Hi! I am working on a circuit that needs to produce biphasic square wave output of controllable + constant current nature. I am using an op amp configured as an integrator connected to a MOSFET as my VCCS (voltage-controlled current source). However, when the circuit is in conduction mode the output voltage of the op-amp momentarily spikes before reaching the desired level set by the feedback loop and in this period the output current spikes as well distorting the square wave. I need the output to be a clean square wave or at least the spike to be as short as possible. Can someone take a look at my circuit and suggest some possible changes? Thanks

the circuit in the image is connected to a mosfet h bridge above it

 

Hi ash,
Welcome to AAC.
Is this a simulation or an actual built circuit.?
Please post your LTSpice asc file.
E

 

Hi,
try a resistor about 200 Ohm in series to 'C'

 

Hi ash,
Welcome to AAC.
Is this a simulation or an actual built circuit.?
Please post your LTSpice asc file.
E

Hey! Thanks for replying. This is in LTspice, however, I have built the circuit physically and it behaves in the same way. The components I used physically vary a little tho. Physically, I used LM324 op amp and IRLR110 mosfets.

Things I have tried: I tried adding a 10Meg ohm resistor in parallel with the capacitor, I tried using a bigger feedback resistor to slow down the input signal to not cause a spike, however, I can't find the right balance of components that gives a clean square wave or at least just a smaller distortion of about 1us-3us.

 

Hi,
try a resistor about 200 Ohm in series to 'C'

Doesn't seem to affect the waveform in LTspice.
If you could also state your thought process along with the suggestion then that would be much appreciated.
Thanks.

 

hi ash,
The OPA is not fast enough, look at this option.
E

 

Both the OP747 and the LM324 are rather slow devices. The OP747 in your sim has an output slew rate of only 0.2V/uS, so can't respond well to the brief sharp input pulses.

 

Doesn't seem to affect the waveform in LTspice.
If you could also state your thought process along with the suggestion then that would be much appreciated.
Thanks.

The remedy in form of a resistor was described for a little bit differently built integrator. ( Input terminal was tied to the inverting input via resistor)

Concerning your circuit. It seems, unwanted pulses at the output can be eliminated or attenuated if input signal dV/dt rate is lowered. Put a resistor in series to '+' input, and a capacitor between '+' and Common terminal. I suggest , 200 Ohm and 500 pF for beginning. Try several values

 

thank you everyone for your replies! using a faster does seem to work in the simulation and I had considered it, however, I was hoping to get things done with lm324 because I have them on hand. Nevertheless, I'll try using a faster op-amp and update this thread if anyone is interested. Having said that, if someone can find a capacitor-resistor combination for the previous integrator circuit then that help is welcomed as well. Again, thank you everyone!

 

The remedy in form of a resistor was described for a little bit differently built integrator. ( Input terminal was tied to the inverting input via resistor)

Concerning your circuit. It seems, unwanted pulses at the output can be eliminated or attenuated if input signal dV/dt rate is lowered. Put a resistor in series to '+' input, and a capacitor between '+' and Common terminal. I suggest , 200 Ohm and 500 pF for beginning. Try several values

Thanks for replying!
By the common terminal you mean the output terminal right?

 

hi ash,
The circuit needs a fast response OPA, in order to eliminate the spike.
E

 

Hi! I am working on a circuit that needs to produce biphasic square wave output of controllable + constant current nature. I am using an op amp configured as an integrator connected to a MOSFET as my VCCS (voltage-controlled current source). However, when the circuit is in conduction mode the output voltage of the op-amp momentarily spikes before reaching the desired level set by the feedback loop and in this period the output current spikes as well distorting the square wave. I need the output to be a clean square wave or at least the spike to be as short as possible. Can someone take a look at my circuit and suggest some possible changes? Thanks

the circuit in the image is connected to a mosfet h bridge above it

Hello,

If you can get away with using a device like the LM324 then why do you need a mosfet on the output?
The mosfet can introduce an unwanted increase in loop gain and capacitance which can easily cause oscillation. Try using an NPN bipolar (not PNP) you should get better results.

 

Thanks for replying!
By the common terminal you mean the output terminal right?

No. Ground point

 

hi! i am working on the same circuit again. Update: I used a fast response op amp lmh6644. While my problem with the current spike was resolved I ran into a new problem. One of the biphasic square waves has a high-frequency noise, when I bring my hand close to one of the wires the noise frequency changes. Any ideas on how I could eliminate the noise?

 

hi! i am working on the same circuit again. Update: I used a fast response op amp lmh6644. While my problem with the current spike was resolved I ran into a new problem. One of the biphasic square waves has a high-frequency noise, when I bring my hand close to one of the wires the noise frequency changes. Any ideas on how I could eliminate the noise? View attachment 292624

What circuit are you using now?

 

It is not "noise". instead it is a high frequency oscillation changed by the capacitance of your hand near the wiring.

I notice that your schematic has some wires out of place then maybe the circuit is built the same with a mess of wires all over the place on a breadboard?

 

What circuit are you using now?

I am using the same circuit as last time just a new op amp lmh 6644. you can find the schematic close to the start of the thread

 

It is not "noise". instead it is a high frequency oscillation changed by the capacitance of your hand near the wiring.

I notice that your schematic has some wires out of place then maybe the circuit is built the same with a mess of wires all over the place on a breadboard?

hi! I am using a general purpose board and I believe the wiring is pretty neat (at least by my standards). I tried to shorten the wire that goes to the + input of the op amp, this did change the high freuency noise's frequency but did not eliminate it. What do you think is causing the high frequency oscillation? my guess from your response would be parasitic components in the circuit

 

Maybe the input should be a shielded audio cable and the shield will prevent the input from picking up the output and prevent other interference.

 

I am using the same circuit as last time just a new op amp lmh 6644. you can find the schematic close to the start of the thread

Ok so did you try a bipolar instead of mosfet yet? The circuits end up more stable.