severe ringing on scope with new snake oil capacitor

[pjd3]

Hi there. I just found something that I didn't expect an was hoping that some of you might comment on what the implications of this are.

I've been using scopes in work as an electronic R&D tech for years and just got my own scope for Christmas. One of the things I wanted to be comfortable with was measuring capacitors by way of injecting a square wave into a series R/C. Then, using cursors to find the 63% of the amplitude then do the math to arrive at the value.

I just received an order of components for a JMC800 2204 that I've began to build. Just for shits n giggles I decided to scope out the big chunky 1000V 500pF Sozo cap to see how close my new method would arrive at the caps value. To my surprise, there was severe ringing at the rising and falling edge of every square wave. Although I'm quite familiar with ringing occuring in many situations, during the 100-200 scope/squarewave measurements i've taken on caps, I've never seen even a small amount of ringing on any cap. And I've measured every type of cap, ceramic, mica, electrolytic, poly, and all values from pico farads to power filter cap size. Never any ringing from any of them, and I did alot over many nights to try to get good at this.

I play a game with myself and allow myself to purchase one or two silly even snake oil level component for my builds. For this build, it was about spending near 10 bucks on a damn 500pF cap, that I could have spent 50 cents for nice black shiny bean pod mica. So, I'm surely not the last word on capacitor specifications but, I guess I just didn't expect such heavy ringing on this mighty snake oil and expensive Sozo cap.

Do you think this is anything normal or expected? Or am a finally doing something wrong with my test approach? Seems like I'm doing the same as always. I even changed the 1M resisitor to a 10K and adjusted the frequency for a reasonable waveform but nope, that ringing was prevalent no matter what was changed. And we are talking a first ring peak that is about 50% higher than the level off amplituded. I'm stumped and feel either pretty dumb right now, or pretty fooled!

I would really like to hear what you think about this.
Thanks for your time and patience!
Phil Donovan

[pjd3]

I think I figured it out.

Normally, I applied the signal at the resistor with the return at the other side of the cap. The scope would look across the cap with the probe between the cap and resistor and the scopes ground along with the signal generators return at the caps end.

I believe I set the circuit up backwards to that, applying the signal at the cap, and the return at the other end of the resistor. But, whatever misnomer I found myself under, the waveform looked perfectly normal with a nice smooth charge curve with no overshoot or ring. And the 500pF cap calculated out at 462uF which I take as reasonable.

So, my bad. I'll need however to take a closer look to understand why I was seeing such overshoot with the square wave applied at the cap first. I don't at this point understand the implications of that, unless, the cap monentarily looks like a short not having the resistor first in line? I had assumed that in this little R/C series simple circuit that the time constant would be the time constant any way that you looked at it. Well, time to try and understand yet another question about the way electrons and components can behave.

Thank you for stopping by,
Best,
Phil

[martin manning]

If I understand correctly, in the first case you had the R and C wired like a high pass filter on the signal generator output (series C, shunt R), and in the second case you have a low pass filter (series R, shunt C), and in both cases the scope probe was on the junction of the R and C. The time constant is indeed the same either way, but the waveforms are very different. With the signal fed into the cap it charges almost instantly to +/-Vp-p, and discharges exponentially. If you feed the signal into the resistor, the cap charges and discharges exponentially to +/-Vp. The expected traces are shown below, and you can see that the charging/discharging time is the same for both configurations. There shouldn't be any ringing going on.

[pjd3]

Thank you Martin, this does infact clarify what I suspected, that the injection of the square waves in to cap first acts like a short initially, until the cap starts to fill up. I did however have a conceptual issue with the fact that the resistor is still in the series circuit and would thus still force the cap to charge up according to the resistor/cap impedance/time based situation. Although it is obvious now there is a completely different if not nearly opposite voltage waveform across the cap, I may still need to think over a cup or two precisely what is driving these events. Hey, I'm 64 years old, not as sharp as I used to be, but still tryin".

There was however, a third scenario (the initial one) that I was not able to replicate - a true square wave with typical but very heavy ringing, much like many graphs of filters I've seen. And the charging curve was there as well. Well, there is only so many ways a simple RC circuit can be implemented and measured so, maybe with some diligence I can replicate that again, just to see how that was occurring. I'll return to this thread should I rediscover that.

Thanks again Martin for your well executed explanations. Very clarifying and helpful

Best,
Phil D

[trobbins]

Another item to note could be the scope probe's effect. For no probe, just the scope input (eg. 1Meg and parallel C) would parallel the DUT C. A 10:1 probe would present 10M and lower C.

Re-arranging the DUT to have R across the scope would place DUT C in series with scope C, and perhaps not so simple an RC charging circuit.