the other position at which point I was Now armed with another clue in the
He seemed to be quite certain that the greeted by an encouraging green glow timing of the fault occurring, I won-
"SAT" logo appearing on the screen was from beneath the switch button and dered whether this could be a power
not something being generated by the now everything was working. supply problem after all, but the owner
TV so the problem had to be elsewhere. I quizzed the elderly couple about wasn't convinced of this since despite
We were running out of elsewheres. how the switch could have been living at the end of the road at the tip
The last straw caused the frustrated bumped off since there were obviously of a long peninsula, it couldn't be a
owner to send a curt text to my phone no children around at which point the power voltage drop because the lights
a couple of days later that 'Now every- lady admitted that one of her cats had remained bright.
thing had cut out, nothing was work- chased a mouse behind the TV the pre- Now clutching for straws and want-
ing – I'll pay for you to call'. I phoned vious day and must have jumped on ing to resolve the problem before I
him and waited while he fetched an the switch. This seemed quite plausi- headed away for a brief break in a few
electric drill to plug into the wall to ble since the switch button was quite days, I connected the new TV in my
test that the power was on. The drill large and the switch took little effort shop to a variable transformer just to
whirred into life so I said I'd think to operate. eliminate that from the mix of possi-
on it and let him know. He was slow- So, the immediate no-power prob- bilities. I slowly reduced the voltage
ly convincing me that the TV was to lem was solved but I was fairly sure from the normal 230VAC and I was
blame and had somehow now caused this had no connection (pun intended) surprised that the TV worked perfectly
everything to fail. with the original complaint so I carried down to 80VAC, at which point it cut
I talked with the brand's head tech- on with swapping the TVs and after the out. As soon as the voltage came up a
nician again and he agreed to cover loan TV was connected and working, bit, the set burst back to life without
the considerable cost of a call out if it I checked some of the menu settings so much as a hiccup. I reset the TV a
proved to be a TV fault. I phoned the on the decoder and sound bar. These second time and adjusted some of the
owner and he was happy for me to call were both connected with HDMI ca- menu settings.
on the basis that if it was not the TV at bles to the TV. The next day was a public holiday.
fault then he had to bear the cost. He I’ve experienced some odd behav- Despite that and being so keen to get
was quite convinced that the TV was iour in the past by allowing such appli- on top of the problem, I loaded the
faulty, so was confident that he would ances to talk to each other. I went into owner's TV into the van. I included
not be writing a cheque. the menu of each device and switched the variable transformer and set off
I called the next day armed with a off this interactive feature to eliminate with some trepidation, knowing that
small loan TV of the same brand with arguments between them as a possi- if I did not solve the problem this time
a twofold job for it – first, to pacify
the owner by leaving him something
to watch and second, to act as a test
unit while I checked his new TV at
Servicing Stories Wanted
Do you have any good servicing stories that you would like to share in The Serviceman
my shop.
column? If so, why not send those stories in to us?
The first thing I noticed when I ar-
rived at his home was that there was We pay for all contributions published but please note that your material must
indeed no power. The TV, decoder and be original. Send your contribution by email to: [email protected]
sound bar/sub-woofer were all dead. Please be sure to include your full name and address details.

50  Silicon Chip siliconchip.com.au


around, I would have to admit defeat. nological feature of the power outlet I suggested the owner have an elec-
Who would want to pay me for the multi-box which I did not realise it trician check his power supply as
many hours I had already invested in had. In any other location, the pro- such a severe voltage drop could sug-
this job if I couldn't provide a solution? tection would most likely not have gest a high resistance in one of the
On arrival, I connected the vari- tripped until a real and rare brown- three phases supplied to the property,
able transformer to the wall socket out event happened and I would have which in turn could pose a fire risk.
and the multi-outlet box in turn to remained blissfully unaware of the All of this happened in late summer
the transformer and turned every- feature. so it was destined to be much more of
thing on. As I slowly reduced the After the power was restored and an issue come mid-winter when heat-
voltage from 230VAC to just above the protection reset, the viewer would ers, hot water and ovens are all work-
200VAC there was a click and every- continue to enjoy uninterrupted tele- ing harder.
thing died. Turning the voltage back vision for many more months or even The “SAT” word which side-
up a little produced another click years, before such an event might hap- tracked me briefly must have been
and everything except the sound bar pen again to trip the protection. programmed into the new TV when it
powered up again. Since I was never present in the was installed by the retailer. I was sur-
The clicking noise was coming from owner's home when the unprompted prised he had gone to this much trou-
the power multi-box and it was now fault happened, I can only suspect his ble. It is a feature within the menu of
obvious to me that this was a feature mains voltage is regularly dropping some TVs which allows you to enter
of the device to protect connected below 200VAC. I was not planning to your preferred label for the various
appliances against brown-out damage spend an evening there monitoring the inputs. Resetting the TV the first time
due to prolonged low voltage supply. mains voltage to find out. I suggested erased this entry and it defaulted back
A press of the power button on the to him that he purchase a $30 plug-in to the HDMI1 label.
sound bar remote control turned that power monitor available from Jaycar Another little twist was that despite
on again. to check for himself; it is not uncom- purchasing a nice sound bar and sub-
When I repeated the exercise, I mon for there to be power fluctuations woofer, again maybe an upsell by a
noticed the TV came back on almost in rural locations. keen salesman, the new owner was
immediately after the voltage was The owner was hoodwinked into missing out completely on the much-
restored but the Sky decoder took believing he did not have a power improved sound these systems offer
several minutes to reboot itself. In problem by a different technology over the standard TV speakers.
the meantime, the TV patiently dis- – CFL lights which maintain their From the very first time the power
played a blank screen with “HDMI1” brightness despite the voltage drop. dropout happened, the speaker system
showing in the top left corner, until it Unlike incandescent lamps which give had not automatically restarted. It was
once again received a signal from the a visual indicator of the voltage of your the only device not to do so and de-
decoder and then displayed a perfectly power supply; flickering or obviously spite a scan through its menu, I could
normal picture. dimming when the voltage dropped. not find any reference to power set-
These were exactly the symptoms Yet another hidden and overlooked tings which would allow me to select
the owner had described. Eureka! After clue was that the Sky decoder was auto power on. It probably relies on
substituting the power multi-box with housed in a cabinet with a tinted glass the HDMI interaction feature which I
a cheap discount store version, with- door, so it was not obvious to the own- mentioned earlier, switching itself on
out the brown-out protection feature, er that the decoder was going through or off in sync with the TV.
he has had no further problems – so far. its reboot sequence after the power in- The owner had been using the TV
terruption. All he saw was the blank speakers only and I know this because
Epilogue TV screen which came back to normal he complained that he had to hold
So I had been hoodwinked by a tech- by itself after a few minutes. the TV remote control high in the air

