LECTURE 6 – ANALYSIS USING BLOCK DIAGRAMS

Block diagrams are a means to displaying graphically the connections between various
components of a system. There are rules for reducing any block diagram down to just one
box with an input and output. That box represents the transfer function of the system.
Therefore we assume zero boundary conditions at all times when working with block
diagrams. We can work either in the time or frequency domain as follows.

We recall
𝐹𝐹(𝑠𝑠)
ℒ[𝑓𝑓′(𝑡𝑡)] = 𝑠𝑠𝑠𝑠(𝑠𝑠) and ℒ[∫ 𝑓𝑓(𝑡𝑡)𝑑𝑑𝑑𝑑] = 𝑠𝑠

In the block diagram form we have

We see functions or multiplying factors are in boxes and variables are associated with the
connecting lines.

An illustrative example

We want to find the transfer function of

the mass/spring/damper system.

We know 𝐹𝐹𝑖𝑖 = 𝐾𝐾𝐾𝐾 + 𝐶𝐶𝑥𝑥̇ + 𝑀𝑀𝑥𝑥̈ so draw

a summation junction making the
highest differential the output.

We will work in the complex domain for this solution using 𝐾𝐾𝐾𝐾, 𝐶𝐶𝐶𝐶𝐶𝐶, 𝑀𝑀𝑠𝑠 2 𝑋𝑋 etc.

The output is 𝑥𝑥 with input 𝐹𝐹𝑖𝑖 . So to obtain 𝑋𝑋 alone we must divide the 𝑀𝑀𝑠𝑠 2 𝑋𝑋 by 𝑀𝑀 and
then use the following:
We also note we can derive 𝐾𝐾𝐾𝐾 using

and 𝐶𝐶𝐶𝐶𝐶𝐶 using

So the complete block diagram is

This has left us in the position of needing ways to reduce the block network down to one
block that will contain the transfer function.

RULES FOR REDUCING BLOCK DIAGRAMS

Combinations of boxes can be reduced to one box using the following rules.

RULE
1 Multiplication
→ (Serial)

2 Addition/Subtraction
→ (Parallel)

3 Feedback loop (positive or

→ negative)
4 Moving a junction on
→

5 Moving a junction back

→

Now returning to our example, using Rule 3 gives

Using Rule 1 gives

Finally using Rule 3 gives

Simplifying the contents of the box gives

𝑋𝑋 1
𝐹𝐹𝑖𝑖
= 𝑠𝑠(𝑀𝑀𝑀𝑀+𝐶𝐶)+𝐾𝐾 , the transfer function.

Just to complete the analysis, say that 𝐹𝐹𝑖𝑖 was a step function of magnitude 𝐹𝐹. We will find
out what the displacement 𝑥𝑥 looks like.

𝐹𝐹 𝐹𝐹 𝐾𝐾�
𝑀𝑀
∴ 𝑋𝑋 = 𝐶𝐶 𝐾𝐾 = 𝐾𝐾 ∙ 𝐶𝐶 𝐾𝐾
𝑀𝑀𝑀𝑀�𝑠𝑠2 + 𝑠𝑠+ � 𝑠𝑠�𝑠𝑠2 + 𝑠𝑠+ �
𝑀𝑀 𝑀𝑀 𝑀𝑀 𝑀𝑀

To illustrate the inversion, we use

𝐹𝐹 𝜔𝜔 2
𝑋𝑋(𝑠𝑠) = 𝐾𝐾 ∙ 𝑠𝑠(𝑠𝑠2 +2𝜉𝜉𝜔𝜔𝑛𝑛 2)
𝑛𝑛 𝑠𝑠+𝜔𝜔𝑛𝑛

𝐾𝐾 𝐶𝐶
We made 𝜔𝜔𝑛𝑛 = �𝑀𝑀 , 𝜉𝜉 = 2√𝐾𝐾𝐾𝐾 ; 𝜔𝜔𝑛𝑛 is called natural frequency, 𝜉𝜉 is called damping ratio.
Hence using tables, the result is

𝐹𝐹 1 1�
𝑥𝑥(𝑡𝑡) = 𝐾𝐾 �1 + 1 ∙ 𝑒𝑒 −𝜉𝜉𝜔𝜔𝑡𝑡 ∙ 𝑠𝑠𝑠𝑠𝑠𝑠 �𝜔𝜔𝑛𝑛 (1 − 𝜉𝜉 2 ) 2 ∙ 𝑡𝑡 + ∅��
(1−𝜉𝜉 2 ) �2
with
1 �
�1−𝜉𝜉 2 � 2
∅= 𝑡𝑡𝑡𝑡𝑡𝑡−1 𝜉𝜉

Such an expression says we have oscillations that are decaying away. So finally, what value
does 𝑥𝑥(𝑡𝑡), 𝑡𝑡 → ∞ settle at?

