Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Problems with my UPS

Status
Not open for further replies.
Hi, i'm having some problems my UPS (Kanji 1200va lcd) with it activates, under high loads, it starts to make a CLANK CLANK noise. It's supposed to withstand 720W, and my PC reaches 80%(576W) in the display at maximum load. (yes, it high end stuff). I already RMAed it once because after 4 months it started doing the same noise but even at 30%. It uses 2 12V 7Ah sealed lead-acid batteries in series. I tryed uping the load little by little, from 30%, and when it reaches 70% it starts. If it activates while onder high load, it does that noise but very loud, and then the pc shuts down. (which defeats the purpose of an ups) (i still haven't opened it yet, because of the waranty) i think it might be the batteries, which as they are from factory, they might be made-in-china™ and because of that, can't withstand the load. (but i don't know about this things so i may be wrong) the page is www.kanjitech.com.ar (it's more like a phantom page, very cheap, and missing links, even the email doesnt work xD) ...i should had bought a brand one, APC or a known one, but this was quite cheap and wanted to try) any idea of what might be the problem? EDIT: i forgot, i have the 24' LCD (switch), and the router and modem (transformer) plugged to the ups also
 
Last edited:
what's it's "continuous load" rating? i think you'll find it's about half of what you think it is. sounds like you're looking at it's "peak load" or "transient load" rating. also remember that the power factor of your power supply might be very low, so the actual current drawn is so out of phase with the voltage that the actual VA is much higher than the "wattmeter" on the UPS is saying it is. this is a common problem with switchmode power supplies, is that their PF is somewhere between 0.3 and 0.5, because they present a primarily capacitive load to the line this tends to throw most "wattmeter" readings way off, because the "wattmeter" is assuming the voltage and current are in phase. cheap UPSs tend to have cheap "wattmeter" circuits, where better UPS designs will have a wattmeter that also measures PF and give a closer approximation of the true VA of the load.

to paraphrase Audioguru, you're probably reading a "WHATmeter"...
 
Last edited:
it starts to make a CLANK CLANK noise.
Sounds more like a diesel engine which has run out of lube oil than something electronic.:eek:

JimB
 
Sounds more like a diesel engine which has run out of lube oil than something electronic.:eek:

JimB
in fact, yes, now that i think of it, it really sounds like a diesel engine xD

what's it's "continuous load" rating? i think you'll find it's about half of what you think it is. sounds like you're looking at it's "peak load" or "transient load" rating. also remember that the power factor of your power supply might be very low, so the actual current drawn is so out of phase with the voltage that the actual VA is much higher than the "wattmeter" on the UPS is saying it is. this is a common problem with switchmode power supplies, is that their PF is somewhere between 0.3 and 0.5, because they present a primarily capacitive load to the line this tends to throw most "wattmeter" readings way off, because the "wattmeter" is assuming the voltage and current are in phase. cheap UPSs tend to have cheap "wattmeter" circuits, where better UPS designs will have a wattmeter that also measures PF and give a closer approximation of the true VA of the load.

to paraphrase Audioguru, you're probably reading a "WHATmeter"...

(f**, firefox died, and my reply got deleted before i submited it x.x)
summarizing, my supply has really good line regulation (below 3% for each voltage line), and the PF is always above 0.8 (active PFC). the LCD has a switchsupply inside but it's roughtly ~80W, so doing some heuristics, yes, i'm probably overloading it :-(, but still i'm not sure, i'm not even using the 60% of my supply capasity, and at 0.8, 700W would be 875VA....by 0.6 = 525VA. the lcd lets say 0.4 then 80W would be 200VA, then the modem and the router transformers are 4W, at 0.1(random, don't really know how much) would be 40VA x 2 = 80VA
525 + 200 + 80 = 805VA which is still below the 1200VA rating
even with the supply at 100% (whch is far from being real) everything would sum 1155VA, still below, at it's limit bstill below.

so then, any idea of could that sound be?, something is breaking or is it just an alarm, or just plain cheap ups that can't withstand what it says?
 
Last edited:
Sounds like your UPS is DOWN :D

To open it and change the batteries should not void the warranty, lets face it batteries dont last for ever.

One thing i found with ups is the heat generated inside the enclosure drys the batteries out, and i have taken to extending the battery leads and putting the batteries outside the ups, this gives 5 times the life from the batteries.

