Transconductance circuit with arduino due

Thread Starter

deeweerf

Joined Nov 20, 2018
5
Hi, im trying to model a transconductance circuit (or VCCS)
to power a pneumatic proportional valve, using an Arduino DUE
with an opamp and a resistor. The valve specs are 24Vdc and
current from 0 to 85mA, need to be current controlled. It's
simply a resisance of about 200ohm. The signal comes from Arduino
DAC into the amplifier.

So, the arduino signal goes into the noninverting input of the
opamp (i'm using an LM7171, dont know if it's the optimal one),
a precise resistor of 27.4ohm is inserted from the inverting
input to ground and the valve from opamp output to inverting
input. The opamp is single supplied from 30V (V+) to ground (V-).
When arduino applies 2.7V to opamp input, the output should
supply 100mA.

Ive tried simulating this circuit with some programs and seems
working well. But I'm not really good in electronics and i don't
wanna burn the opamp or arduino for a stupid mistake.

Is this circuit a good choice? Is this opamp a good choice too?
Should I add a resistor from arduino output to opamp noninverting
input to prevent high current flowing from or to arduino input (it
can bear like 20mA, i know ideally there should not be any current).
Is it a problem that the load is not grounded?
Should i add some capacitor somewhere in the circuit (I'm not sure
what I'm asking here but I've seen a lot of circuit with capacitors
for stability).

Some more information: this circuit controls a pneumatic valve, so
it's a mechanical system. I chose a fast opamp even if i won't give
an input signal with a freuency higher than 500Hz. I chose it basically
because it can give me 100mA and be supplied with 30V.

Thanks in advance for any answer :)
 

danadak

Joined Mar 10, 2018
4,057
Thats a classic circuit.

upload_2018-11-26_8-31-30.png

If load is inductive I would use a clamp diode on it (strapped across it),
to minimize transients due to load disconnect possibilities.

Regards, Dana.
 

ebeowulf17

Joined Aug 12, 2014
3,307
Thats a classic circuit.

View attachment 164443

If load is inductive I would use a clamp diode on it (strapped across it),
to minimize transients due to load disconnect possibilities.

Regards, Dana.
Maybe I misread the description, but I didn't think the thread starter had a transistor - just the op amp sinking all the current.

@deeweerf , do you have a schematic image you can upload? Much easier to check than verbal description!
 

Thread Starter

deeweerf

Joined Nov 20, 2018
5
Maybe I misread the description, but I didn't think the thread starter had a transistor - just the op amp sinking all the current.

@deeweerf , do you have a schematic image you can upload? Much easier to check than verbal description!
It's like this circuit, but Vi comes from arduino due DAC and ZL is my valve, which is a resistance of 200ohm. Im not using a transistor
 

Attachments

danadak

Joined Mar 10, 2018
4,057
Of course you have to pay attention to heat dissipation so thermal
design just as important as electrical.

And board layout, separate the current path to take valve
current down its own path and meet at board edge. That would
be power pins on 7171. Maybe bypass them with a good polymer
tant and MLC for bypass.


Regards, Dana.
 

ebp

Joined Feb 8, 2018
2,332
You have a feedback element that will have substantial inductive reactance that is probably not constant, plus I'm assuming it is likely to be remote from the very high speed op amp. This is a recipe for an oscillator that is likely to behave very badly.
 

Thread Starter

deeweerf

Joined Nov 20, 2018
5
Of course you have to pay attention to heat dissipation so thermal
design just as important as electrical.

And board layout, separate the current path to take valve
current down its own path and meet at board edge. That would
be power pins on 7171. Maybe bypass them with a good polymer
tant and MLC for bypass.


Regards, Dana.
What about thermal design? How should i do it?
I dont understand what you saying about the board layout, could you explain it better (I know
a bit of electronics, but im not so good)?

You have a feedback element that will have substantial inductive reactance that is probably not constant, plus I'm assuming it is likely to be remote from the very high speed op amp. This is a recipe for an oscillator that is likely to behave very badly.
The resistance of the valve is not constant, but that's why i chose this circuit. It should be able to deliver
the current i need without considering the load (of course if it has enough power to deliver). I wasn't concerned
about the load reactance, should it be a problem? Could you explain it?
However this opamp is very fast, but i wont give an input signal of frequency higher than 500Hz because it
controls a mechanical system that difficulty will vibrate with frequency of MHz. So why should it become
an oscillator (a bad one)?
 

Thread Starter

deeweerf

Joined Nov 20, 2018
5
Well, for circuit test i'll use a simple breadboard, so the board layout is not a
concern to me for now.
About the thermal problem, i tried following what's written in the LM7171
dataheet. Please tell me if I'm wrong:
Pd=Pq+Pl
Pq is constant at 195mW
Pl=(100mA)*(30-24-3) 30V is the supply, 24V for the load, and 3 is the voltage
forced by the opamp at the inverting input.
Pl=300mW -> Pd=495mW way lower than 730mW it can dissipate at 25°C.
So thermal problem should not be a problem? :)
 

danadak

Joined Mar 10, 2018
4,057
Max Load for OpAmp occurs when 100 mA flows thru 200 Ohms = 20V
across load. From graphs 100 mA sourced or sunk by OpAmp output =
~ 4V drop across OpAmp driver transistor @ 100 mA =~ 400 mW.

So

Pd = Pq + Pl = 195 mW + 400 mW = 595 mW.

Tjmax = 150 C

DIP THETAj-a = 108 C/W

Tj = 108 x .595 = ~ 64C (above ambient)

So looks like plenty of margin.

Regards, Dana.
 

Thread Starter

deeweerf

Joined Nov 20, 2018
5
Alright, so it's thermally ok. Do you think it will work or should i expect any problems?

You have a feedback element that will have substantial inductive reactance that is probably not constant, plus I'm assuming it is likely to be remote from the very high speed op amp. This is a recipe for an oscillator that is likely to behave very badly.
I still dont understand why i need to consider load reactance. Well, i know that when i will feed with Arduino a sinusoidal signal with, for example, a frequency of 100Hz, the reactance will show. But should it be a problem? As i said, i'm not going to feed with very high frequency, because the circuit will control a mechanical system which hardly will oscillate at high frequency. And even if it can, i think the oscillation will be tolerable. I always has been taught that electrical systems are much more sensitive and faster than mechanical ones.
 
Top