Low voltage when UPS is in battery mode.

Status
Not open for further replies.

Rajiv_6

Commendable
Nov 28, 2016
14
0
1,510
Hi,

I found same strange voltage readings (using multi-meter) on APC BX600C‑IN 600VA, 230V Back UPS.

The INPUT voltage during this test was =~ 250 volts.


1) No Load Test:

Main Mode (UPS is not in battery mode) : The output voltage is in range= 247-250 volts.
Battery Mode: The output voltage is in the range = 142-150 volts.

And transferring the UPS from battery to mains (by switching on the power supply) gives a sudden voltage spike of 411 volts and even 500 volts sometimes.

2) On Load Test:

Main Mode: Output voltage range is 245-248 volts.
Battery Mode: The output voltage is in the range = 167-170 volts.

And transferring the UPS from battery to mains (by switching on the power supply) does not show any voltage spike.

Please suggest:

Are low voltage during battery modes normal? and what does the VOLTAGE SPIKE on no load suggest? (I guess i need to measure current in this case).

Is not this spike very dangerous when somebody leaves Mains, UPS and PSU ON (but PC OFF) and the UPS goes from mains->battery->mains; sending spike to primary CAP of PSU and damaging it.

Thanks.



 


What does output 230v nominal means 5% or 6% ... of input voltage?
Yep, battery is good (=~ 15 minutes of backup). Generally UPS manufacturers mention output as % of input voltage, say +/-5% of 250 volts i.e. the output voltage range of the device will be 237-263 volts ; but the APC specs sheet for the model does not mention anything about OUTPUT voltage range.
 


Yep, I contacted them and it was very horrible experience. The issue was escalated to branch manger as site engineers were not punctual, responded erratically and angrily.

They visited on the fourth day and changed the PCB of the UPS. Later when my PC started to reboot I contacted them again they told me they have already done what was to be done and suggested using UPS on TV to verify.

They are also suggesting to change my PSU (Corsair VS 450...1 year 3 months old....still under warranty) . Instead of checking UPS on TV; I thought of using Multi meter......and now I am afraid that "UPS from battery to mains gives a sudden voltage spike of 411 volts and even 500 volts sometimes on no load".

The specs sheet at http://www.apc.com/salestools/VGUA-9WHCL9/VGUA-9WHCL9_R0_EN.pdf does not mention exact nominal voltage percentage.

Thanks for your quick reply.

 


Yep, It does mention 230 V TYPICAL (but the question still remains the same: What is the TYPICAL IN "PERCENTAGE"?). And moreover we MUST ALSO CONSIDER THE FACT THAT IT IS THE MODIFIED [APPROX.] SINE WAVE.

Can this be the reason WHY MULTI METER IS ONLY SHOWING 167-170 VOLTS?

The irony is that the most of the company's product sheet is always misleading. Did you look at TRANSFER TIME i.e. 4-8 ms. What does this mean? Sometimes it will take 4 ms and sometimes it can take even up to 8 ms??

Moreover earlier I had Intex 725 protector http://www.intex.in/it-accessories/ups/ups-protector-725-2 and it also showed less voltage on battery mode (145-160 volts).

And according to INTEX no Backup UPS in the world can show OUTPUT close to 230 volts on BATTERY MODE. (This is not true for online or other efficient UPS with CVT).
 
It only affects DVMs that are not true RMS meters: http://www.apc.com/us/en/faqs/FA157483/ If you are using an inexpensive DVM, then your readings are normal. A better way is to use an oscilloscope as that displays the voltage and the waveform.

Depending on how quickly input voltage drops, it can take up to 8 ms for the UPS to transfer to battery mode. A 4-8ms transfer time is excellent and no system will shutdown if switching occurs that quickly; a PSU that meets the ATX specs should be able to work properly when the outage is 16ms or less.

What's a UPS with CVT? A converter used in India? There is a difference between a real UPS and a converter (a product that isn't sold in most countries).
 


Thanks GhislainG for clarifying the "LOW VOLTAGE ON BATTERY MODE" . Actually APC is big name and I switched from Intex to APC for better protection; at least they are honest while mentioning TRANSFER TIME IN THE RANGE OF 4-8 ms.

I don't have a Oscilloscope. But I can access one at my friend's LAB.

I bought multi-meter by checking reviews on amazon : http://www.amazon.in/Mastech-MAS830L-Digital-Pocket-Multimeter/dp/B0093H048Y/ref=sr_1_1?ie=UTF8&qid=1499961021&sr=8-1&keywords=multimeter

I guess it should not be used to check stepped sine wave , Can I measure current flow using http://www.amazon.in/Mastech-Digital-Voltage-Current-Resistance/dp/B00PNIHCL4/ref=sr_1_2?ie=UTF8&qid=1499961021&sr=8-2&keywords=multimeter
Do these device give reliable readings ? I mean there is no need to CUT THE CABLE to check the current in series using 2nd device.

I guess I need to shift my attention from UPS to PSU as "Corsair VS 450" belongs to TIER 4 list at

http://www.tomshardware.co.uk/forum/id-2547993/psu-tier-list.html

CVT stands for "Constant Voltage Transformer" I was thinking of using CVT in series after UPS (UPS powering CVT), but its not possible because 600 VA * O.6 (PF OF UPS) * O.8 (efficiency) = ~ 288 watts < 450 watts (> 350 watt approx power is required by my PC).

