Going about PSU testing all wrong?

[VirtualLarry]

Originally posted by: jonnyGURU
So light bulbs are out. Too difficult and expensive to properly regulate.

I would think that the real reason is that they are a static load, mostly purely resistive in nature, while PC power requirements are rather volatile and dynamic. Not just that it's hard to start switching-up the connection of various bulbs to more correctly approximate an "evenly balanced" power draw across all of the rails. I saw one of the review tests on the internet, used a mixture of both - some bulbs for increasing the overall actual load, combined with running an actual PC at the same time to more accurate test the dynamic-load behavior. That sounded like a decent compromise to me, for a home-made MacGyver-esque testing rig.

Originally posted by: jonnyGURU
I took a couple days off from work and drove up to his office where he showed me an actual PSU testing machine (SunMoon SM-268) and we did a series of tests on some different power supplies.
He had told me that the only "good" PSU test online was the one that Tom's hardware just did. But I was feeling a bit let down about that opinion because they only "benchmarked" efficiency at different loads, gave an aesthetic overview of the PSU and then concluded "none of them blew up." How is that a good review?

"Good" as in "accurate and correct way to test a PSU", or "Good", as in, "politcally correct review - don't want to single out any particular company and make their product look bad"?

Originally posted by: jonnyGURU
He told me that if you actually know how to use the test equipment and what a good "balanced" number for a power supply's given specs is, you can load a PSU up to 100% load for eight hours and not have a single problem. Tom's review was only trying to be "fair and balanced." Nobody isn't going to load a power supply to 100% load for any prolonged period of time. And all of the power supplies tested, ANY power supply that is UL listed, is GOING TO live up to it's label.

Hrm. I would have to research this more, but I wasn't aware that the UL guys actually tested the accuracy of the specs of the PSU, I thought that their only job, more or less, was to test the safety issues of the supply, whether or not it would burst into flames, withstand high-voltage arcing, shut itself down if overloaded, etc? So just because it has the UL number, doesn't mean that the supplies' mfg ratings are "accurate". (But it does mean that the mfg cared enough about the quality and safety of their supplies to submit it for testing, and that helps differentiate those supplies that even have approval at all, from those bottom-of-the-barrel ones that don't.)

I especially wonder about that, with regards to continuous-load numbers, since many cheaper (but UL-listed) supplies don't even list the continous-load numbers, only "max" load numbers, with a caveat about max load being sustainable only for 10ms or something like that. So given that, how would the UL guys even theoretically be able to test a continous load for eight hours, if they don't even know what that load is supposed to be! So, color me skeptical on that scenario.

Originally posted by: jonnyGURU
So what aspects of a power supply do you test? Features and efficiency.
So he's saying to buy a power supply based on features and the specs on the label and if efficiency is important, look at that as well?
Hmm.... Not good enough for me. I like blowing stuff up.

*chuckle* Sounds like a good plan to me.

Originally posted by: jonnyGURU
We tried to knock a few tests out.... We did an OCZ, an Enermax, a PCP&C, Antec and Ultra X-Connect and X-Finity before we ran out of time (it was getting late and I had to get home so I could go to work in the morning.)

Much like THG's review, all of the aforementioned power supplies handled the full loads I hit them with as per their label... AND THEN SOME. Some characteristics are worth noting and that's what I'm doing here. ALL of these power supplies are VERY GOOD despite smell, heat and small puff of smoke. NONE of these power supplies should be frowned upon because in the end, they were ALL run OVER label specs, and like I said, none blew up.

Hey, no fair, I thought that you said that you were going to blow stuff up?
(I'm with Zepper on this one - test 'em till they die. That's the only way to find out how much actual "headroom" there is. Much like testing CPU frequencies for "binning".)
Still.. did you have an O-scope hooked up, to gauge the regulation/voltage-swings, under heavy load? Just raw load-testing doesn't do all that much good, not only does the supply have to supply the power, but it has to be stable enough to use. Most of the cheaper supplies will supply that much power, in the absolute sense, without blowing up (hopefully - as long as they are UL listed), but that power may be too unstable to use. That's the key metric as far as I'm concerned (once you get past the, "didn't explode on the test-bench" stage).

