square output UPS vs. bulk capacitor voltage

[i4004]

Re: square output UPS vs. bulk capacitor voltage

dunno why paul gave up there; somebody should of told him caps don't need to be inside the psu case...hehe...

[davmax]

Re: square output UPS vs. bulk capacitor voltage

Just a note to state that I have called off the need for APC to respond. We understand the problem and work arounds. The identified issue relates to a modified square wave UPS being attached to a capacitance load without any substantial resistive load to stop waveform ringing.

[rau]

Re: square output UPS vs. bulk capacitor voltage

Hi, I had read this topic a while ago and had also brought this up on a brazilian forum a while ago (I'm brazilian) and some more knowledgeable people there had come up with the same conclusions posted here and they were surprised about this issue, they had never thought about it before...

I'm posting this message because I wanna know if APC replied about this issue.

I know we now understand the causes of it and workaround, but if APC haven't replied on this, wouldn't it mean they're trying to hide the fact? I think they'll never admit an UPS can demage a PSU, but I think we should insist on getting an answer from them. They should do something to fix it at least on new models released.

My APC UPS has some kind of feature that after a few seconds the APC PowerChute softwares commands Windows to hibernate, the UPS automatically turns itself off, it starts blinking it's LEDs and a few seconds later, turn off, but it only does this if powerchute commands Windows to hibernate. The UPS remains on if I turn off the PC manually. That shows me APC might know about this issue as well as other brands, but they are too afraid of losing money by admtting it. So afraid that they don't even talk about it or give any recommendations on the manuals. They should at least warn people to turning the UPS off after turning off the computer.

A few months ago I posted about this subject on the APC Forum, but no one replied!

[gonzo0815]

Re: square output UPS vs. bulk capacitor voltage

I think that APC has certainly done an internal problem analysis (if it really reached their attention), but may be they decided that the practical adverse effect are not that problematic, then they may sound first.

The reason for this may lie in the following arguments:

Capacitors, especially those from the known good sources tend to accept some higher voltage without consequence.
Capacitors just reformat their isolation layer and adjust to the higher voltage (within some limits).
May be leakage current increase, but this should not affect the operation significantly.
May be endurance decrease, but i think again probably not really significant.

APC recommends to use PFC appliances, as this will reduce the needed USV capacitance considerably. Apart from this, most cooperations / professionals use PFC controlled devices any way, as they have to pay for their PF.

So in the end, it may not worth the hassle with negative press, recalls etc..
Then they may just decided to ask their insurance company to bear the minor damages which may occur.

[Hibone]

Re: square output UPS vs. bulk capacitor voltage

from this post
http://www.jonnyguru.com/forums/show...7&postcount=43

and this article


seems quite clear that the problem is boost converter...

since transistor on the primary shoud support more than 400v, do you think that replacing primary caps with higher rathed ones can solve the problem?

[rau]

Re: square output UPS vs. bulk capacitor voltage

I think that would be a solution, but would make the PSU's more expansive, wouldn't it? What's the price difference between a 400V and 500V capacitor? Or a 200V to a 250V?

Isn't there a way for the UPS manufactures change this kind of bahaviour in the UPS itself? I think that would be a more reasonable approach to the problem.

[PCBONEZ]

Re: square output UPS vs. bulk capacitor voltage

The problem exists in older designs of UPS that were [relatively] inexpensive consumer grade models. The much more expensive server grade UPS with sine wave vs stepped square wave (or similar) don't have the problem.
When those consumer models were designed and sold consumer grade PSU's with PFC were rare or non-existent.

The UPS manufactures have no interest in reworking old designs to make them work with newer PFC type PSU's that they weren't originally designed to be used with.
[That's rather like asking motor car companies to redesign 50's/60's cars to use unleaded gas.]

They would rather you buy a nice shinny new UPS that's designed to handle PFC.
.

