My desktop PC's PSU is acting a bit strange as of late. I'm experiencing unexpected shutdowns, especially while the system is under load. One thing I noticed is that the PSU feels quite warm to the touch after these unexpected shutdowns, too warm in fact.

I've checked the voltages with a cheap multimeter (DT-830D) with a mild load of 5 amps on the 12v rail and all voltages seem to be well within tolerance. In fact, I even checked them while the system was under load (FurMark) via a spare molex connector and everything seems to be fine.

Now, I wish to see the DC ripple and see if it's within tolerance as well (ATX specifies a maximum ripple of 120mV peak to peak) but unfortunately I don't have access to an oscilloscope, and neither I've enough money saved-up for a decent PSU. Which brings us to the question: Is there ANY way I can measure the ripple with the cheap DMM I have?

I apologize if it sounds foolish but I've next to zero knowledge when it comes to electronics. I've been Googling about this for several days now but can't seem to find a definite answer on this topic as the information regarding this matter is surprisingly sparse.

P.S This is my very first post here so I'm sorry if this isn't an appropriate place to ask such a question. I just can't seem to find helpful information anywhere regarding this matter and 'All About Circuits' seems to be the most appropriate place on the whole worldwide web to ask such a question!

Thanks a lot for your patience and bearing with my sheer ignorance and bad English!

 

Usually that ripple specification applies over a specified bandwidth and will probably be the RMS value so whatever method you use It is not simple to compare a measured value with the specification.

 

The last Power Supply Design Guide for desktop systems I remember reading was Revision 1.31 going back to 2013. There are likely newer revisions to be had out there. The link tells you about anything you could possibly want to know including things like allowable voltage tolerances and also output ripple noise and under what load conditions. Page 16 of the linked design guide.

Something you can try is using a good quality DMM designed to measure true RMS voltage, not an average responding RMS reading but a true RMS meter and measure the AC component on your DC PSU outputs. The specifications call out mV pk to pk so what you read as RMS multiply by 1.414 to get pk and then X2 to get pk to pk value. Lacking a good scope this is a poor man's option and will only give you a rough idea and even that depends on the bandwidth of the meter used. Again, this method is a far cry from how it is done so do not put much stock in the data you get. Years ago I had a PSU which would randomly shut down. It became frequent and one day I restarted it, there was a loud pop, a puff of smoke and game over. It was also a rated name brand.

Ron

 

The specifications call out mV pk to pk so what you read as RMS multiply by 1.414 to get pk and then X2 to get pk to pk value.

This is only true for a sine wave which the ripple on the output of a smps is unlikely to be.

 

The last Power Supply Design Guide for desktop systems I remember reading was Revision 1.31 going back to 2013. There are likely newer revisions to be had out there. The link tells you about anything you could possibly want to know including things like allowable voltage tolerances and also output ripple noise and under what load conditions. Page 16 of the linked design guide.

Something you can try is using a good quality DMM designed to measure true RMS voltage, not an average responding RMS reading but a true RMS meter and measure the AC component on your DC PSU outputs. The specifications call out mV pk to pk so what you read as RMS multiply by 1.414 to get pk and then X2 to get pk to pk value. Lacking a good scope this is a poor man's option and will only give you a rough idea and even that depends on the bandwidth of the meter used. Again, this method is a far cry from how it is done so do not put much stock in the data you get. Years ago I had a PSU which would randomly shut down. It became frequent and one day I restarted it, there was a loud pop, a puff of smoke and game over. It was also a rated name brand.

Ron

You're right. PSUs die, just like all electronics. The problem is my PC which is a branded HP Z400 server. Its PSU is proprietary so replacing it isn't easy... or cheap! Not only I'll have to buy a new PSU, but also an ATX to HP Z400 PSU adapter. And if that's not enough, I'll also have to drill holes for the screws in the new PSU as Z400's original PSUs have different screw placement which isn't ATX compatible.

It's a lot of hassle!

In any case, I read somewhere that one can measure DC ripple of a power supply by selecting AC voltage on the DMM and then reversing the probes i.e COM probe on the live wire and the other on GND. Unfortunately my DMM doesn't go below 200V on AC so there's no way I can test it for myself.

Do you think there's any truth in that?

Thanks!

 

This is only true for a sine wave which the ripple on the output of a smps is unlikely to be.

Uh Oh, yes, that would only apply to a true sine wave so my bad on that note as I never thought it all the way through.

Thanks
Ron

 

Welcome to AAC!

A good quality DVM can do it. My Simpson 467 can capture positive and negative peaks:


But a scope would be better.

 

You're right. PSUs die, just like all electronics. The problem is my PC which is a branded HP Z400 server. Its PSU is proprietary so replacing it isn't easy... or cheap! Not only I'll have to buy a new PSU, but also an ATX to HP Z400 PSU adapter. And if that's not enough, I'll also have to drill holes for the screws in the new PSU as Z400's original PSUs have different screw placement which isn't ATX compatible.

It's a lot of hassle!

In any case, I read somewhere that one can measure DC ripple of a power supply by selecting AC voltage on the DMM and then reversing the probes i.e COM probe on the live wire and the other on GND. Unfortunately my DMM doesn't go below 200V on AC so there's no way I can test it for myself.

Do you think there's any truth in that?

I feel your pain as been there and done that with propitiatory server type and workstation type power supplies. You can try a good quality AC voltmeter. Having a 200 V range sort of sucks and as to reversing the probes I never heard of having to do that. You do want a True RMS responding meter. Really gets hard to call why the PSU randomly shuts down.

Ron