UPS Question

[jrga]

Hey all, just bought a PowerWalker VI 2200 LCD for peace of mine however after about 30 mins runtime it will always click into AVR mode and the fan ramps on. The Input is steady at 250VAC but the output then drops to 213VAC. This is my first UPS I have no experience with them so whats going on here and how do I fix it? 

 

I have two systems plugged into it: A Freenas server running a 3770k (600w corsair psu), and my main rig with a 3900x and GTX1080 (1000w corsair PSU). When running fur mark and CPU-Z stress I was still under 50% load on the UPS so I'm not overworking it. 

 

Thanks in advance for any advice!

[Bombastinator]

I’m not 100% sure it’s broken. Machines run fine using it powered yes?  If you pull the plug out of the wall do they continue to work?  If so it could just be the UPS going “dude.  You only actually need 213v to run this stuff”

[AbydosOne]

1 minute ago, Bombastinator said:

I’m not 100% sure it’s broken. Machines run fine using it powered yes?  If you pull the plug out of the wall do they continue to work?  If so it could just be the UPS going “dude.  You only actually need 213v to run this stuff”

That's not how UPSes work. The UPS has no idea what the voltage rating of the attached devices are. Undervolting can damage some power supplies.

 

37 minutes ago, jrga said:

Thanks in advance for any advice!

Did you buy it new? I would contact the manufacturer and see if they have a better idea.

 

Off-hand, it sounds like your input may have periodic spikes or dips that are causing it to need to correct the voltage. Alternatively, when running at very low loads, computer power supplies are a very reactive power sink, and can mess up systems expecting more resistive-like current loads. Do you have anything else connected to it?

[jrga]

11 minutes ago, AbydosOne said:

That's not how UPSes work. The UPS has no idea what the voltage rating of the attached devices are. Undervolting can damage some power supplies.

 

Did you buy it new? I would contact the manufacturer and see if they have a better idea.

 

Off-hand, it sounds like your input may have periodic spikes or dips that are causing it to need to correct the voltage. Alternatively, when running at very low loads, computer power supplies are a very reactive power sink, and can mess up systems expecting more resistive-like current loads. Do you have anything else connected to it?

Yep purchased new from Amazon, I have opened a ticket with the manufacturer to see what they say. I only have the two systems noted plugged into it nothing else, it could be it switches when one of the systems idles but I have had it happen whilst watching youtube and having Avid MC open so I doubt it. 

[jrga]

9 minutes ago, Caroline said:

can it output all 250 or does it work at a lower voltage? check the labels or manual to find out

if it can only output let's say 220 volts but your mains is 250 it treats that as an overvoltage and uses AVR to lower it

Manual states it can output between 170 - 280 VAC I assume thats with AVR though oops was mistaken thats input rating, says output is 230 VAC +/-10%, being that this is my first UPS im not super familiar with all the terms/specs so I cant really make assumptions about the specifications given so here is the manual: https://powerwalker.com/manuals/VI 650-2200 SH(L)/PowerWalker VI SH(L) - Quick Guide EN.pdf

[Bombastinator]

12 minutes ago, AbydosOne said:

That's not how UPSes work. The UPS has no idea what the voltage rating of the attached devices are. Undervolting can damage some power supplies.

 

Did you buy it new? I would contact the manufacturer and see if they have a better idea.

 

Off-hand, it sounds like your input may have periodic spikes or dips that are causing it to need to correct the voltage. Alternatively, when running at very low loads, computer power supplies are a very reactive power sink, and can mess up systems expecting more resistive-like current loads. Do you have anything else connected to it?

Of course it doesn’t know what they’re rated for, But it could measure the outgoing power over time.  There apparently was a power drop after a period of time.  It might not of course.  There’s a difference between it’s not working as intended (broken) and it’s not the right device/setting for the use case.  

[AbydosOne]

1 minute ago, Bombastinator said:

Of course it doesn’t know what they’re rated for, But it could measure the outgoing power over time.  There apparently was a power drop after a period of time.  It might not of course.  There’s a difference between it’s not working as intended (broken) and it’s not the right device/setting for the use case.  

Voltage is almost always held constant; current is what varies when a device requires more/less power. Again, no way for the UPS to know the necessary input voltage.