Use the final value theorem on the definition of 𝑋𝑋.

1 𝐹𝐹 𝐹𝐹
lim𝑡𝑡→∞ [𝑥𝑥(𝑡𝑡)] = lim𝑠𝑠→0 [𝑠𝑠 �𝑆𝑆 ∙ 𝑀𝑀𝑠𝑠2 +𝐶𝐶𝐶𝐶+𝐾𝐾�] = 𝐾𝐾 , in other words the resulting extension is that
due to a static force 𝐹𝐹 extending the spring, which is what we would expect, i.e. 𝐹𝐹 = 𝐾𝐾𝐾𝐾.

Block diagrams for reservoirs

We will analyse the 𝑟𝑟ℎ𝑠𝑠 diagram

treating 𝑞𝑞𝑖𝑖 as an input and ℎ0 as an
output.

The net flow into the right-hand

reservoir:

𝐴𝐴ℎ̇0 = 𝑞𝑞𝑖𝑖 − 𝑞𝑞01 − 𝑞𝑞02 , keeping the sign convention which becomes:

whereas the net flow into the left-hand

reservoir is directly:

ℎ0 ℎ ℎ
However, 𝑞𝑞01 is made up of (by superposition) − 𝑅𝑅1 and 𝑞𝑞02 = 𝑅𝑅0 which gives
𝑅𝑅1 1 2

and

Combining all the sub block diagrams we get:

 All variables have been written for
the time domain just for a change.

Treat the top part as a feedback loop and apply Rule 3 to get:

Remember:

Now apply Rule 3 to the bottom loop to get:

which reduces to one block using Rule 3 to get:

𝐻𝐻0 𝑅𝑅2 (𝑅𝑅1 𝐴𝐴𝐴𝐴+1)

∴ = (𝑅𝑅
𝑄𝑄𝑖𝑖 2 𝐴𝐴𝐴𝐴+1)(𝑅𝑅1 𝐴𝐴𝐴𝐴+1)+𝑅𝑅2 𝐴𝐴𝐴𝐴
Block diagrams for electrical circuits

We will analyse the 𝑟𝑟ℎ diagram

treating 𝑒𝑒𝑖𝑖 as an input and 𝑒𝑒0 as an
output.

Applying Kirchhoff’s Laws, we know

1
𝑒𝑒𝑖𝑖 = 𝑖𝑖1 𝑅𝑅1 + 𝑒𝑒2 , 𝑒𝑒2 = 𝑖𝑖2 𝑅𝑅2 + 𝑒𝑒0 , 𝑒𝑒0 = 𝑐𝑐 ∫ 𝑖𝑖2 𝑑𝑑𝑑𝑑
2
and
1
𝑒𝑒2 = 𝑐𝑐 ∫ 𝑖𝑖3 𝑑𝑑𝑑𝑑 ∴ 𝑖𝑖2 = 𝑐𝑐2 𝑒𝑒̇0 and 𝑖𝑖3 = 𝑐𝑐1 𝑒𝑒̇2
1

Hence we can draw, knowing 𝑖𝑖1 = 𝑖𝑖2 + 𝑖𝑖3

Next step is to move the junction at 𝑖𝑖2 to 𝑒𝑒0 using Rule 5 then apply Rule 3 to the right-
hand loop to get:

Now move the top point of 𝐶𝐶1 𝑠𝑠 box to the output using Rule 4 and sum all the boxes in the
feedback path using Rule 2 to get:
Finally, use Rule 3 to make one box, using 𝑇𝑇1 = 𝐶𝐶1 𝑅𝑅1 , 𝑇𝑇2 = 𝐶𝐶2 𝑅𝑅2 , to get:

This is the same result as in the example in LECTURE 4.

FINALLY! The general case:

Say we have a general differential equation relating input to output. How does it translate
itno a block diagram?

Take the equation:

𝑦𝑦⃛ + 4𝑦𝑦̈ − 5𝑦𝑦̇ + 6𝑦𝑦 = 3𝑥𝑥̇ + 4𝑥𝑥 [𝑦𝑦 = 𝑓𝑓(𝑥𝑥)] ,

rewrite as:

𝑦𝑦⃛ = −4𝑦𝑦̈ + 5𝑦𝑦̇ − 6𝑦𝑦 + 3𝑥𝑥̇ + 4𝑥𝑥 , and start by creating all the right-hand terms.

Integrate 𝑦𝑦⃛ in stages to 𝑦𝑦 .

Put 𝑥𝑥 on the left-hand side and
differentiate. Then feedback each
term needed with coefficients.

If this block diagram is reduced to one box it will generate the transfer function

𝑌𝑌 3𝑠𝑠+4
𝑋𝑋
= 𝑠𝑠3 +4𝑠𝑠2 −5𝑠𝑠+6