Measure Power Supply Rise Time at Power Up

[AMR Labs]

I would like to know how much time it takes a certain power supply to rise from zero output to the rated output voltage (15VDC) once power starts appearing on its output. I wanted to find out if the power supply will meet some critical parameters for the intended load, which has a build in processor that might corrupt its own firmware (apparently due to a manufacturer design glitch) because somehow it might get into an incorrect state if the ramp up of the power supply voltage is too "slow", or even unstable. Not sure how or why exactly this problems happens as I am not intimate to the design of the device, but I was advised of the possibility of this occurring with a slow rise time power supply. I had the device connected to a variable LM338K based power supply, all seemed to be going well, and then about 1 hour into it suddenly the device stopped working. I sent it back and sure enough, the firmware was corrupted and had to be reloaded.

So I just got this small stand-alone 35W switching power supply which seems to be very stable, and now more out of curiosity I wanted to see if there was a way to actually measure the output voltage rise time of it. With a DVM set on a fixed range of course the change from zero to 15V appears to be instantaneous, but I wonder how many milliseconds it actually takes to fully rise to 15V.

I am assuming the obvious way to go would be with a scope, I do have a Tektronix 2247A, 2213A and a 222. Since the parameter to be measured is not really a repetitive event in the usual sense of a waveform, I am guessing I will need a storage scope to capture this one time event, and I am hoping my 222 with do the trick. But I wanted to ask opinions as to how would be the best setup to capture it. Also, its been a while since I last used the 222 (my main scope is the 2247A), so if anyone could refresh my memory on how to setup the 222 scope would also be helpful and probably save me some fiddling time.

Thanks in advance.

[AMR Labs]

Well I fiddled a bit with the 222 and got the correct settings in place to capture the rise time event. It looks like this switching type power supply in question takes about 7mS to get up to the rated 15V output from power up at the time the switching circuit kicks in. Screen settings are 5V/Div and 1mS/Div. Trigger point 1.11V (its as close as I managed to get to 1V with these controls).

[tautech]

On this topic there are desirable attributes for a PSU, namely NO power ON overshoot.
For more complex circuits power rail sequencing is very important too as you don't want the inputs of a device's max rating to be exceeded in its OFF state.
Check Daves vid about power ON PSU rail sequencing:
https://www.eevblog.com/2017/05/21/eevblog-995-power-supply-sequencing/

[AMR Labs]

Hi, interesting video, a bit too long though... the guy sure likes to talk

My application only requires a single rail of +15V, but it has to have a fast (or at least not slow) ramp up time. The device to be powered up has a known weird bug that will trash the firmware if it is powered up with a slow ramp-up power supply. Apparently something to do on the brownout detector going bonkers. Just wanted to make sure the output rises reasonably fast even while under load. BTW rise time increased from 7mS to 14mS when loading the power supply with 0.8Amps, and increased to 160mS with a load that draws 2Amps. Power supply is rated for 2.4Amps max. So far I would think it should be good to go.

[tautech]

Sure but now you have a better understanding so to help avoid or mitigate any problems at power ON. 
Otherwise use a PSU that's already powered ON and then hard switch its output.

[Ice-Tea]

Otherwise use a PSU that's already powered ON and then hard switch its output.

The PSU may not appreciate that. The inrush will occur after any grace period it has and it may cycle. Kinda not what you want with equipment that behaves badly during startup. Or the hard switch (mechanical?) may bounce (same issue).

[AMR Labs]

Otherwise use a PSU that's already powered ON and then hard switch its output.

The PSU may not appreciate that. The inrush will occur after any grace period it has and it may cycle. Kinda not what you want with equipment that behaves badly during startup. Or the hard switch (mechanical?) may bounce (same issue).

Agree. I had actually thought about putting a relay between the device and the power supply, and the relay being driven by a window comparator that would sample the output voltage of the power supply, so the device would only get power when it was at the correct level, not bellow or above the window. But then the contact bounce issue came to mind and I abandoned the idea. Maybe it could be done without bounce with a MosFet as a switch instead of the relay, something similar as it was done in the video to sequence power supply rails. But I think I will be OK with this power supply, there is absolutely zero output until the switcher kicks in and then it will almost instantly jump to 15V within probably 125mS with the ~1.7Amp initial load I am expecting. So that should hopefully keep the device happy. Will be doing a power up testing this afternoon. Watch out for any mushroom shaped clouds on the horizon ;-)

[tautech]

Otherwise use a PSU that's already powered ON and then hard switch its output.

The PSU may not appreciate that. The inrush will occur after any grace period it has and it may cycle. Kinda not what you want with equipment that behaves badly during startup. Or the hard switch (mechanical?) may bounce (same issue).

Agreed, however many lab PSU's are output switched using relays and if a load, any load is unforgiving then it needs modification IMHO. Maybe even split staging so the MCU is certain to get power before any other part of the circuit.