APC, overpriced cheap products.

sblantipodi

sblantipodi

2[H]4U
2FA
Joined
Aug 29, 2010
Messages
3,805
I have two BR1500GI from APC both rated as 1500VA/865W.
I tested this units with similar rigs, one with a 5930K and another one with a 5820K, both with GTX980 Ti SLI, one with Corsair AX860i and another with an Enermax Platimax 850W.

Both configurations runs at 800W when running Far Cry Primal Benchmark.
I see 800W from the UPS LCD and from a Watt Meter that I have between the peripherals and the UPS. So the measurement is quite good.

With blender rendering the load is at about 750W and the UPS handle the load without problem, if I stop the AC the ups switch on battery and I have the time to poweroff the PC.

With Far Cry Primal benchmark at 800W both BR1500GI shutdown with F02 error with long beep. I need to power off and power on the UPS to stop the beep.
This on both Enermax Platimax 850W and Corsair AX860i.

APC home unit are cheap overpriced crappy products.
 
nope. they do have pretty good stuff, and next to CyberPower (ups) one of the only brands i would consider getting.

the problem is, most UPS are not producing a "real" sinus signal, which is needed for (pc) psu's.

try the CyberPower 1500PFC or GX1325 and you will see.
 
So you main issue is that your 865w rated UPS can't handle a continuous 800w load? Looks like you cut the overhead really close and got burned. I'll keep using APC but thanks for the laugh.
 
still, 800 isnt 850/865 and thats what the ups is rated for, so it needs to do that without trouble.

independent from that, i still would swap them out for cyberpower (real) sinus wave ups,
or you will throw out the PSU's pretty soon.
 
Last edited:
Ocellaris said:
So you main issue is that your 865w rated UPS can't handle a continuous 800w load? Looks like you cut the overhead really close and got burned. I'll keep using APC but thanks for the laugh.
Click to expand...

Continuos is not the right word.
I would be happy to have 1 minute or less of time to shut down my PC if an outage occurs during load (800W).
On offical specs is written that this things should handle up to 865W for four minutes.

By BR1500GI can't handle 800W neither for one seconds, it simply shut down instantly if the load is near 800W.
 
Fry178 said:
or you will throw out the PSU's pretty soon.
Click to expand...

inaccuracies or popular beliefs, call it how you want.
simulated sine wave is good for every decent PSU.

I'm using simulated sine wave on Active PFC PSU since the first years of the PFC and I never got a problem or seen a PSU dying.
 
sblantipodi said:
inaccuracies or popular beliefs, call it how you want.
simulated sine wave is good for every decent PSU.

I'm using simulated sine wave on Active PFC PSU since the first years of the PFC and I never got a problem or seen a PSU dying.
Click to expand...

Yeah, same here... "stepped approximation" (APC's terms) or "Simulated Sinewave" (Cyberpower's terms) are both fine with modern PSUs. My desktop machine and fileserver box have both been through quite a few power outages and they're still fine today. (PSU in the fileserver box has to be going on 5-6 years old now)
 
Last time I bought an APC 2200VA unit for the server room where I work, it took them 3 replacements to get me a fully functioning one.

Original purchase: We had a power outage a couple months after receiving it, and the one outlet never came back up after we got power back.

I did full tests on each replacement before putting them into service - every one of these had their QA pass sheet attached to them.

1st replacement: One whole bank would not power on unless it was done manually

2nd replacement: The cooling fan was faulty and very very loud

3rd replacement: It actually worked and is still in service.

Never buying APC again. Too much hassle and not reliable enough.

On the other hand, I have not had a single issue with ANY of the CyberPower units we have here. I have one 2200VA and a bunch of smaller ones that are used for various desktops/laptops that have to stay powered up in case of an outage.
 
very very disappointed from this units. it costs three times more than a cyberpower and they are even worse.
 
@
sblantipodi
are you talking about running the psu on AC power (thru UPS), or the output signal when the battery provides the power?
2 different things.
and, i was referring to the CP with "real" sinus wave, not the ones running simulated (then u can just get the APC).

not saying they wont work, but i remember seeing tests that showed almost all of the psu's (aPFC) dying even before the warranty was over, when running on battery.

i had a 750w TX that died after 3y connected to a APC UPS, while experiencing about 10-30 brown/black outs per month.
my replacement worked for 5y without any issues connected to a UPS with real sinus signal, and the seasonic i got after that, is now running 3y without problems with about 20-40 brown/blackouts per month.
not sure how that can be popular belief...
 
Last edited:
Fry178 said:
@
sblantipodi
are you talking about running the psu on AC power (thru UPS), or the output signal when the battery provides the power?
2 different things.
and, i was referring to the CP with "real" sinus wave, not the ones running simulated (then u can just get the APC).

not saying they wont work, but i remember seeing tests that showed almost all of the psu's (aPFC) dying even before the warranty was over, when running on battery.

i had a 750w TX that died after 3y connected to a APC UPS, while experiencing about 10-30 brown/black outs per month.
my replacement worked for 5y without any issues connected to a UPS with real sinus signal, and the seasonic i got after that, is now running 3y without problems with about 20-40 brown/blackouts per month.
not sure how that can be popular belief...
Click to expand...

popular beliefs, say that your PSU died for a simulated sine wave is like saying that christ died for cold.
 
sblantipodi said:
popular beliefs, say that your PSU died for a simulated sine wave is like saying that christ died for cold.
Click to expand...

