Op-amp circuit help

[Humphrey B. Bear]

Hello all, I've been trying without success to design a circuit that has 2
inputs and one output, such that the output is:- Vout = Vin1 + (Vin1-Vin2):-

Vin1----|

---- Vout

Vin2----|

I'm using a single 5V supply and both input voltages come from op-amp
voltage-followers, (originally derived from PWM signals). I need the output
to swing from 0V to as close as possible to the 5V supply.

The circuit is intended as part of a closed-loop speed control system for a
DC motor.

Can anyone help with ideas, links etc.?

.... Humphrey

 

[Tony Williams]

Hello all, I've been trying without success to design a circuit
that has 2 inputs and one output, such that the output is:- Vout
= Vin1 + (Vin1-Vin2):-
Vin2----|

R1 R2 R1=R2= 10k-ish.
Vin2---/\/\---+---/\/\---+
| _ |
+--|- \ |
| >---+---> Vout = 2*Vin1 - Vin2
Vin1-------------|+_/

I'm using a single 5V supply and both input voltages come from
op-amp voltage-followers, (originally derived from PWM signals).
I need the output to swing from 0V to as close as possible to the
5V supply.

Look for an opamp described as rail-rail input/output.

 

[Humphrey B. Bear]

R1 R2 R1=R2= 10k-ish.
Vin2---/\/\---+---/\/\---+
| _ |
+--|- \ |
| >---+---> Vout = 2*Vin1 - Vin2
Vin1-------------|+_/


Look for an opamp described as rail-rail input/output.

Too easy. Looks good to me. Thanks, Tony.
Do you think I'd strike any problems if I used 1M resistors instead of 10K,
so that I could do away with the voltage-followers and take the inputs
directly from the PWM filter capacitors? (Vin1 is the requested motor speed,
a 20kHz PWM signal, and Vin2 is the motor speed feedback signal, PWM with a
frequency ranging from 100Hz at 60RPM to 13.33kHz at 8000RPM, derived from a
100 pulse-per-revolution, 2", optical encoder disk Vin1 is smoothed with a
1uF cap and Vin2 is smoothed by a 10uF cap)

.... Humphrey

 

[Tony Williams]

Too easy. Looks good to me. Thanks, Tony. Do you think I'd strike
any problems if I used 1M resistors instead of 10K, so that I
could do away with the voltage-followers and take the inputs
directly from the PWM filter capacitors? (Vin1 is the requested
motor speed, a 20kHz PWM signal, and Vin2 is the motor speed
feedback signal, PWM with a frequency ranging from 100Hz at 60RPM
to 13.33kHz at 8000RPM, derived from a 100 pulse-per-revolution,
2", optical encoder disk Vin1 is smoothed with a 1uF cap and Vin2
is smoothed by a 10uF cap)

Yes you could do. You could even probably reduce
the circuit to just one filter. But if this is a
first go then retaining the followers allows you
to look at Vin1 and Vin2 separately.... useful for
diagnostics.

 

[Humphrey B. Bear]

Yes you could do. You could even probably reduce
the circuit to just one filter. But if this is a
first go then retaining the followers allows you
to look at Vin1 and Vin2 separately.... useful for
diagnostics.

Tony Williams.

Good one. Thanks again for the help, Tony.

 

[Ban]

Too easy. Looks good to me. Thanks, Tony.
Do you think I'd strike any problems if I used 1M resistors instead
of 10K, so that I could do away with the voltage-followers and take
the inputs directly from the PWM filter capacitors? (Vin1 is the
requested motor speed, a 20kHz PWM signal, and Vin2 is the motor
speed feedback signal, PWM with a frequency ranging from 100Hz at
60RPM to 13.33kHz at 8000RPM, derived from a 100
pulse-per-revolution, 2", optical encoder disk Vin1 is smoothed with
a 1uF cap and Vin2 is smoothed by a 10uF cap)

... Humphrey

Hi Humphrey,
I think with your idea it will be very difficult to control the motor speed,
because of long time constants. If your motor is a DC motor, you could
actually use the fact that speed is proportional to applied voltage, if you
correct for armature resistance. So just a current sensor in form of a
resistor is needed instead of this encoder.
http://www.elecdesign.com/Articles/ArticleID/6199/6199.html
http://www.elecdesign.com/Articles/ArticleID/6240/6240.html
These articles will give you some ideas I hope.