The clunk, clunk is more than likely the transformer energizing and switching off then energizing again etc. This normally results in killing the mosfets after a few cycles of trying to boot up.

This could be due to overloading the ups or the batteries not being able to supply enough current for the initial start up surge. I would guess the later.

If you put a meter on the batteries and monitor the voltage on start up you should see what is going on, and if they take a big dive in voltage below the cutout point than the batteries are shagged.

Pete.
 
now that u mention it, on thing that happened the first time was that the charge meter of the batteries went down to 0-26% from 80% went the sount started.
Maybe it is the batteries, if the charge it calculated by the voltage. after the waranty i could mod it to plug some bigger deep cycle bateries from the outside like the profeesional UPSs
 
how many times have you run the batteries down and not recharged them right away? a UPS is not made to be used like a offline inverter. it's only supposed to provide enough power for you to shut the PC down safely. usually they're supposed to be able to run a few minutes at full load (10 or 15 minutes i think) which should be long enough to shut everything down safely. they're not meant to be run for extended periods. if you have been trying to use it as an offline inverter for extended periods, you may have drawn the batteries down often enough and long enough that they have become sulphated. they may hold a charge, but they can't supply anywhere near the current that they used to. the other thing to check is the connections to the batteries. any loose connections there will severely limit the amount of current that can be drawn from the batteries. spade lugs are notorious for getting loose as the spring tension in them relaxes with heating-cooling cycles when current is run through them. if the batteries are connected with ring lugs and bolts, you should feel no play in the lugs, bolts, and nuts if you wiggle them. the same goes for the battery cable connections to the boards.

also you mentioned the batteries are 2 12V 7AH batteries.... it doesn't sound like what should be in a 750W UPS (and no, i won't even think about the model number actually having anything to do with the wattage in this case, as it sounds too good to be true). many 500 and 600VA UPSs use 9 and 10AH batteries
 
Last edited:
If you are going to run external batteries make sure they are not live with mains voltage. I have a UPS that i was going to fit external batteries to but it wasnt isolated .
 
that would definitely be something to watch out for. better UPS supplies use a small transformer for the charger. ceaper ones "cheat" and use the main transformer to power the charging circuit. and some use a direct offline rectifier and DC-DC converter which is not isolated at all
 
yes, some times (about 4) when there was a blackout, i stayed about 20 minutes, on batteries, waiting for it to return. (yes that were 20 minutes powering the cpu and the lcd). Maybe more times because my machine runs 24/7 as a server.
The user manual says that every 3 or 4 month one should check how much the ups can hold on batteries to check the their health, that once they reach about 80% of the initial time, they should be changed.
btw, the manual says the it uses 7Ah/8Ah batteries, but i'm not sure if it means it can use both or if they are for a difrent model (it covers about 4 models, made in china style)
if u say that this is too much for some 7Ah batteries then they must be the starting type. Wikipedia says that UPSs use deep cycle ones(suppose it's talking about the brand ones)

about the ups using the main to charge the batteries. It's a standby ups, and u can hear the inverter when it's working on batteries, so i'm pretty sure that when u shut it down it doesn't have energy at it's terminals.
What type and specs of batteries should i look for to do the mod? so it can hold for about 20 minutes and not die after 4 discharges
 
you might want to use external liquid lead acid batteries if you're intending to use the UPS as an offline inverter. you may also want to think about modifying the inverter's cooling system to handle the extended operation. UPSs are only intended for 5-15 minutes of operation, as i said, in order to safely power everything down. any longer than that is "offline inverter" usage, and two problems present themselves, battery capacity and cooling capacity, both of which are inadequate for the your application.
 
Test ur batteries before assuming they are the problem...Remove them from UPS and attempt to recharge with good regulated 12 volt SLA charger. Remove charger and measure OCV (open circuit voltage) after a 1 to 2 hour rest period. Allow to sit for 2 days unused and remeasure OCV to see if batteries have a high internal discharge. If OCV is still good, continue with load testing.
You will need a low value high wattage resistor to do this quickly or extend the discharge time using a higher value lower wattage resistor. Find a discharge curve for ur batteries. I looked at curve for Yuasa NP7-12. Its rated for around 30 minutes at 7 amps OR 20 hours at 0.35 amps. If u can get a 2 ohm 200 watt resistor that should work for quick discharge OR get a 30 to 35 ohm 15 watt resistor for 20 hour discharge. Recharge immediately after the test if batteries appear to handle load test.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top