ualiEco
Circuits Pty Ltd.

siliconchip.com.au June 2017  51


to adjust the volume. The sound bar I really did want to view them. I de- decided to change those too, as I felt
was blocking the infrared signal to the cided to remove the hard drive for a the symptoms dictated a faulty elec-
TV, unless the remote was raised suf- better look of the main board. And it tro somewhere.
ficiently above it. was here that I struck pay dirt or so I I didn’t have the correct values but
When I explained to the owner thought. Under where the hard drive decided to use what I had at hand
that he needed to deliberately power had been sitting I could clearly see rather than waste any more money
the sound bar back on after a power three electrolytic capacitors with their buying new ones. In all I ended up
interruption, I could tell he had not tops bulging. changing CE2, 3, 11, 15, 16, 24, 29
been doing so. In fact, he was com- I didn’t have any of the correct val- and CE223.
pletely unaware that it had not been ues (all three were 220µF 16V types) Some of the capacitors I used were
working and that he had been missing so after a quick visit to the local elec- physically bigger, so I had to leave
out on much improved sound fidelity tronics store, I wasted no time fitting their legs fairly long and bend them
from his entertainment system. the capacitors and replacing the main over into what space I could find so
I had another job recently with board. The HDD was reinstalled, pow- as to reinstall the HDD which, as
a near new TV which came in for er supply connected and switched on. stated, fitted over the whole lot. The
service with both speakers blown; this Well, you can imagine how I felt photo shows some of the replacement
despite the owner having a sound bar. when this time there was even less capacitors.
I wonder if this was another case of response from the unit than before I I didn’t hold much hope of success,
the sound bar not being able to come had started. In disgust, I switched off, so it was a surprise when the machine
back on automatically after a power pushed it to the rear of the bench and booted up and all appeared well. That
interruption? that’s where it sat for 12 months. was several months ago now and I am
During that time it got in my way on gradually catching up on episodes of
Faulty capacitors don’t several occasions but I just ignored it the weekly shows I had recorded some
need to be bulging as best I could. I did make an attempt 12 months ago.
L. W., of Logan, Qld, replaced bulg- to boot the HDD from my PC in an However, it did bug me in not know-
ing capacitors on the main board of his effort to retrieve the information from ing which capacitor was actually the
malfunctioning PVR and thought that it but even with the assistance of the culprit. So one night I fired up the
would fix it. But it wasn’t that simple internet I wasn’t able to achieve an scope and oscillator and measured
and it took a lot more detective work end result. the ESR of the capacitors that I had
to complete the repair. Having recently purchased a replaced the second time around.
About 12 months ago, my TEAC replacement machine, the time came Again to my surprise, while it was a
HDR 1600T Personal Video Recorder to decide whether to have another go 220µF capacitor like the bulging ones
began to exhibit erratic operation until at fixing it or consign it to the wheelie I had originally replaced that was ob-
it finally failed and would not complete bin. It was still in a dismantled state viously faulty, all the others were at
its start up boot process, hanging appr- and picking up the main board, I various stages of deterioration, with
oximately half-way through. noticed that there were several other much higher ESR than they should
I had experienced problems before similar capacitors in the same area as have exhibited.
with the unit which had turned out the ones I had previously changed. So there you go; another piece of
to be a bulging electro in the power All were black with blue label- electronic equipment saved from the
supply circuit, so it was with some ling while most of the other electros tip for just the price of a few electro-
hope that I removed the cover to take on the board were black with white lytic capacitors. You can be lucky
a look inside. labelling. So for no other reason I sometimes. SC
After spending some time examin-
ing the power supply and finding no
signs of distress, I turned my attention
to the main board. Only part of this
board was visible as the hard drive is
mounted above it. There was nothing
obvious here either so I decided to
switch on and measure the outputs
from the power supply.
I waited for it to complete as much
of the boot process as it could and
then took some measurements. There
was no joy here as all appeared to
be well within tolerance. At this
point I started to lose interest as the
machine was now getting a little long
in the tooth and much better machines
with larger hard drives are available
on the market. Bulging 220µF 16V electrolytics weren't the only issue in this video recorder as it
However, there were a couple of wasn't until some of the larger capacitors were changed due to high ESR values
weekly shows that I had recorded and that the set came back to life. This photo shows the new capacitors installed.