Originally posted by: jonnyGURU
The PCP&C sat there like I didn't have a load on it. Seriously. The rails never fluctuated and it never got hot. I actually laughed out loud. Even though the power supply was 510W, it was doing 600W easy and the 12V was at 12, the 5V was at 5, etc. Too bad the power supply costs $225!

That's very impressive, and proof that those supplies (and the Fortron units that they are apparently based upon), have a lot of extra "margin" in them, that the cheaper supplies don't have. But that's also why they cost as much as they do. :|
(I am slightly curious about the "never got hot" part, I thought those had a reputation for "running warm"? Just wondering if perhaps the "on" button didn't get flipped for that test... was it even warm at all?)

Originally posted by: jonnyGURU
The X-Connect was the most stable on the heavy 12V, but had the biggest +/- going from idle to full load. Always within spec, though. The 5V test sent the 12V into a serious dip, but the Ultra is engineered with a high 12V for Prescotts and SLI, so that didn't surprise me. There was a slight smell, but not as bad as the Enermax. When we shut the machine off, I did so suddenly after the full load. I guess I learned that I need to gradually let the PSU down because the sudden stop of the fan caused a sudden concentration of heat, which resulted in a small puff of smoke. Now I can see why Raidmax's ECASO (Enhanced Cooling After System Off, where the fans spin for three minutes after power off is so important. After the PSU cooled back down, I fired it back up and it seemed to be OK.

See, that's actually interesting. Back in the day, the +3.3v/+5v lines were "combined wattage" (as you will see on most labels). Based on your testing, it seems as though the +5v/+12V on those Ultra X-Connects may be "combined wattage" instead, which would perhaps explain a few things, and how they manage to pull off such a high rating on each, based on the size of the components used to build it. (Going by rule-of-thumb here, not really scientific, but I hope you can get what I'm suggesting.) That actually could make some sense, as these days, either the +5V, or the +12V, is going to be loaded heavily, but not generally both. That might also explain, partially, failure under testing, if this +5v/+12v-combined issue wasn't taken into account. I would attempt to follow up on that with your contacts at Ultra and their engineers, to find out if that is indeed the case.

As far as the fans running after the PSU is shut off, I've wondered about that too, and not just the PSU, but the CPU heatsink's fan too, on some really high-powered systems, and even a bank of actively-cooled HDs. Killing fan power will allow their temps to briefly spike before settling back down. (Same reason most cars these days will continue to run the radiator fan off of battery power after the car is stopped, especially on a hot day.) I would like to see that feature implemented on more systems in the future, it sounds like a really good idea on the whole. It's good that you brought up that issue. Hopefully more PSU makers will take note.

Originally posted by: jonnyGURU
He nodded and with the 12V already at 30A, I started taking the 5V up from 20A. I don't remember exactly where the amperage was at for the 5V, but I remember seeing 700W on the load tester and just over 900W at the AC outlet when the lights went out in the office. I guess between the PSU and all of the other test equipment on the same outlet, we pushed the wiring in the building beyond IT'S capability and the breaker tripped. My friend then told me that that was to be expected because the X-Finity 600W is essentially two 400 or 450W power supplies mounted on top of each other and then "underclocked" to a total 600W so it didn't get too hot. Nice.

Hrmph. Sounds like this company had no business testing PSUs, or running a PSU testing jig off of a non-dedicated circuit in the building.

Originally posted by: jonnyGURU
He said back in Taiwan, they have racks loaded with these so they could test several units at a time. He could lend me one unit as long as I reported to him with any findings, problems, etc. So now I'm back at work with a SM-268 in a cardboard box sitting in my trunk. I will have to take the family to the fair on Saturday, but come Sunday I'll be hooking all of this stuff up and will continue testing power supplies.
If anyone has any questions or wants me to try anything, please ask! And Anand... if you want to do a PSU shoot out, let me know!