I will not comment on PSUs with Active PFC literally dying when fed a stepped wave, simply because luckily I haven't had one die yet. If power goes out more more than 5-10 seconds, I have clients set to shutdown ASAP and people trained to shutdown the UPS units as well.

The failures I've had were always 100% failed bad caps on the PSU's secondary (happened to about 10% in the course of 7 years).
As far as UPS devices go, I've had slow self-cooking of a few due to power outages in the night.
The UPS units (Line Interactive AVR single 12V VRLA 500W real power) happily kept the sleeping monitor and the motherboards going for several hours during a weekend long outage.
Noone killed me for the beeping.

But I have seen an ability to generate rapid smooth sine waves both on AC (mains passthrough mode most likely) AND on battery power as listed as a requirement for warranty service on one device or two.
Could have been a Corsair PSU FAQ or the FAQ of an UPS company, but that latter would probably wreck the point I'm trying to make :)

Either way the beef I personally have with 'stepped wave' on battery is that you're messing with the assumptions the PSU's designer was working with whilst designing.
It's a stretch to call it a simulated sine wave when it's far from it still.

With a sudden (fast) step up of an amplitude, the charging of capacitors will in some cases also happen more rapidly. Please, why won't anyone think of the little relays.
Also did the designer check the cap's specs for surges this big?

My worst personal experience with a stepped wave PSU was with audio actually.
Some mic/mixer/recorder/monitor combos wouldn't play nice on battery until I connected a 'pure sine wave' UPS.
 
I'm not technician, so i'll just tell my experience as a customer. I use 2 of APC's SMART line of UPSs, the SMT 1000 (http://www.amazon.com/Smart-UPS-SMT...UTF8&qid=1458317071&sr=1-7&keywords=APC+SMART) & the SMT 750 ( http://www.amazon.com/APC-Smart-UPS...UTF8&qid=1458317071&sr=1-5&keywords=APC+SMART ) for several months, and i'm extremely pleased with the purchace. I used some other brands in the past, and i had several minor - but still frustrating:rage:-, issues with my PCs. After the UPS's replacement these issues never appeared again, so there is a very good chance that they were UPS related.
The bottom line is that i've learned my lesson well enough: I'll never again cheap out on such a critical component like the UPS is. APC's SMART line might be an expensive line, but all the issues that i faced in the past by using cheaper units never appeared again!!:happy:
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
APC's SMART line might be an expensive line
Click to expand...

Refurbished SMART units can be had for less than $200 shipped with brand new batteries so its not like they are that expensive if you want a server quality UPS for home usage. And they come with 2 times the battery capacity of the home units.
 
Last edited:
@sblantipodi
didnt know my personal experience is popular belief.
having all my psu's (running on stepped) die BEFORE the warranty was over (incl some at work) , but not after switching to ups with real sinus output, is enough for me NOT to buy stepped ones again.

michalrz
i normally dont run it for more than a min or two when there is a blackout (while surfing/streaming etc),
but i wont lose a ranked game to extend psu life ;-)
 
Sorry for the late reply but the explanation is simple.
The inrush current of the psu trips the overload.
The capacitors in the psu are depleted holding during the switch over(psu hold time is important here) so the current is much higher for a fraction of a second charging them. This will not show up on cheap watt meters as their sample time is too slow.
At 800W it does not have the reserve to meet the inrush current.
I always keep my ups load at 75% max of its rating and prefer 50%.
 
stormy1 said:
Sorry for the late reply but the explanation is simple.
The inrush current of the psu trips the overload.
The capacitors in the psu are depleted holding during the switch over(psu hold time is important here) so the current is much higher for a fraction of a second charging them. This will not show up on cheap watt meters as their sample time is too slow.
At 800W it does not have the reserve to meet the inrush current.
I always keep my ups load at 75% max of its rating and prefer 50%.
Click to expand...

a product that cost 5 times a normal UPS should be over engineered and if it shows 865W on the specifications should handle a good quality PSU like the Corsair AX860i at 800W of load without a problem.
if this is not the case I see no point in sending a premium for a product that does not meet even the advertied specification.
 
In AC power there is this thing called Power Factor. It causes additional current without a change in real power (watts). Energy Star rating requires PSUs to have a power factor of at least 0.90, but non-Energy Star models have been tested with power factors as low as 0.55. This only applies to alternating current, since it has a leading/lagging component based on the inductance/capacitance of the load that DC does not have.

Without a unity (1.0) power factor, a UPS (any brand, really, and applies to generators in general) not hit a Watts limit, but will hit a current limit, and force it into overload. The farther away from 1.0 you are, the lower the watts will be for a given current (or the higher the current will be for a given wattage).

