circuit anomally figured out.

Published by hobbyist in the blog hobbyist's blog. Views: 63


OK here is the situation,
I'm experimenting with designing a transistor PWM circuit with current feedback to control motor torque.

Everything is working properly, So now I am ready to start measuring voltage and current values so as to determine the best current sensing restance needed in series with the mosfet and motor load.
So I set up my multimeters one across the motor, another across the sense resistor and another in the feedback link and I hook up my osciloscope, to the output of my PWM board to detect increase in duty cycle during motor loading.

Everything works properly, I put a load on the motor and I see the voltage across the sense resistor increase as the motor voltage decreases slightly I watch the feedback voltage increase properly and the duty cycle of the PWM wave increases to feedback more voltage input to the gate of the mosfet, increasing motor torque under load.

Great however when I disconnect the feedback circuit link I notice theres some slight duty cycle change in the PWM wave when I put a load on the motor. That should not be.

Now a little explanation is in order for the circuit topology. I have the input to the gate of the mosfet drivien by a opto isolator, and the mosfet is on its own circuit supply.

So the PWM board is electrically isolated from the motor driving circuit. The PWM board is supplied by a 9 volt battery eliminator, and the mosfet circuit is supplied by my 30V. supply.

I have all feedback links disconnected, so all I have is the PWM output driving an opto isolator LED, which in turn is optically driving an NPN transistor in the opto isolator which applies input signal to the mosfet gate terminal.

Yet when I put any kind of a load on the mosfet drain the PWM waveform increases in duty cycle, wether the load on the mosfet is active such as the motor, or even a passive load, a 7 ohm power resistor.

Mind you, there is complete electrical isolation between both boards the PWM board and the mosfet board, the only connection between the two is the ground, to allow proper gate terminal voltage input. Without the ground connection between the two boards the mosfet has no gate input from the opto isolator, the mosfet turns on due to high gate impedance, when the common ground is connected the waveform controls the mosfet action.

So I thought maybe its the ground link between the two, some how its interacting to cause the PWM board to detect a load put on the mosfet.

So I disconnected the entire circuit boards and started over, I hooked up the PWM board on its own 9V supply with a battery, and I hooked up the mosfet board on the 30V supply. I did not connect a common ground between the two boards, then I hooked up only the one wire from the PWM board opto isolator, to the gate of the mosfet, the moment I did that the motor turned on full, and I had no control over its speed with the gate drive. This I know is due to the mosfet being static driven just by touching ists gate, so I know it needs a common ground to the PWM board.

So I connected a common ground from the 9V battery to the 30V supply, and now I have perfect controle over the motor speed by varying the PWM wave duty cycle, as expected.

HOWEVER when I put it to the next test by removing the mosfet load, the PWM waveform duty cycle remained stable, it did not change, so I tried other loads on the mosfet and the PWM waveform remained stable, no matter a load or no load on the mosfet.

Then I removed the 9V battery and went back to using the battery eleminator as in the original curcuit setup, the PWM waveform still remained stable as expected.
Well that's exactly what is supposed to happen, after all the boards are electrically isolated, from each other, problem solved.

However what was the original cause of the PWM waveform being unstable earlier.

Here is the answer, in the second setup I did not bother to set up my multimeters across the motor and sense resistor, I purposely left them out to allow the circuit to have no outside influence whatsoever, the only test equipment I had hooked up was my oscilloscope to the output of the PWM board on the LED side of the opto isolator, so as to have no physical connection to the gate at all.

Sure enough it was the multimeters hooked into the circuits that was causing some sort of deisolation between the boards.

Moral to the story if a circuit isn't functioning the way I expect it too, I better check my test equipment connections they could be the cause of circuit loading.
You need to be logged in to comment