Checkng Smart UPS Inverter funcion

[smallfreak]

I have acquired a cheap Smart UPS 700 that has been sold because "batteries are old". Sure enough it moans about replacing battery.

It took about 2 hours to load the batteries. At least the BCHARGE status went from 10% to 100% in that time. Just to check general function I placed a small load on it (~8%) and removed the mains cord. The device made some clicking for about a second and then it was off, indicating "battery fault". Voltage on the battery clamps dropped to about 21.6V for a short time. It does operate in online/filter mode.

Just to be on the safe side I want to check whether the inverter IS OK and it's really the battery that is faulty. Is it OK to clamp a 27.6V power supply to the UPS instead of the battery or additionally? At least that's the voltage of the pack after getting "fully charged".

Do I risk to blow the bench power supply due some current backflow? This one is no premium gadget too, but one of the cheap Chinese 300W 30V parts.

[Vovk_Z]

Yes, there is a risk to blow a regulated power supply because of backflow (if it is not designed to work as a charger too). But it became safe to add a series diode (I use 5-10-20 A Shottky diode). So you'll not have backfeed from the UPS charger.

The only typical problem with most (APC) UPSs - 22 uF electrolytes became dry instead of being wet (and some larger ones too). After that the charger floating mode voltage rises up and batteries die fast. So, if UPS is 10 or more years old I would check all electrolytes and possibly most of them need recap.

[Vovk_Z]

You may check batteries (and invertor) without a load too. The UPS itself eats about 1-2 Amperes DC from batteries.

[smallfreak]

Quote from: Vovk_Z on Yesterday at 04:42:13 pm

You may check batteries (and invertor) without a load too. The UPS itself eats about 1-2 Amperes DC from batteries.

Unfortunately, no as the UPS does shut off within two seconds even without load. This may be because of a dead battery OR because of a bad inverter.


Quote from: Vovk_Z on Yesterday at 04:29:48 pm

The only typical problem with most (APC) UPSs - 22 uF electrolytes became dry instead of being wet (and some larger ones too). After that the charger floating mode voltage rises up and batteries die fast. So, if UPS is 10 or more years old I would check all electrolytes and possibly most of them need recap.

So when the charging voltage ends at 27.6V then for each of the two batteries that's just the recommended 13.8V for "regulated power/standby" use - as is noted on the batteries. The capacitors might not be a problem for now.

I checked with the support line for my power supply, but the answer was rather inconclusive  :

Under normal use, the charging voltage should not be higher than the rated charging voltage of the battery is no problem.

A quick check with an 8R power resistor told me that each of the (single) batteries is able to deliver 1.2A which is quite low and the voltage drops from 12.6 just below 10V and slowly falling even with such a moderate load. I captured a voltage profile on the combines 24V battery connector during the "APC self-test" without any load.



This pattern is done twice before the UPS declares the battery bad.

Dos anyone know whether this big drop is to be expected?

[Vovk_Z]

All syptoms are pretty simple and clear I guess. You just have dead batteries, as UPS say. Charging voltage is fine too, so UPS (now) is fine.
Those old dead batteries may be useful if you need to test other UPS or else until they are totally dead. I mean, they are useful because they can't deliver large current so they are 'more safe' then new ones for testing something.

[smallfreak]

Ah, yes. You are perfektly right.

Isn't the "Smart" UPS' main feature that it is an online-UPS? That would mean, that the output is ALWAYS produced by the inverter?

So if the inverter is bad it either goes into 'bypass' or the output is off, even when mains is OK. Both does not happen.

I'll get a new set of batteries and fine.

But I learned some interestin new stuff about batteries and UPS in the last days. 🥴

[Vovk_Z]

Isn't the "Smart" UPS' main feature that it is an online-UPS?

No. There are different series Smart-UPS. Those cheaper are classic line-interactive type. I mean SmartUPS with rated power >=1000 VA are of two types: line-interactive and online.
Smart UPS 700 has to be only line interactive, as I know.
Back-UPS series is off-line type. (I may be wrong).

[smallfreak]

No. There are different series Smart-UPS. Those cheaper are classic line-interactive type.

I could check an old data sheet and it indeed notes "line-interactive" for that model and that the "inverter is permanently connected" and used both ways. So I guess, in case of a bad inverter the battery would not charge either.

Eventually an opinion about the replacement batteries might help.

What I have read about batteries for UPS use (not the solar-technology type), there seems no clear preference over the different technologies used. AGM batteries in average seem to be more expensive compared to Gel batteries but seem to offer no significant advantage when you expect a rather short power drop a few times and maybe a full cycle each year with a moderate (~25%) load.

But the topics I found were not the newest ones. Does this still hold? Modern systems most likely use modern cells like LiFePo instead of Lead, that may account for the lack of recent discussions about Lead-Batteries.

It doesn't make a fortune for just a small UPS, but choosing an "RBC5 battery pack" (= 2x 7.2 Ah) between 32€ (Q-Batteries, AGM?) and a 180€ (Original Schneider) with seemingly no obvious differences make me think 600% difference might have a reason beyond marketing or brand name. Certainly, every merchant praises his goods. As expected, they are all "superior", "best of technology" or similar. 