Here's a decent intro without getting too physicsy

Power Factor Correction FAQ - AnandTech Forums
 
Brian_B said:
In AC power there is this thing called Power Factor. It causes additional current without a change in real power (watts). Energy Star rating requires PSUs to have a power factor of at least 0.90, but non-Energy Star models have been tested with power factors as low as 0.55. This only applies to alternating current, since it has a leading/lagging component based on the inductance/capacitance of the load that DC does not have.

Without a unity (1.0) power factor, a UPS (any brand, really, and applies to generators in general) not hit a Watts limit, but will hit a current limit, and force it into overload. The farther away from 1.0 you are, the lower the watts will be for a given current (or the higher the current will be for a given wattage).

Here's a decent intro without getting too physicsy

Power Factor Correction FAQ - AnandTech Forums
Click to expand...

a UPS of this price should simply work good and it doesn't.
it is better a cheap cyberpower that this extremely expensive APC.
APC products are crap.
 
I believe this behavior is actually a function if the APC unit being a better product meaning it protects itself from overload where a cheaper unit lacks such protection or may not protect itself as well. Although to the user the better protection seems like a misfeature compared to the cheaper UPS. Well that is until the cheaper UPS self destructs because it is running its components over their rated values.
 
Last edited:
drescherjm said:
I believe this behavior is actually a function if the APC unit being a better product meaning it protects itself from overload where a cheaper unit lacks such protection or may not protect itself as well. Although to the user the better protection seems like a misfeature compared to the cheaper UPS. Well that is until the cheaper UPS self destructs because it is running its components over their rated values.
Click to expand...

if this is the case the UPS should write this on the display: "I'm so cool and smart that I protect my self and I shut down all the connected peripherals"
in real life my APC write this on the display: "I'm so dumb that I write battery output short"

Battery output short with F02 error is no way a smart autoprotection feature.
 
They have a teardown of the CyberPower 1500VA on TH. I'm still carrying beef about TH though, so I won't link to it.
 
sblantipodi said:
if this is the case the UPS should write this on the display: "I'm so cool and smart that I protect my self and I shut down all the connected peripherals"
in real life my APC write this on the display: "I'm so dumb that I write battery output short"

Battery output short with F02 error is no way a smart autoprotection feature.
Click to expand...

Or, hey, the battery could actually have a short, and catch on fire or explode. Yeah, that's pretty dumb protection there...

I'm not an APC fanboy or anything, but wow. You seem totally oblivious to the real problems, and just keep blaming the unit because it won't melt itself down when you try to do stupid things to it.

You could always just make your own UPS with a 12V DC battery, charger, and an automobile inverter, and forgo all safety protection, if that's what you want.
 
LightningCrash said:
They have a teardown of the CyberPower 1500VA on TH. I'm still carrying beef about TH though, so I won't link to it.
Click to expand...

What is TH?
can you link me this comparison in PVT please?

Brian_B said:
Or, hey, the battery could actually have a short, and catch on fire or explode. Yeah, that's pretty dumb protection there...

I'm not an APC fanboy or anything, but wow. You seem totally oblivious to the real problems, and just keep blaming the unit because it won't melt itself down when you try to do stupid things to it.

You could always just make your own UPS with a 12V DC battery, charger, and an automobile inverter, and forgo all safety protection, if that's what you want.
Click to expand...

I don't get your point, I paied an eye for a product that is advertised to support up to 865W and at 800W it power down?
do you call it a feature?
 
sblantipodi said:
What is TH?
can you link me this comparison in PVT please?
Click to expand...

Toms Hardware. On 9/11 a lot of websites were down. You couldn't get to CNN or news sites because of the crush of traffic. Since a lot of people could still get to HardOCP, Kyle was posting 9/11 news information on [H]. Tom acted like a douche about it on TH.
 
I'm not actually a big PURCH fan these days in general.

I never was a big Tom's fan even before a few acquisitions these days.

TH stopped being real probably 25 years ago.
 
Yeah, I was more worried about the battery age. I have a Back ups from 1994 that gets new batteries every 4 years. Its super solid. It would be a shame if their quality has been falling over the years. I still trust SmartUPS to my servers. It would suck if they started failing.
 
Jorona said:
Yeah, I was more worried about the battery age. I have a Back ups from 1994 that gets new batteries every 4 years. Its super solid. It would be a shame if their quality has been falling over the years. I still trust SmartUPS to my servers. It would suck if they started failing.
Click to expand...

my batteries are fresh new.
 
just another F02 during a blackout. just to remember to people to not waste money on APC products.
they seems premium products but they aren't.
 
sblantipodi said:
just another F02 during a blackout. just to remember to people to not waste money on APC products.
they seems premium products but they aren't.
Click to expand...
remove the batteries and measure the voltage of each battery when not connected to anything including each other and post the numbers here.
Sounds more like junk batteries than a bad ups.
 
I have had good luck with Cyberpower UPS. Went to one when I switched from XP to Win 7 since Win 7 dropped native serial port support and that is what my 90's era BackUps had. Newer APC units didn't play well with several Active Power Factor computer PS at the time. Oddly, the old BackUps worked just fine with APF PS. Neither brand has suffered a smoke liberating event.

In my experience, APC units are picky about 3rd party replacement batteries.
 
You must log in or register to reply here.
Back
Top