Measuring PSU Voltages

[Reloadron]

Originally Posted by Capper

well, wait one......companies are actually rethinking the multi rail design....and many are in fact offering top of the line 1000W and 1200W PSUs with a single 70-95A rail. regardless, its an ever evolving field, like the rest of the tech sector......the main principles should remain unchanged......basically we are wanting a reliable way to test each rail, the total power, both under load and at idle

Yeah, the rail thing for 12 Volts seems to be forever evolving. Part of the whole thing with PSU manufacturing is a lack of standardization to a point. There is a PSU Design Specification but really nothing binding manufacturers to it. Thus everyone seems to be trotting along to the beat of their own drum. That is partially why I seemed to feel a good test system should be somewhat modular.

Testing individual rails just involves the ability to start with little or no load on them and adding load in increments. Something like 5 Amps per step. Maybe the ability to add loads in multiple steps like 5 OR 10 Amp steps. As the load on each rail is increased the current and voltage are monitored. The watts are calculated. Most of that can be done using some minor automation. That was the basic concept in the link I posted some time ago. There was a very simple load bank for a 12 Volt rail.

Some of the problems. Switching the loads is an issue. Well from a cost point. Using even manual switches to increment a load you need switches capable of handling the current. That shows up well in the link.

Also, extending beyond the loads do we care about anything else? Do we just want to add load to the rails till the voltage folds over or the PSU shuts down? Do we even care about a PSUs regulation response times? How about overall efficiency? Do we want to look at the AC line input in watts as compared to the power delivered to a load?

Oh great! My very first thread here and I have come to the realization I managed to hijack my own thread. Sure just hint at PSU Test Benches and Ron loses it! Back on topic, if anyone has any interest in what I was looking for when I started the thread plaese let me know.

Ron

[Capper]

What we really need is something all these lines...Ii don't want to overcomplicate it

SPCR's PSU Test Platform V.3 | silentpcreview.com

I am just worried about a reliable, honest, and accurate testing method that gives people a good idea of how a PSU will perform

[Reloadron]

Capper, that was a great link. I much enjoyed the read. They approach things a little differently than I would in a few aspects. However, the seem to hit on some of the same concerns I had. Things like change in PSU design and ATX PSU Form factor.

Depending on what you guys come up with I would be intrested and happy to contribute whatever I can as to design efforts and parts.

They use smaller power resistors which as I think about it would work just fine. They also use pretty basic shunt resistors where I would go with slightly better precision shunts, simply because I have them.

They also made refrence to the Intel specification and the multiple 12 Volt rails as well as how quickly a PSU Test Bench can become somewhat obsolete.

Let me kick things around into the weekend. I should have some time to think outside the work box. If you are interested I will post some thoughts. I also want to find time to mess with my thoughts that started this thread.

Ron

[PrOLifIC_onE]

Originally Posted by Reloadron

Oh great! My very first thread here and I have come to the realization I managed to hijack my own thread. Sure just hint at PSU Test Benches and Ron loses it!

HAHA Sounds like something I would do! It's ok man!

[Hitman]

Originally Posted by Reloadron

I have been actively discussing it with Manta.

I am trying to imagine a conversation between Manta and Ron. It would be like a conversation in a different language. I get lost just reading half of what they post LOL

[Jafah5150]

Heh..I messed around with this idea once but not to the extent of automated recording.

Basically Hitman's idea of multiple DMMs...get a bunch of cheap DMMs, tear them apart and build them into a cheap plastic tool box.

A little off topic but something I would like to see manufacturers do is come up with better connectors(molex, etc..) or gold plate them(or the mobo) to give the connections more consistency.
I've seen hard core overclockers solder all the mobo connections just for added stability.

but I digress...it will be interesting to see what you come up with :)

[Capper]

Originally Posted by Reloadron

Capper, that was a great link. I much enjoyed the read. They approach things a little differently than I would in a few aspects. However, the seem to hit on some of the same concerns I had. Things like change in PSU design and ATX PSU Form factor.

Depending on what you guys come up with I would be interested and happy to contribute whatever I can as to design efforts and parts.

They use smaller power resistors which as I think about it would work just fine. They also use pretty basic shunt resistors where I would go with slightly better precision shunts, simply because I have them.

They also made refrence to the Intel specification and the multiple 12 Volt rails as well as how quickly a PSU Test Bench can become somewhat obsolete.

Let me kick things around into the weekend. I should have some time to think outside the work box. If you are interested I will post some thoughts. I also want to find time to mess with my thoughts that started this thread.

Ron

Mainly, what we need to do is focus on this from the user perspective....which is why I don't want to overcomplicate it. We can save on cost and design time by focusing on this from the perspective of what we're trying to accomplish......to help people make good buying decisions regarding the power supply for their computer.

[Hitman]

Originally Posted by Jafah5150

Basically Hitman's idea of multiple DMMs...get a bunch of cheap DMMs, tear them apart and build them into a cheap plastic tool box.

My Idea? I don't remember posting anything about that.

[Capper]

Originally Posted by Hitman

My Idea? I don't remember posting anything about that.

probably a heavy drinking night, or morning, or afternoon.

[Reloadron]

Originally Posted by Capper

Mainly, what we need to do is focus on this from the user perspective....which is why I don't want to overcomplicate it. We can save on cost and design time by focusing on this from the perspective of what we're trying to accomplish......to help people make good buying decisions regarding the power supply for their computer.

OK, I will kick it around and make a few suggestions. Simple and functional with accuracy. Thinking cheap!

I will also mess with the thread subject this weekend and get something posted.

Ron