Oh, ok, I take that last comment back, if these things are meant to be portable. I guess they aren't quite as physically large as I had assumed? (Size of a beloved patriot-tonk piano? Smaller? I guess...?)

 

[jonnyGURU]

Yeah man... Read the whole thing before you reply. Geez!

I wasn't at a factory. I was in the guy's office. I think he just had the tester there so he could show potential clients load capabilities. Keep in mind, we had the AC Usage monitor, the load tester AND the power supply all plugged into the same outlet of an office.... not a factory. God knows how many amps were being pulled through that outlet.

Portable? It's a 4U rackmount. So even though I could easily carry it to my car, it's not really "portable."

"Back in the day" as you put it (aw, hell. I put it that way too.) 3.3V and 5V combined was just that much more important simply because the other rails WEREN'T as important. It has nothing to do with how the two combined interact with the other rails. It was just a verbal indicator that this power supply has a combined wattage of X on these rails. A more important number "back in the day" than total wattage.

I was talking to another guy who's actually pretty good with PSU's, but not too up to date with the way the power was being used today. He was wondering why a certain 500W power supply had had less 3.3V and 5V combined than another 350W. He was actually trying to imply that the 500W couldn't possibly be a 500W because it didn't have as high of a 3.3V and 5V combined. I told him it was because it had more on the 12V and that he has to factor that into the total wattage. He then said, "then why is it less on the 3.3V and 5V?" I then had to explain to him that the load isn't as high on those as they used to be and manufacturers were migrating over to a stronger 12V. Otherwise, if they just had made a stronger 12V rail on a PSU with the same 3.3V and 5V specs, you would have a 600 or more PSU that was way overkill because it would have way more power than anyone could ever need.

It really doesn't have anything to do with one being MORE interactive with the other. It's a switching power supply. They're all interactive and all of the PSU's above would drop voltage on one rail if the amps weren't squat on the other. The only power supply above that didn't have one rail drop when the other was not loaded was the PCP&C and that's just because that is "God's Power Supply" (God has assloads of moolah.)

Remember that article I linked a couple PSU threads ago? The old IBM XT the guy talked about. The 5V rail was the primary spec on the PSU becuase the CPU, RAM, everthing but the hard drive motor, used the 5V. Since the 12V rail was only used if a hard drive was installed, and a switching power supply is naturally interactive, they had to install giant ballast resistors on a 12V connector if there was no hard drive in the machine. Otherwise, the 5V would drop so low that the PC wouldn't even turn ON!!!!

The ONLY solution to that "problem" would be to have one PC power supply actually be 4 or 5 power supplies built into one housing. But look at how big a simple 6V 1.5A power supply is for a paper shredder (have one right here.) It's big. It's heavy. Now lets say that needs to be able to do 20A (and 5V instead of 6V.) How big does that power supply need to be and how heavy? Now you need one for the 12V, the 3.3V.... Not practical.

Even if you look at one of Antec's "True Power" PSU's which claim to have "independent voltage rails," you'll find that it's actually the same old power supply, just with independent regulation. The rails are still interactive and still have the same load/no load issue... only the voltage is "cleaner" due to the independant regulation. The grossest misrepresentation of independant rails is Silverstone and Enhances quad rail PSU. Add up the numbers and you'll see, those aren't really four 12V rails. Just one 12V rail with four additional regulators. So one would think that you could rob Peter to pay Paul if you overloaded one rail and there's another that isn't as loaded up. That's what I thought. I was wrong. The regulators actually PREVENT that, so the power supply is either going to shut down do to overload protection or the voltage is going to drop until the PC locks up or crashes because it's under spec.

As for the UL, they do test the PSU's for the accuracy of their labeling BECAUSE of safety. A PSU is submitted with a sheet with specs. That's the specs the power supply is tested under and that's not to say that the specs submitted to the UL aren't going to be different then what's on the label (although unlikely.) Let's say a PSU is submitted that's documented to do 500W and they load it to 500W and it blows up. THAT'S not safe. Wiring diagrams are also submitted, but aren't published online likely due to potential patent problems.