[PCBONEZ]

Re: square output UPS vs. bulk capacitor voltage

What's the price difference between a 400V and 500V capacitor?

400v - $12.83
http://search.digikey.com/scripts/Dk...name=P10063-ND

500v same series - $30.74
http://search.digikey.com/scripts/Dk...name=P13315-ND

And remember the UPS models with problems were the 'economy' models of the time.

.

[KeriJane]

Re: square output UPS vs. bulk capacitor voltage

Wow!

Thank you for all the information and wonderful links.

When I bought my Mac Pro, I researched about what kind of UPS to buy. The concencous was that a Mac Pro does not play nice with square or stepped square wave UPS and that only a 1000VA or higher Pure Sine Wave UPS gives acceptable results, especially if the computer wakes out of Sleep mode.

Even an APC 1500VA BackUPS often overloads when a Mac Pro wakes out of sleep (but not a normal start from "off") with its 980w. APFC PSU (with 3 450v primary caps BTW) APC acknowledged this and recommended using a Sine Wave UPS with Mac Pros.

So I saved up for a bit and got this used APC SmartUPS SUA1500 Pure Sine Wave UPS that weighs a TON!
It works great too. Even with the used batteries the system will run at light load for nearly an hour and that's with 4 internal drives, a 23" monitor and an external 2-drive RAID enclosure.

The APC software is OK under Windows, but there isn't really any good Leopard software. If I plug in the USB cable Leopard itself gives some rudimentary data from the UPS and offers different "Low Battery" and "Runtime Dependent" options but that's about it.

Having read this thread and some of the links, I'm really glad that I got this one. Really, with a $3000 computer that uses a proprietary $300 PSU, it would be stupid to get anything else.

Have Fun!
Keri

PS- My used SUA1500 was way under $200- closer to $100 including shipping!
PPS- No, I did NOT pay full list price of $2799.99 for this computer!

[i4004]

Re: square output UPS vs. bulk capacitor voltage

smartups are not awfully expensive as they used to be....

[mdm]

Re: square output UPS vs. bulk capacitor voltage

One question, is it safe for me to use a non sine wave UPS in 220VAC for HX850w APFC since it has two capacitors?

http://www.jonnyguru.com/modules.php...tory4&reid=153

Thanks'

[mdm]

Re: square output UPS vs. bulk capacitor voltage

Travis @ jonnyguru forum told me:

It probably can cope with simulated sine-wave but two primary capacitors don't change anything here. But those caps are connected in parallel, not in series.

so I got an ansfer, but have one other question for you guys:

I noticed that when I have my halogen lamp dimmed and PC even shut down (easier to hear) my PSU make buzzing noise, and stops when lamp off or full power. And of course when lamp unplugged from the wall no buzzing sound either, and same behavior even with different lamps (with dimmer). I know the reason is some harmonic resonace through the line with the capacitors or coils of the PSU etc..

But my question is, can it harm/damage the PSU in any way? With that noise coming from the grid and interfering with the PSU I am afraid it can maybe cause bad input voltage noise, change the input sine wave or EMI (dirty power) to the PSU that the PSU would not like and transfer it to my components too. And finally damaging / killing my PSU or components.
And if so should I put a line filter like for example outlets for PC with voltage filtering having capacitor, coils and varistors, an AVR or similar? If yes what kind do you suggest?
(And moving the lamp/s doesn't help - same line)

Thanks in advance!

[PCBONEZ]

Re: square output UPS vs. bulk capacitor voltage

No.
If you are on 220 VAC it probably can't cope with a simulated sine-wave supply.
With 120 VAC service [like Travis has experience with] your odds are better but no guarantees there either.

Your noise is because your PSU's input filter is insufficient or is defective.
[Perhaps already at it's limits due to a simulated sine-wave.]
I wouldn't expect actual damage from it, but then again, if it's loud enough to hear... ??
I suggest trying an incandescent lamp or get rid of the dimmer.
.