[Bombastinator]

15 minutes ago, AbydosOne said:

Voltage is almost always held constant; current is what varies when a device requires more/less power. Again, no way for the UPS to know the necessary input voltage.

re: voltage

then this avr mode he is talking about shouldn’t exist.  It seems to be there for voltage normalization.  It also seems to think 213v is what it should output.  It could be because it’s broken.  It could because it’s simply set to 213.  It could be something else.  It’s a weird number. Why it is 213 is the question. Hence not 100%. Looking at the next post the claim is output is 230 plus or minus 10%. So 207-253.  Still might not be broken.
 

Re: again

Again of course not. 

[jrga]

15 minutes ago, Bombastinator said:

re: voltage

then this avr mode he is talking about shouldn’t exist.  It seems to be there for voltage normalization.  It also seems to think 213v is what it should output.  It could be because it’s broken.  It could because it’s simply set to 213.  It could be something else.  It’s a weird number. Why it is 213 is the question. Hence not 100%. Looking at the next post the claim is output is 230 plus or minus 10%. So 207-253.  Still might not be broken.
 

Re: again

Again of course not. 

The System is very limited in terms of settings, the WinPower app it comes with only lets you see a crude diagram of the power flow and a % load figure. No temp readings or anything else. I don’t see how I could manually set an output voltage or anything so it seems to have just automatically decided that 213 is what it’s going to output and stick it on AVR with a very loud 40mm fan running all the time. It’s super annoying. I’ll wait to hear back from the manufacturer to see if this is intended behavior or not. If it is I’m returning the unit because my work requires a pretty low noise floor in my space. 

[AbydosOne]

20 minutes ago, Bombastinator said:

re: voltage

then this avr mode he is talking about shouldn’t exist.  It seems to be there for voltage normalization.  It also seems to think 213v is what it should output.  It could be because it’s broken.  It could because it’s simply set to 213.  It’s a weird number. Why it is 213 is the question. Hence not 100%. Looking at the next post the claim is output is 230 plus or minus 10%. So 207-253.  Still might not be broken.
 

Re: again

Again of course not. 

There will be voltage variation coming into the UPS because wires are not perfect conductors. As more power (read: current) is drawn, the resistance of the wires (due to V = IR) will drop the apparent voltage at the receiving end (V_nominal - V_{wire_loss} = V_apparent), which is what AVR is correcting for. As more current is drawn, input voltage may decrease, requiring AVR to kick in. Also, reactive (vs resistive) loads can push the voltage above the nominal, which AVR also can correct for. These effects also works at a greater scale in the power grid, and you can end up with slightly depressed (or even raised) voltages based on the loading to your part of the grid. AVR mode is quite useful, you see.

 

But, as I said, the UPS doesn't know what input voltage the connected devices need, and the UPS should be regulating to the nominal output voltage. The nominal output of the UPS is 230VAC (per the spec page). 213VAC output is weird, which is why I recommended contacting the manufacturer, but not out of spec, correct. Based on my understanding of PC power supplies, 213VAC is low for 230VAC, but not dangerous (the rating on my Corsair CX650M is 100-240VAC, but your milage may vary).

[Bombastinator]

5 minutes ago, AbydosOne said:

There will be voltage variation coming into the UPS because wires are not perfect conductors. As more power (read: current) is drawn, the resistance of the wires (due to V = IR) will drop the apparent voltage at the receiving end (V_nominal - V_{wire_loss} = V_apparent), which is what AVR is correcting for. As more current is drawn, input voltage may decrease, requiring AVR to kick in. Also, reactive (vs resistive) loads can push the voltage above the nominal, which AVR also can correct for. These effects also works at a greater scale in the power grid, and you can end up with slightly depressed (or even raised) voltages based on the loading to your part of the grid. AVR mode is quite useful, you see.

 

But, as I said, the UPS doesn't know what input voltage the connected devices need, and the UPS should be regulating to the nominal output voltage. The nominal output of the UPS is 230VAC (per the spec page). 213VAC output is weird, which is why I recommended contacting the manufacturer, but not out of spec, correct. Based on my understanding of PC power supplies, 213VAC is low for 230VAC, but not dangerous (the rating on my Corsair CX650M is 100-240VAC, but your milage may vary).

Re: doesn't know

You keep saying that, and I keep saying “yes and?..” 

 

re: not broken. 
I agree in that I don’t like 213 particularly.   That digital meter implies it might be simply setable, by single volt, but OP says no ability to do that.   It’s possible this thing is a lot more primitive than its digital meter implies. 

 

[AbydosOne]

10 minutes ago, Bombastinator said:

You keep saying that, and I keep saying “yes and?..” 

 

You're the one who suggested the UPS was adjusting the output voltage based on required input (somehow)...