As for blowing things up.... I told you I wasn't done! I only did some preliminary testing while I was over there. The only PSU I tried to blow up got the best of the mains. I'm going to do the same three tests with all of my other power supplies and then I'll come up with a new test to throw at them until I blow something up. That's why I'm up to suggestions. Throw some REALISTIC scenarios at me and I'll program the loader accordingly.

 

[jonnyGURU]

I missed this one.....

Originally posted by: VirtualLarry

"Good" as in "accurate and correct way to test a PSU", or "Good", as in, "politcally correct review - don't want to single out any particular company and make their product look bad"?

Good as in accurate. Much to my surprise, most of those PSU's were not given to THG by manufacturers as per reps I know at different companies that had units tested, so you can throw that theory out the window.

I know you want to see things blow up, and I'm not a big THG fan, but how would the review not be "good as in accurate" if they tested the PSU within spec? If the guy knows how to use the machine, tests the PSU's at max spec, which is actually a worst case scenario, and nothing blows up, that doesn't make it a bad review.

If a reviewer blows something up and doesn't state at what specs something blew up at and doesn't try to figure out why something blew up and doesn't contact the manufacturer when something blows up then THAT is a bad review IMHO. I've got the memory set on this rig to three different load tests. Two additional memories are used up for dual rail PSU tests. If something blows up, I'll tell you when and where it blew up.

Now, I did have a PSU blow up. THIS was scary actually, because I didn't expect it to blow as it was a fanless 400W Coolmax and I only had a 200W load on it for less than 3 minutes. I only had the load tester set to 12V: 10A, 5V: 10A and 3.3V: 3A. Smoke started POURING out of the unit almost immediately as if a little man with a bad nicotine addiction was inside. It had to have been a defective unit because it was the only thing I smoked and I smoked it at half of it's max rating in a very short period of time. I wish I had pictures.

 

[VirtualLarry]

Originally posted by: jonnyGURU
"Back in the day" as you put it (aw, hell. I put it that way too.) 3.3V and 5V combined was just that much more important simply because the other rails WEREN'T as important. It has nothing to do with how the two combined interact with the other rails.

My understanding was that the actual power-regulation circuitry for those "combined" rails was in fact, directly "combined". Hence why Antec introduced their TruePower line, with independent regulation for each of the primary voltage rails. As you said further down:

Originally posted by: jonnyGURU
The ONLY solution to that "problem" would be to have one PC power supply actually be 4 or 5 power supplies built into one housing. But look at how big a simple 6V 1.5A power supply is for a paper shredder (have one right here.) It's big. It's heavy. Now lets say that needs to be able to do 20A (and 5V instead of 6V.) How big does that power supply need to be and how heavy? Now you need one for the 12V, the 3.3V.... Not practical.

That's more-or-less essentially what Antec did. (Btw, if you're talking about a power-brick type supply, for purposes of comparison, most of the cheaper ones of those are linear, not switching, AFAIK, and that's why they are "big and heavy". Switching supplies are lighter and smaller, of course, although they cost slightly more to make. The only "power bricks" that I've seen that are switching, are usually laptop AC adaptor/charger units, and you can tell that they are much lighter. The other ones aren't much more than a custom-wound xformer and a bridge rectifier, and maybe a cap if you're lucky.)

What I was actually suggesting, with the Ultra X-Connect being (potentially) designed differently, with the +5V/+12V rails combined, was actually to suggest that you were right earlier, claiming that the testing on the Ultra was flawed, if indeed those rails are shared. In prior supplies, the +12V was largely independent of the +5V, so placing a heavy load on one wouldn't directly affect the ability of the other to supply power. If the test jig was set up to test this way, placing a heavy load on both at the same time (since in prior supplies, the +5V and +12V were usually independent), but were combined on the X-Connect, then yeah, that could explain why it blew up during testing. Therefore, it would be very interesting to find out if that is true, and if it is designed slightly differently in that way.

