EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: riscy00 on January 26, 2018, 04:44:31 pm
-
This is about Close Loop Feedback analyser and other PSU analyser feature contains in PWR plugin for DSOX30**T scope series.
Omicron Lab Bode 100 Vector Network Analyzer and Ripley Engineering AP300 and similar equipment which is expensive and specialised. Until I discovered DSOX3034T scope with PWR upgrade, which is capable of Close Loop Response analysis.
Has anyone actually compare between DSOC30xxT scope with PWR to actual dedicated equipment.
I found it more cost-effective solution than a dedicated unit where scope is must have instrument for all kind of electronics works.
I familiar the scope has limited resolution of 8 bit, but how this significantly impact the close loop response measurement?
Has anyone has some major issues when using PWR plugin?
-
Actually the frequency response analyzer(FRA) is not dedicated only to PSU stability measurements. Once you get to used to it, you will not want to be without. Stability measurements/PSRR/transfer functions/impedance measurements/component measurements - it is a very versatile piece of equipment. The limit of 8bit resolution is very real. You need to realize that the injected test signal is of very small amplitude. If your test signal is too big, you are moving the operation point of the psu and are measuring rubbish. In some cases, the test signal can be as low as -30dBm. And you are trying to measure this signal riding on ripple and switching noise of your regulator....
I've played with the earlier version of bode plot on Keysight scope, it was not really usable. Friend of mine tried to use 3000T on flyback when our FRA was being used elsewhere. Did not have much luck with it either. It might be usable in some cases, but I did not have much luck with it
If you want to measure isolated regulators, you might get by with sound-card with opamp buffers. The loop bandwidth is usually very much in the sound card bandwidth.
Personally I'd much rather have a 100Mhz 4channel scope and FRA, than 350MHz scope and no FRA. As usual, YMMV
Either way, I would highly recommend to contact your local distributor and ask for a test drive - both FRA and the DSOX30**T
Few observations regarding loop measurements:
1) Use resistive loads
2) Monitor the switch node. If you have constant input voltage, constant load, you should have a constant duty cycle/small jitter. There is no point of measuring loop response if your controller is jittering like hell/pulse skipping.
3) Rectified mains and 330uF cap is not constant input voltage...
4) Set the right amplitude for injection signal
5) know what you are expecting - you should know how the transfer function should look like. If there is an extra resonance peak, there is something fishy.
6) Watch out when doing Bode plot measurements, bode plot readout(text book phase margin and gain margin) is only valid if you are crossing the 0db line with 6db slope. If not, you need to use Nyquist plot. In first order system, the point closest to instability point are the zero crossing of phase and zero crossing of magnitude. If you have higher order system(voltage mode controller, and closing the loop before the ESR kicks in and fixes the slope to 6dB for example), you might have stability problem at different frequency, but you wont see this from Bode plot. From Nyquist you will.