Quote from: trtr6842 on 08 May, 2024 08:29I initially didn't see the small capacitor that appears to be in series with one of your transformer windings, what value is it? In my experience no series capacitor is needed, so long as your signal generator as a truly low DC offset when it's set for 0V. ...Thank you for such a thorough response! That makes a lot of sense.There are a two things still puzzling me:
- Why do we measure the Vsen+ terminal as the input signal, when that doesn't reflect that actual injected signal? If I look at the measurements in the time domain, the input signal looks like 0 at low frequencies even when injecting a 1.5V p-p signal. This makes sense because the control loop is "doing whatever it can" to keep the feedback signal it receives constant at the set Vout. But then why do we measure this constant signal?I'm guessing the input signal must not actually be flat and my eyes aren't the right tool for measuring that, since at higher frequencies the input signal is definitely not flat. If there is some remnant of the actual injected signal that appears on the Vsen+ terminal, wouldn't that be error we're measuring, the part of the injected signal that the control loop couldn't perfectly compensate for?This is showing the input (CH4) and output (CH1) measurements in the time domain with the sig. gen. outputting 100Hz at 5V p-p. Input shows no signal by eye.
- I can't figure out why my phase measurement is off by 180°. The obvious answer is that the polarity of my injected signal across R3 is backwards, but since I was quite careful about the polarity of the injection transformer build and connections, that means I'm misunderstanding something.The injection transformer is wired so that the output is in phase with the input. Then the positive transformer output terminal is connected to R3 on the Vsen+ side. It would seem to me that when the signal generator sine wave output is high, the transformer positive output is high, and so Vsen+ should see a positive perturbation. The control loop should then cause Vout to decrease so that Vsen+ sees the correct setpoint. So sig. gen output goes high, Vout goes low, that's a -180° phase shift.However, the actual measurement at low frequencies shows a phase shift of +20° at 500Hz. You pointed out why that makes no sense, and yet it's definitely what the scope says. So the measurement is wrong, but why?Here is a sketch of the entire measurement setup:
Some setup clarification:There isn't a capacitor in series with the injection transformer, that is just a 500mA polyfuse for protection. The wires are just a blue and blue/white twisted pair from some cat3 cable I had laying around.For each measurement probe on the board, I've soldered a twisted pair to the board and then I wrap those wires around the probe tip accordingly to get low noise measurements. I've found this to be better than using the spring-tip probes by picking up less switching noise and also not requiring so many hands.