52  Silicon Chip siliconchip.com.au


SAVE UP TO $150
60W ESD SAFE SOLDER/DESOLDER
REWORK STATION
Complete solder/desolder station for
production and service use. Microprocessor

SAVE $150
controlled for precise control of thermal
performance. 60W Power.
• Backlit LCD displays NOW
• Temperature range: 160 - 480°C
• 215(W) x 225(L) x 155(H)mm
$
199
SAVE $150
TS-1574 ORRP $349

DUAL OUTPUT LABORATORY


POWER SUPPLY
The two outputs can be operated
independently, connected in parallel, or series
for multiple output currents and voltages.
• Output voltage: 0-32VDC (x2)
• Output current: 0-3A (x2) NOW
16 CHANNEL 720P
ANALOGUE HIGH DEFINITION (AHD) DVR
NOW • 185(H) x 260(W) x 400(D)mm $
319
View live footage on a Smartphone. Supports monitors
with VGA and HDMI input. Plug & Play easy network setup.
$ 599 MP-3087 WAS $399 SAVE $80

Includes DVR, 1TB HDD, mains power adaptor and mouse. 2-IN-1 HANDHELD
SAVE $150 SCOPE & MULTIMETER
• Dropbox cloud storage
• Motion trigger recording All the functions of a 4,000 count True RMS CAT III
digital multimeter and a 10MHz oscilloscope. Includes
• 300(W) x 227(D) x 53(H)mm
USB interface and PC logging software.
QV-3149 WAS $749
• Autoranging, AC/DC voltage (1000V),
current (20A) NOW

720P AHD BULLET CAMERAS TO SUIT 99


$
QC-8668
• 50MSa/s sample rate
QM-1577 WAS $579
$
499
• Ideal for a wide range of SAVE $80
surveillance applications SAVE $80
• 1/3" CMOS sensor
• IR LEDs for day/night vision
• IP66 rated SAVE $60
• Includes 18m cable and power supply
VARI FOCAL QC-8668 ORRP $179
$
149
QC-8670
PAN & TILT QC-8670 ORRP $249 SAVE $100

NOW NOW
$
129 $
139
SAVE 25% SAVE $60 SAVE $60

2 X 50WRMS COMPACT 2 X 75WRMS COMPACT


BASIC DUINOTECH STEREO PA AMPLIFIER STEREO AMPLIFIER
EXPERIMENTERS KIT Delivers the quality of a Class AB amplifier Powerful 2 channel (Stereo) in a compact size.
The entry point for learning and with the efficiency of a Class D. Solid • Built-in digital signal processor
experimenting with Arduino®. Contains aluminium body, banana socket speaker • Volume control and on/off
many parts to get you up and running terminals, 3.5mm stereo input & 6.5mm switch on the front panel
including duinotech Nano board, NOW headphone socket. • Includes power supply and cables
breadboard, jumper wires and plenty of
peripherals. XC-4285 WAS $79.95
$
59 95 • Includes power supply and audio cables
• 78(W) x 150(D) x 50(H)mm
• 165(L) x 95(L) x 30(H)mm
AA-0505 WAS $199
SAVE $20
AA-0488 WAS $189
2 CHANNEL WIRELESS UHF 900W DMX FOG MACHINE
1080P REAR VIEW MIRROR MICROPHONE SYSTEM Use remote control or hookup
CAR EVENT RECORDER Ideal for schools, churches, to a DMX512 controller for total
weddings, gyms, karaoke, etc. customisation of stage
Clips over your existing mirror and features a cool blue
Supplied with two microphones. or party effects.
anti-glare tint, IR night vision, motion detection recording,

SAVE loop recording, emergency file lock, time and date display. AM-4114 ORRP $149 AF-1213 WAS $189
• 4.3" colour LCD

$110
• MicroSD card required (up to 32GB)
QV-3860 ORRP $199

NOW NOW NOW


$
89 $
89 $
129
SAVE $110 SAVE $60 SAVE $60

Catalogue Sale 24 May - 23 June, 2017 To order phone 1800 022 888 or visit www.jaycar.com.au

You might also like