Originally posted by: jonnyGURU
It really doesn't have anything to do with one being MORE interactive with the other. It's a switching power supply. They're all interactive and all of the PSU's above would drop voltage on one rail if the amps weren't squat on the other.

The issue of a switching supply requiring a load, is seperate and different from the issue of a supply sharing part of the circuit that regulates two DC output rails. When I mentioned the interactive loads, I meant the demands of the device being powered, not the inputs in the supply. That's a function of the regulation ability and accuracy, and is not directly related to whether the rails are combined, although combined rails likely would show less-precise regulation under load.

Originally posted by: jonnyGURU
Even if you look at one of Antec's "True Power" PSU's which claim to have "independent voltage rails," you'll find that it's actually the same old power supply, just with independent regulation. The rails are still interactive and still have the same load/no load issue... only the voltage is "cleaner" due to the independant regulation.

If they've changed from combined regulation to independent regulation circuits, how is that the "same old" PSU? The load issue still does exist, yes, since that's a factor that's apparently inherent in how switching supplies work. But having independent rails, should allow any particular one of them to be loaded down to the max, without significantly affecting the regulation or ability to supply current of the others, at least up to the limits of the input AC power source, given the AC-to-DC conversion efficiency factors involved.

Originally posted by: jonnyGURU
As for the UL, they do test the PSU's for the accuracy of their labeling BECAUSE of safety. A PSU is submitted with a sheet with specs. That's the specs the power supply is tested under and that's not to say that the specs submitted to the UL aren't going to be different then what's on the label (although unlikely.) Let's say a PSU is submitted that's documented to do 500W and they load it to 500W and it blows up. THAT'S not safe. Wiring diagrams are also submitted, but aren't published online likely due to potential patent problems.

Yes, but likewise, as long as the PSU doesn't blow up, that doesn't mean that the specs printed on the side that claim "500W", after UL testing, also somehow "guarantee" that that PSU will be able to provide a properly-regulated, in-spec, 500W of continuous-load power to the PC connected to it. All I'm saying is that, AFAIK, the testing done by the UL is safety testing only, and doesn't test the quality of power-regulation, nor the accuracy of the claims thereof. As long as it doesn't fail and cause a safety issue, they get the "approval". Again, AFAIK. I could well be wrong on this. But the number of supplies that I've seen, that have the UL logo, that aren't exactly.. "accurate" specs, along with the rampant re-labeling of lower-end supplies (stickers on top of sticker), tends to lend credence to that.

Originally posted by: jonnyGURU
As for blowing things up.... I told you I wasn't done!

Cool. We want vids. (Like the infamous TH AMD/Intel CPU "burnup" videos.)

 

[jonnyGURU]

AFAIK all of the rails are combined on all switching power supplies so far. 3.3V+5V rails were never independent of the 12V rail on any power supply I know of. As I said, 3.3V + 5V combined was only a label. When I said that the Antecs were regulated I meant to say that they were regulating a rail that has already been regulated. Some people have the misconception that the TruePower is actually independent switching power supplies in one housing. Apparently you are one of them. That is simply not true. Even my Antec rep will tell me that. Not to say that a TruePower isn't an excellent power supply. There's just nothing pioneer about them.

You are correct about power bricks being linear and not switching. My bad. Bad analogy, but I couldn't think of any other and saw the big ol' brick for my paper shredder sitting there. A multiple switching power supply is still going to be big and heavy and run hot. My friend told me he had a client interested in putting out a true quad rail power supply (not sure if PCP&C is truley quad rail as I don't think it's in production yet and nobody has seen anything more than pictures) The thought was to stack two power supplies. Put a 5V/3.3V/-5V/-12V/5VSB on one side (a regular PSU without a 12V rail) and another 12V only power supply on the other side that had four seperate rails. He said that the 12V only power supply would still be as big as a regular power supply in order to give adequate amperage on each rail and keep them somewhat "independent".