1 hour ago, AbydosOne said:

“dude.  You only actually need 213v to run this stuff”

...which isn't true. It's a misleading interpretation of how electrical power works. At no point should the UPS be dropping the output voltage in response to decreased demand.

 

I'm trying to explain what the AVR is used for because you said...

42 minutes ago, Bombastinator said:

then this avr mode he is talking about shouldn’t exist

...and I've given the reasons for why it exists. (Though I should have originally specified "output voltage" is held constant; the input voltage variation is what the AVR is correcting for).

 

Sorry for any misunderstanding.

[jrga]

7 hours ago, Bombastinator said:

It’s possible this thing is a lot more primitive than its digital meter implies. 

 

Oh for sure its primitive (hence 2200VA for only £130), sure it has a display but only one button on the device: power. 

[Bombastinator]

11 hours ago, AbydosOne said:

 

You're the one who suggested the UPS was adjusting the output voltage based on required input (somehow)...

...which isn't true. It's a misleading interpretation of how electrical power works. At no point should the UPS be dropping the output voltage in response to decreased demand.

 

I'm trying to explain what the AVR is used for because you said...

...and I've given the reasons for why it exists. (Though I should have originally specified "output voltage" is held constant; the input voltage variation is what the AVR is correcting for).

 

Sorry for any misunderstanding.

Re: required input

it could be possible if the thing was measuring how much it was drawing and adjusted whatever metric it had to do so.  Which appeared from the digital readout to be voltage. It wouldn’t need to know what the requirements were, just what was being drawn.  Does assume more capability than it apparently has though. It’s apparently not happening. 
 

re: I’m trying to explain..

Yes.  You did that.  You missed the IF. 
 It’s one of the things a UPS is for. It’s like a surge protector that Gets the low side as well as the high side.  The issues seem to be that “not 100% sure” got taken as “not”, and a speculation as to a possible reason, got taken as “the” reason.
My fault for not being sufficiently vague  or not sufficiently specific that i was being vague I guess. Or something.

 I was reacting to a statement in the original post that contained an assumption that the device was assumed to be defective.  My statement was intended to be only that it only might be defective, and there were numerous reasons that could be.  I made a speculative example of one possible. It assumed a lot more capability than the thing apparently has.   What currently seems to be the case is that the thing does not have the capability to do such things and i was misled by the presence of a digital readout into thinking it could possible do a lot more than it seems it can.  The thing seems designed to get the PSU to take the slop rather than adjusting itself in more than a merely rough way.  I have a nasty suspicion that attempting to raise the output voltage to an actual 230w. Would require something like opening it and banging on something with a hammer, which would then only cause other problems.

 I badly over estimated the capacity of the device. At this point I almost want to put the thing on a power strip.  This 280 max is a lot lower than the multiple thousand watts a surge can produce.  With luck there a fuse in there that will go off it goes above 280 shutting it down and diverting to battery power. 

[jrga]

So spoken to the support and basically they said my input voltage is annoyingly between the threshold of when AVR kicks in. 
https://support.powerwalker.com/kb/faq.php?id=39

It sits stable at 250 but if it ever goes over 252 it will kick AVR on but won’t turn it off until it drops to 247 so it basically means it’s going to be very random when it kicks into gear. Their response was effectively “buy one of our higher end units”

[AbydosOne]

17 minutes ago, jrga said:

So spoken to the support and basically they said my input voltage is annoyingly between the threshold of when AVR kicks in. 
https://support.powerwalker.com/kb/faq.php?id=39

It sits stable at 250 but if it ever goes over 252 it will kick AVR on but won’t turn it off until it drops to 247 so it basically means it’s going to be very random when it kicks into gear. Their response was effectively “buy one of our higher end units”

oof, that a pretty bad place to be... though I give them credit for publishing the graph of activation voltages.

 

You'd probably want to look into other manufacturers/models and see if there's a better option. Running it in buck mode almost constantly will probably decrease its lifespan.

[jrga]

7 hours ago, AbydosOne said:

oof, that a pretty bad place to be... though I give them credit for publishing the graph of activation voltages.

 

You'd probably want to look into other manufacturers/models and see if there's a better option. Running it in buck mode almost constantly will probably decrease its lifespan.

weird thought, would connecting via an extension lead make a difference to the input voltage? 

[AbydosOne]

2 minutes ago, jrga said:

weird thought, would connecting via an extension lead make a difference to the input voltage? 

Theoretically, yes, but you'd need close to 400 meters to get down to 240VAC at idle load. It's not the sort of hack I would recommend (though it would be funny to see).