At least I would have got PURE SINE WAVE (better for PSU with Active PFC) in both mains and battery mode.

Online UPS are very expensive > 20,000 INR in India and this solution (assuming 1200va UPS and CVT ) would have cost Rs 11000 approx.
Thanks.



 
A True RMS meter costs a lot more: http://www.amazon.in/Fluke-115-True-RMS-Multimeter/dp/B00K17CMKS/ref=lp_6395977031_1_19?s=industrial&ie=UTF8&qid=1499965293&sr=1-19 This seems to be the least expensive on amazon.in.

The meter that you linked isn't True RMS; therefore it can't be accurate.

What's wrong with your PSU? Do you have a high-end and very expensive system? A tier 4 PSU isn't necessarily a bad unit for an office system, but it isn't recommended if you have a powerful GPU. I ran an office system 24x7 using a VS550 and I never had issues with it.

You should use a UPS or a CVT, not both in series.

I know that a pure sinewave UPS is expensive; that's why I usually buy them used for a fraction of the original cost.
 


Okay, the main problem is that my PC reboots when there is very SLIGHT FLUCTUATION (only visible on my room's tubelight) and I can't hear any UPS TRANSFER sound. and twice PC rebooted during A COMPLETE BLACK OUT (I was working on my PC & it rebooted before I could hear any UPS TRANSFER SOUND).

5 times this has happened when I was near my PC (3 times on slight fluctuation and 2 times during BLACKOUT) and many times when I was not near it.

The problem is still present after PCB was changed by APC.


Sad but true this may be a PRANK by someone (but I can't figure it out.. WHAT is being done TO CAUSE THIS TYPICAL FLUCTUATION - ...harmonic distortion....frequency is being changed.... but how?).

We are residing in a Tower with 48 flats and POWER LINES, CHANGEOVER IS ACCESSIBLE TO ANYONE (There is Change Over outside my FLAT which changes when mains supply is off and BACKUP GENSET is ON and Vice-Versa).

The only good news is that there were no reboots today.

Windows 7 (UPDATED TILL TODAY) and event Manager shows KERNEL POWER ERROR 41(63).

Processor: Intel Pentium G3220
RAM: Corsair Vengeance 4x2=8 GB DDR3
HDD: 1 TB Seagate Barracuda 7200 RPM SATA 2
Optical Drive: Asus Blu Ray writer.
Graphics: MSI GTX 750 Ti
Monitor AOC 15 inch LED (uses 2 USB 3.0 for POWER)
Keyboard and Mouse: 2 USB 2.0
Network Adapter: TP-link TL-WN722N USB 2.0
Motherboard: Gigabyte-H81M-S1
Cabinet: Cool Master 311 Elite
PSU: Corsair VS 450

Plz Note: PSU & APC UPS were bought on the same day i.e. 14 april 2016 and the problem started on 25 june 2017.

Thanks in advance.
-----------------------------------------------------------------------------
 


Thanks for devoting your time GhislainG.

Please suggest one. (Someone in toms hardware forum suggest to go for : channel well technology i.e. seasonic G series V2 ...good holding time and better for modulated sine wave i.e. when UPS is on battery)

Earlier I forgot to mention 2 important things:

a) I am still confused about Power Loading (@ BIOS which enables or disables Dummy Load. When the power supply is low load a self protection will activate causing it to shutdown or fail. If this occurs please set to enabled).

I have 3 options here : auto (currently selected), enabled (previously selected) and disabled. What will be the best option?

b) Main AC socket which connects UPS to the AC source does not have a valid earthing (problem with the whole building), whereas OUTPUT from UPS to PSU shows a valid earth point.

Thanks once again.
 
CWT and Seasonic are very different manufacturers. You can't presume that a quality Seasonic has a very good hold-up time because that isn't always the case. Same with CWT and other manufacturers. Since hold-up time is very important in your environment, check reviews like http://www.tomshardware.com/reviews/fsp-twins-500w-redundant-psu,4864-5.html (if you check the hold-up time graphs, you'll notice that a Corsair RM650x is very good). So is the Corsair SF450: http://www.tomshardware.com/reviews/corsair-sf450-power-supply,4512-4.html

a) Does your PSU shutdown at low loads? If not, then that setting in the BIOS should have no effect and I would leave it on auto. If I understand your problem, the system sometimes reboots when the voltage drops just not enough for the UPS to take over; it doesn't occur when the UPS takes over.

b) That certainly doesn't help, but there's nothing that you can do about it (unless a grounded water pipe is close by).

Honestly the best solution is an online pure sine wave UPS, but a PSU with a longer hold-up time may help. Even though you don't need it, I would consider a quality 550W PSU like
 
Ok, that sums it up.
1) Checking UPS thoroughly by oscilloscope and fluke rms multi-meter.
2) Corsair SF-450 W seems to be a complete peace of mind at Rs 9000 approx with 7 years warranty.

Thanks GhislainG.
 
Status
Not open for further replies.