UL claims that they obtain frequent samples and specification documentation, so I don't know how to address your point beyond what I've already said. It sounds like they keep a pretty good leash on those who pimp the UL logo. On the flip side: You'll find that some PSU's aren't actually UL listed, but their components are. I was told that ANY change in any spec can effect their listing. For example: The PC Power and Cooling: It's made overseas, but it's assembled in the USA. There's potentionmeters made readily available to the assembler/tester so the PSU can be tweaked. I don't believe this final product is UL listed. Look at a PCP&C and you'll see a UL logo, but no number. That Coolmax that blew up? It had a UL logo, but no number. When I looked at the website, it said "assembled from UL components" which leads me to believe that the PSU was UL tested as a normal PSU with a fan, but not without the fan in it's aluminum housing. If it had been, I don't think it would've gotten a UL number and I think it would've been displayed on the box somewhere!

I'd love to shoot video. I've even got a DV camera, but that would take SOOO much time and tape. I mean, testing will take enough time and then I have to set up the camera and edit out the dead air. Proper testing is to put the PSU on the tester for an hour to "warm up" even before obtaining data, and then you have to let it run for 8 hours for it to be a "pass." I might get done by X-Mas '05!!!!

 

[jonnyGURU]

Oh, and what I said about the Ultra... I was just repeating something their rep told me. I'm not going to even pretend to understand HOW it works. It's got "Continuous Power" (which is supposed to be what the 'X' means) for "power on demand" (it does ramp up better on the tester than the other PSU's) but I don't know how this is accomplished.

I know what you're saying, and it sounds logical... but the fact remains that THG tested the same power supply with pretty much the same equipment at 100% load and it didn't blow up. I'm just leaving that one as chalked up as one of two things: Bad PSU or operator error.

Operator error is highly likely. After I was "trained" to use this machine, I almost blew up two or three power supplies by just touching a button. Fortunately, there was someone nearby to grab my hand.

 

[justly]

jonnyGURU, thanks for all the information from your field trip .

I think you have put out more usefull information on the subject than most (if not all) the "hardware review" sites have with their "roundups".

I will definatly re-read this thread again (to make sure I dint miss anything) and I might have a question or two after I do, thanks again for sharing your experiance.

 

[jonnyGURU]

Just got some more.....

I'm getting a couple dual rail power supplies. The two rails on one are rated at 18A and 16A. We're under the impression that they are not really dual rail (like how you add up the four rails on the Silverstone/Enhance and come up with something completely different.)

The test: Load test it as if it were a single 12V rail and bring it up to 34A. If it fails, it's truly a dual rail. If it passes, it's not a dual rail, it's just a mock up in an effort to help sell power supplies. We'll see.

PC Power and Cooling doesn't lie about their dual rails. Note how they rate their dual rails as a single rail combined maximum and not individually. I will reveal the name of the imposter after I perform my tests next week.

 

[DanDaMan315]

Why does it say in the subtitle that you are stopping the testing?

 

[jonnyGURU]

I'm stopping testing with the light bulb rack.

Guess you didn't read the progression of the thread.

 

[jonnyGURU]

There. I edited it for better clarity.

 

[VirtualLarry]

Originally posted by: jonnyGURU
AFAIK all of the rails are combined on all switching power supplies so far. 3.3V+5V rails were never independent of the 12V rail on any power supply I know of. As I said, 3.3V + 5V combined was only a label. When I said that the Antecs were regulated I meant to say that they were regulating a rail that has already been regulated. Some people have the misconception that the TruePower is actually independent switching power supplies in one housing. Apparently you are one of them. That is simply not true. Even my Antec rep will tell me that. Not to say that a TruePower isn't an excellent power supply. There's just nothing pioneer about them.

You are correct about power bricks being linear and not switching. My bad. Bad analogy, but I couldn't think of any other and saw the big ol' brick for my paper shredder sitting there. A multiple switching power supply is still going to be big and heavy and run hot. My friend told me he had a client interested in putting out a true quad rail power supply (not sure if PCP&C is truley quad rail as I don't think it's in production yet and nobody has seen anything more than pictures) The thought was to stack two power supplies. Put a 5V/3.3V/-5V/-12V/5VSB on one side (a regular PSU without a 12V rail) and another 12V only power supply on the other side that had four seperate rails. He said that the 12V only power supply would still be as big as a regular power supply in order to give adequate amperage on each rail and keep them somewhat "independent".

