MC33072 - Am I exceeding something?

Thread Starter

atferrari

Joined Jan 6, 2004
4,764
Agustín, here is another way to do this. The LM317 regulator is inside the op amp's feedback loop, and basically acts as a bullet-proof voltage follower within the loop.
You don't have to consider the 1.25v vref when calculating Vout/Vin
Hola Ron,

I spent almost one day relearning how to import models and subcircuits into LTSpice. I was totally at lost but thanks to a post from Eric and one from you, I solved that. Now, back to the original subject.

After simulating your variation and my circuit I am not sure what is the actual advantage of yours, because I want an output from -0.25V to -30V.

The only way I found is applying a (+1V) offset after amplifying the control signal as in my attached circuit.

Comments appreciated.
 

Attachments

Ron H

Joined Apr 14, 2005
7,063
OK, here's what I was proposing, modified for -0.25V to -30V with 0 to 2.048V in. You can use any combination of reference voltage and R4 to get the -0.25V offset. I used 1V, because that's what you used.
Note that that Vout/Vin is independent of the LM337 reference voltage, because it's inside the feedback loop.
With your circuit, any drift or regulation issues with the LM337 will show up in the Vout/Vin relationship.
 

Attachments

Thread Starter

atferrari

Joined Jan 6, 2004
4,764
OK, here's what I was proposing, modified for -0.25V to -30V with 0 to 2.048V in. You can use any combination of reference voltage and R4 to get the -0.25V offset. I used 1V, because that's what you used.
Now I see how it could be done.

Two things:

What is the reason for R5? If I eliminate it, nothing seems to change.

R4 should be 58K instead of 581K.

Gracias for so much help and for your time!!
 

Ron H

Joined Apr 14, 2005
7,063
Now I see how it could be done.

Two things:

What is the reason for R5? If I eliminate it, nothing seems to change.
R5 compensates for input bias current. Ideally, R5=R2||R3||R4=9.2k, so 9.1k would theoretically be slightly better than 10k. Run the sim with and without R5, and notice the small difference in the output voltage. You can set Vctrl and Vref to zero and see the output offset voltage. When I did this, the output voltage went from -20mV when R5=0 to -5.6mV when R5=10k. The remaining offset is due to the amplifier's input offset voltage.

R4 should be 58K instead of 581K.
The output voltage is
Vout=-(Vin*(R2/R3) + Vref*(R2/R4))
With R4=58k, the output will go from -2.5v to -32.25v. That is not what you want.

Gracias for so much help and for your time!!
 

Attachments

Thread Starter

atferrari

Joined Jan 6, 2004
4,764
Well Ron, I think I have things a lot clearer than in the beginning.

Would you actually use +5V and -38V to power the MC33071 considering the absolute maximum as per datasheet is 44V?

In my case I took the pain to have +5V and -32V. Could you please comment?

Gracias.
 

Ron H

Joined Apr 14, 2005
7,063
Well Ron, I think I have things a lot clearer than in the beginning.

Would you actually use +5V and -38V to power the MC33071 considering the absolute maximum as per datasheet is 44V?

In my case I took the pain to have +5V and -32V. Could you please comment?

Gracias.
That detail slipped right by me. You are right in avoiding powering the op amp near its limits. Since the current is low, and regulation is not critical, a very simple regulator will work for -32V.
 
Top