Testing an SMPS for stability

[cupoftea]

When you are testing a 400kHz 20vin 60W, 30Vout DCDC Boost SMPS for stability, you obviously look at Vout
over multiple time bases.....say 2us/div to 50ms/div....why not look over that range?...all you have to
do is turn the timebase dial...might as well view it over that full range.

You would also look at the error amplifier output if it was available...on max load....and again over
the same timebase range as above.

As well as that, I would also look at the fet drain voltage pulses....from a timebase of some 2us/div....right up to that
maximum timebase that still allows you to see the drain voltage pulses.....whatever timebase that is....obviously on 1s/div
you wont be able to see the pulse to pulse spacing any more....but might as well look over the widest timebase possible.

However, our contractor, when looking at the drain voltage pulses, he just gets three switching periods on the scope, and if the distance
between the drain pulses is the same, and they are the same width, then he declares stability, and wont bother looking over any wider
timebase Setting. We think this is skimping it.

What say thee about this?

 

[aliarifat794]

While consistency in the switching waveform is a good indicator of stability, it doesn't guarantee it across all operating conditions or under transient loads.It will be beneficial to discuss with the contractor the importance of a more comprehensive stability test that includes observing the waveform behavior across a wider range of time bases.

 

[Easy peasy]

For a stable power supply - if the Vi is varying or the load is varying the duty cycle should be changing all the time - and hence the Drain on or oof times for a pwm system

If the Vin is quite fixed, and the load - observing the Drain at one operating point does not imply or infer stability at high Vin and/ or light load - or over load . . .

 

[cupoftea]

Thanks, would agree...the drain pulses need viewing over all the conditions you kindly describe....and over the widest possible variation of the time base wherein the drain pulses can actually be seen...obviously as you know, at say 1s/div you just get a big blurb.....so not useful at that time base