Ok, well, to attempt to dig into finer details, would be getting beyond my capabilities, since I'm not an EE. But my understanding was that, on many supplies, the +3.3v/+5V lines were indeed combined, and Antec's splitting them out in their "TruePower" line, was supposed to give much-improved regulation and more accurate continuous-load wattage specs than prior supplies. If that's not true, and it was all just a marketing gag, well, I don't know what to say, other than I would be forced to never trust Antec again. (If what you are saying is true, that's rather despicable marketing, IMHO, right up there with "USB 2.0" devices that are only capable of 12Mbit/s transfer rates.)

I mean, at some point, all of the DC outputs coming off of a PC switching PSU, originate at the AC input line. But the difference is at what point they are split off or not. I think that Zepper was on to something when he mentioned transformer windings and current capacity, and jives with what I think you were trying to get at with your "power brick" example. In order to provide truely independent current capacity for each rail, they would have to each be connected to the AC line with their own step-down transformer, I think. If you look at most supplies, they do indeed have two "big" transformers, one usually slightly larger than the other. I'm guessing that the larger one is the combined +3.3v/+5v one, and the smaller one is the +12v one. If that's true, then that is what I was talking about in terms of "combined" vs. "independent" power rails, in terms of the ability to supply current. That matches up with what Zepper said about the increased current draw on one of those lines, decreasing the available current on the other line, if they are both "fed" current through the same transformer. From that point, they have to be rectified, and regulated. You can use multiple voltage regulators in parallel, in order to obtain tighter regulation, even though the limit on overall current capacity is really more likely due to the transformer wiring than anything. (For example, modern CPU VRMs are three-phase, with three parallel but usually identical regulators in operation, and they each share part of the load.) I'm not really sure how they obtain the -5V/-12V lines, since I don't think that they have another set of smaller xformers, although perhaps they do. I should crack open some of my older dead supplies and take a much closer look, I guess. So take that with a grain of salt, since I'm not an EE, but that's my understand of how these things work, and how the "combined" rails are derived.

Originally posted by: jonnyGURU
On the flip side: You'll find that some PSU's aren't actually UL listed, but their components are. I was told that ANY change in any spec can effect their listing. Look at a PCP&C and you'll see a UL logo, but no number. That Coolmax that blew up? It had a UL logo, but no number. When I looked at the website, it said "assembled from UL components" which leads me to believe that the PSU was UL tested as a normal PSU with a fan, but not without the fan in it's aluminum housing. If it had been, I don't think it would've gotten a UL number and I think it would've been displayed on the box somewhere!

That's interesting. I've sort of wondered in the back of my mind what is the real difference between "UL" and "RU" and whether they have a number below or not. I know that the number indicates their file, but it's curious that some can get the logo approval, but not have a file number attached.

Originally posted by: jonnyGURU
I'd love to shoot video. I've even got a DV camera, but that would take SOOO much time and tape. I mean, testing will take enough time and then I have to set up the camera and edit out the dead air. Proper testing is to put the PSU on the tester for an hour to "warm up" even before obtaining data, and then you have to let it run for 8 hours for it to be a "pass." I might get done by X-Mas '05!!!!

Hmm. Perhaps you could stage an early PSU test demo.. say.. for July 4th? You could use a PowMax as the test subject...

 

[jonnyGURU]

Yeah... I'm no EE either, but you and I both know what those transformers are for. The diagram for the quad rail double decker I saw had three transformers. Of course, the 12V side only needed one for all four rails since it's 12V across the board.

I do have one of the Powmax Demon 580's I want to plug in. I'm not sure I want to do it first or last. I guess I'll just wait until I have enough courage to risk potential 3rd degree burns.

 

[jonnyGURU]

Here's where I am at so far:

http://www.jonnyguru.com/psu3.html

Sometime tonight, I'll do the next page which will explain each of the five tests I've assembled and how and why they're executed.