I've had similar problems looking for replacement notebook batteries. They too are all "just great" but tend to last maybe a few months before seriously degrading. Since I'm not in the battery business, one "brand" is just as unknown to me as any other. Any supplier that has been proven to deliver overall reliable results?

[tggzzz]

No. There are different series Smart-UPS. Those cheaper are classic line-interactive type.

I could check an old data sheet and it indeed notes "line-interactive" for that model and that the "inverter is permanently connected" and used both ways. So I guess, in case of a bad inverter the battery would not charge either.

Eventually an opinion about the replacement batteries might help.

What I have read about batteries for UPS use (not the solar-technology type), there seems no clear preference over the different technologies used. AGM batteries in average seem to be more expensive compared to Gel batteries but seem to offer no significant advantage when you expect a rather short power drop a few times and maybe a full cycle each year with a moderate (~25%) load.

But the topics I found were not the newest ones. Does this still hold? Modern systems most likely use modern cells like LiFePo instead of Lead, that may account for the lack of recent discussions about Lead-Batteries.

It doesn't make a fortune for just a small UPS, but choosing an "RBC5 battery pack" (= 2x 7.2 Ah) between 32€ (Q-Batteries, AGM?) and a 180€ (Original Schneider) with seemingly no obvious differences make me think 600% difference might have a reason beyond marketing or brand name. Certainly, every merchant praises his goods. As expected, they are all "superior", "best of technology" or similar. 

I've had similar problems looking for replacement notebook batteries. They too are all "just great" but tend to last maybe a few months before seriously degrading. Since I'm not in the battery business, one "brand" is just as unknown to me as any other. Any supplier that has been proven to deliver overall reliable results?

You must use the same type of battery specified by the manufacturer; see the datasheet and/or operations manual. Replacing a lead acid battery with, say, a lithium battery will lead to "interesting" consequences.

You can, in theory, use smaller capacity batteries than the manufacturer specifies, but I would be cautious about leaving those on continuous charge, and about the peak current that can be drawn during operation.

If your UPS specifies AGM lead acid batteries, then you might notice that there are two slightly different types: ones designed for continuous charging (e.g. UPS), and ones designed for deep discharge (e.g. mobility scooter).

My UPS is used as a power source after mains power has failed until power returns, so I used deep discharge batteries https://www.allbatteries.co.uk/sealed-lead-acid-battery-nx-13-12-cyclic-12v-13ah-f6-35-amp9051.html . When not in use (i.e. most of the time) it is disconnected from both the mains and battery. Once a month I connect the battery and mains, turn it on, perform a self-test. When, after a few minutes, the battery has recharged to 100%, I disconnect the mains and battery again.

[Jeroen3]

You should not change the chemistry of the batteries. Just replace them with new sealed lead acids.
Specific type doesn't matter, this use case in inefficient with all options.
Should not be too expensive, around 30$ per unit. (+WEEE fee)

The sizes of these batteries are reasonably similar, so you can probably find suitable Yuasa or whatever your local shop has. Looks like the "genuine" kit comes with a fuse element, can this be swapped?

When you buy two of the same you could opt to balance them before placing in series in the unit. I'd assume this is the reason for the price of the genuine ones.

[smallfreak]

When you buy two of the same you could opt to balance them before placing in series in the unit. I'd assume this is the reason for the price of the genuine ones.

Just to provide a summary...

I certainly did not want to change chemistry. All my previously listed batteries were Lead-Acid variants ranging from "High-Quality" to "Great Value", depending on the marketing blahblah...

I did order a set from the cheap end of the list, balanced them overnight to +/- 1mV before I put them in series. The UPS attested them a 70% charge and after fully charging they were accepted as good.

Inverter is working fine, run time is within specs. So all perfect with the device. 

I managed to grab an old service manual where there is listed a series of checks to determine the proper function. One of it is a load/runtime test with various loads up to the full 450W (700VA). I did not do THAT, since I assume that draining the entire charge within 4 minutes    will noatably reduce battery life.

I'm glad that this 20 year old UPS (manufactured in 2003) is still good like fresh, although I learned about a long standing "R38" thermal design problem with these models that might need attention in the next 20 years.

[darkspr1te]

If the UPS is one of the many APC smartUPS types then you can reset the battery parameters via the USB/serial cable.
see
https://networkupstools.org/protocols/apcsmart.html
for details


darkspr1te

[Vovk_Z]

Inverter is working fine, run time is within specs. So all perfect with the device. 
......
I'm glad that this 20 year old UPS (manufactured in 2003) is still good like fresh, although I learned about a long standing "R38" thermal design problem with these models that might need attention in the next 20 years.

Nice to hear a feedback.
But the year 2003 may mean it needs recap electrolytes. I don't remember the whole story about whether it was done. Typically all 22 uF 25-50 V, and several large ones.