[370forlife]

Re: square output UPS vs. bulk capacitor voltage

Travis lives in Beijing, China. They use 220V there.

[mdm]

Re: square output UPS vs. bulk capacitor voltage

@ PCBONEZ

Then the more expensive models awaits me I know APC Smart-UPS have true sine-wave, but do they have AVR like the Back-UPS RS series, just for the lamp caused buzzing issue.


I still do not have an UPS, the noise is only because of the lamp! Maybe the input filter make the noise interfering with the lamp? Just guessing. Or should I RMA till its new?!

Because I have an answer at corsair site too from RAM GUY:
It sounds like however the lamp dims itself, its messing with the power signal and causing some harmonic resonance like you mentioned. It would be unlikely that the noise would indicate any increased chance of failure, and I would suspect that the noise is more of a cosmetic problem, rather than a functional one.
Doesn't sound so sure like you, is he?

Thanks for helping

[PCBONEZ]

Re: square output UPS vs. bulk capacitor voltage

Travis lives in Beijing, China. They use 220V there.

Well then, Travis is just wrong.
Pretty well known the 220v folks have more problems than we do with that.

~SOME~ APC Smart-UPS have true sine-wave. - Probably most do but I've run across some that don't. Specifically those I saw that didn't were older rack-mount units but before I saw them I thought [like you] that -all- Smart-UPS were true sine-wave. I was shocked to find one that isn't. Now I double check the spec sheets to be sure before I buy.

[I suspect, not knowing how that dimmer works] The dimmer is introducing a ripple frequency [standing wave] into the AC that is out of the expected freq range of any AC line filters.
It's probably causing the coil [transformer] in the PFC to vibrate a little.
The switchers switching action should inherently kill that through normal operation before it gets to the PSUs output so it won't hurt the PC.
Might eventually kill the PFC transformer but PFC only helps the power company anyway so just bypass the damned thing.

The 'fake' sine-wave is more of a problem on 220v because the PFC drives the input voltage to just short of double what is designed for and many 220v designed PSUs only have 420v or even 400v input caps. [If you have 440v or 450v input caps and other parts on the input side you may be okay.]
Over here on 110v AC the input caps & parts are usually 220v or 240v which is usually just enough. If one gets a cheapie PSU with 200v caps or the power line is pushing towards 120v with anything less than 240v caps that's when we have troubles.

[370forlife]

Re: square output UPS vs. bulk capacitor voltage

Most psu's with two 200v caps have them connected in series with about 320V across them while live. They all run on 220V, when the voltage selector is in 120V position it is in a voltage doubler circuit. My APFC solytech and HEC Zephyr have ~340V across the dc legs of the bridge rectifier. I've never seen them with anything less than 450V, maybe a cheapie, but chances are not many cheap companies are building APFC power supplies. I've also seen them with two 200V caps in series though.

[mdm]

Re: square output UPS vs. bulk capacitor voltage

I checked SC series in not true sine-wave, all other Smart-UPS are.

IF so, who cares about the power company

Saving for expensive UPS then.

[PCBONEZ]

Re: square output UPS vs. bulk capacitor voltage

PFC or APFC is -required- by law in Europe since a few years ago.
The power companies over there used their leverage and got the laws passed through to beef up their profits.
Anyone that wants to sell in that market HAS to have some form of PFC to be legal.
Lots of cheap companies want to sell in Europe.
- Follow... ? - They ARE doing it, and have been.

On a simulated sine-wave through PFC on 220v line the input side can get kicked up to around 420v-440v across those two 200v caps in series. [Specifics depend on how the sine-wave is simulated and the PFC/APFC design in use.]
-
I suspect PSUs that take 120v-or-220v without the switch have less of a problem because those I've seen [when I was paying attention to that] have 440v or 450v caps, but I don't have any non-sine-wave UPS's around anymore to test that theory and I don't have any newer cheapies around with native [non-switched] dual-volts input.
.