Strange OpAmp behaviour (step 5)

[alexdavey]

When I connect the power supply, output TP4 initially looks like this:


Then it slowly transitions to the final (correct, based on other scope screenshots on the forum) waveform:





Note that this happens when the OpAmp chip heats up (NE5532P), and by blowing on it I can get it to revert back to the initial condition (for a while).

I have tried swapping the chip for a NE5532AP, and this gave the following:



Which looks wrong but doesn't display any temperature variation and when I connect a voltmeter to TP4 gives the correct behaviour in the presence of a metallic target (the other OpAmp chip does behave correctly in the final stage, however it is very insensitive in comparison).

I have checked the power supplies (from 12.0V bench PS) - there is a slight spike (~100mv) and some ripple, but apart from that they seem fine.

The flyback voltage seems correct


And the OpAmp input pin doesn't change over time/temperature:

[alexdavey]

Added a video of the OpAmp from power-up to "correct" waveform:
https://vid.me/AF3q

Also I have realised that the "final" waveform is not actually quite correct, especially at the very end where it becomes an almost perfect square wave.

[alexdavey]

I swapped out the OpAmp for the other one again, and tested the whole circuit and it seems to work! Not sure what's going on with the second OpAmp stage - it seems to have unity gain, but I may have made a measuring error.

[Qiaozhi]

I swapped out the OpAmp for the other one again, and tested the whole circuit and it seems to work! Not sure what's going on with the second OpAmp stage - it seems to have unity gain, but I may have made a measuring error.

Check that R16 is 1k, and R17 is 33k. Also, do a continuity test between pins 6 and 7 of U5. If there's a short-circuit, the gain will be unity.

[alexdavey]

Check that R16 is 1k, and R17 is 33k. Also, do a continuity test between pins 6 and 7 of U5. If there's a short-circuit, the gain will be unity.

All of those are fine. I have attached a oscilloscope shot, from the input (yellow, pin 5) to output (blue, pin 7) of the second stage:



However the output of the first stage is already about +- 4v, ie almost saturated, is there a problem with the flyback signal? It is always overdamped, regardless of how much I decrease the damping resistance (I used the damping resistor pot setup you described in another thread). Changing the damping resistance does change the peak voltage, (200-450v). Here's a shot of the flyback signal:



(The damping resistor has values between 185Ω and 1K)

[Qiaozhi]

All of those are fine. I have attached a oscilloscope shot, from the input (yellow, pin 5) to output (blue, pin 7) of the second stage:

[ATTACH]33316[/ATTACH]

However the output of the first stage is already about +- 4v, ie almost saturated, is there a problem with the flyback signal? It is always overdamped, regardless of how much I decrease the damping resistance (I used the damping resistor pot setup you described in another thread). Changing the damping resistance does change the peak voltage, (200-450v). Here's a shot of the flyback signal:

[ATTACH]33317[/ATTACH]

(The damping resistor has values between 185Ω and 1K)

I think you're just misunderstanding the operation of the circuit.
When the MOSFET turns on, the current starts to build up in the coil. At this time the opamp output will hit the +5V rail and go into saturation. When the MOSFET turns off, the flyback voltage is generated, and the opamp output hits the -5V rail, again going into saturation. After a few microseconds, the coil current will decay sufficiently so that the opamp output will come out of saturation. You can see this in your scope shots, as the voltage starts to rise exponentially towards zero. Adjusting the damping resistor will not affect the saturated portions of the waveform. You should be aiming for the exponential voltage to decay in a critical manner (i.e. no ringing, due to being underdamped; and no overdamping, which will decrease sensitivity).
The "ringing" you're seeing in the waveforms is due to external noise being picked up by the coil. No doubt because you're testing inside the house. Try moving the coil around to reduce the noise level.

[alexdavey]

I think you're just misunderstanding the operation of the circuit.
When the MOSFET turns on, the current starts to build up in the coil. At this time the opamp output will hit the +5V rail and go into saturation. When the MOSFET turns off, the flyback voltage is generated, and the opamp output hits the -5V rail, again going into saturation. After a few microseconds, the coil current will decay sufficiently so that the opamp output will come out of saturation. You can see this in your scope shots, as the voltage starts to rise exponentially towards zero. Adjusting the damping resistor will not affect the saturated portions of the waveform. You should be aiming for the exponential voltage to decay in a critical manner (i.e. no ringing, due to being underdamped; and no overdamping, which will decrease sensitivity).
The "ringing" you're seeing in the waveforms is due to external noise being picked up by the coil. No doubt because you're testing inside the house. Try moving the coil around to reduce the noise level.

You are right, thank you . Just to clarify regarding the damping resistance, I expected to see this for the flyback signal (measuring directly across the damping resistor):



Is this not what I should expect to see?

[soverey]

Hi,

I'm also getting a strange OpAmp behaviour , like in the picture I have attached. Could someone tell me what is going wrong? . For example a have added a second picture for comparison, of how the OpAmp looked before.
Thanks.

[Qiaozhi]

Hi,

I'm also getting a strange OpAmp behaviour , like in the picture I have attached. Could someone tell me what is going wrong? . For example a have added a second picture for comparison, of how the OpAmp looked before.
Thanks.

The second image shows that the damping resistor is too high. You can see that it overshoots before settling close to 0V.

[soverey]

The second image shows that the damping resistor is too high. You can see that it overshoots before settling close to 0V.

Thank you very much, I adjusted the damping resistor and now it looks normal.