wesnewell
100 GW
I've got 5 ups's in the house And I'm sick of buying sla batteries for them. I've got lots of rc lipo and I'm thinking about using a 7s pack to replace some old sla's in 24V ups systems. Anyone here tried that yet? Problems?
RC lipo to replace 24V sla in ups?
- Thread starter wesnewell
- Start date
DVDRW
100 W
I think it will work fine. Voltage is right. Example 12V system can use only Lifepo4. But in 24v (7s) 48v (14s) and 72v (21s) systems Lipo/Lion fits perfectly.
But more dangerous than SLA. I'd reccommend NCR18650 (safer and cheaper)
But more dangerous than SLA. I'd reccommend NCR18650 (safer and cheaper)
TheBeastie
1 MW
Yeah I thought about it a bunch of times my self, never done it though, but the UPS is in a shed not a house..
I think if you fireproof the hell out of it with the expectation it will burst into flames then you should be OK..
I like those fibreglass fire blankets and then on top in a steal casing but your gotta be able to let the gas out or else your creating a hydrogen gas bomb and the gas will get out no matter what.
But if your UPS's are in the house then you gotta think about smoke damage.
I think if you fireproof the hell out of it with the expectation it will burst into flames then you should be OK..
I like those fibreglass fire blankets and then on top in a steal casing but your gotta be able to let the gas out or else your creating a hydrogen gas bomb and the gas will get out no matter what.
But if your UPS's are in the house then you gotta think about smoke damage.
Keep in mind most UPS charging circuits, being meant for SLA, continuously trickle charge whatever is hooked up--that's why the SLA die, cuz eventually they outgass everything liquid in them from that. I tried FLA car batteries for a while with a big UPS, and they would get very low in water after a few months, having to be constnatly topped up.
Since it's probably bad to keep charging Lithium batteries constnatly after they are already full, you might want to put a BMS on them to cut off charge when they are full, even if that's it's only function. I think Ykick has a recent thread (last week or so) about a system that would probably work for that.
Since it's probably bad to keep charging Lithium batteries constnatly after they are already full, you might want to put a BMS on them to cut off charge when they are full, even if that's it's only function. I think Ykick has a recent thread (last week or so) about a system that would probably work for that.
TheBeastie
1 MW
Another thought I had was buying cheap "protected" 18650s but I guess at 1amp discharge limit you would need a fair amount of them.
http://www.banggood.com/10PCS-MECO-3_7v-4000mAh-Protected-Rechargeable-18650-Li-ion-Battery-p-992724.html
They have the built in cut off on charge and discharge..
You can run a discharge test, take pictures on them and open a dispute to get your money back for the amount of mAh thats missing..
Or something that might be more trustworthy?
https://www.fasttech.com/product/2204400
https://www.fasttech.com/products/1420/10002180/2769802-authentic-soshine-18650-3-7v-2600mah-rechargeable
http://www.banggood.com/10PCS-MECO-3_7v-4000mAh-Protected-Rechargeable-18650-Li-ion-Battery-p-992724.html
They have the built in cut off on charge and discharge..
You can run a discharge test, take pictures on them and open a dispute to get your money back for the amount of mAh thats missing..
Or something that might be more trustworthy?
https://www.fasttech.com/product/2204400
https://www.fasttech.com/products/1420/10002180/2769802-authentic-soshine-18650-3-7v-2600mah-rechargeable
wesnewell
100 GW
The first ups I'd test it on only pulls 20A continuous max, but another would pull 40A continuous max. Don't think I want to trust that to weak ass 18650's. That's probably more than the hover boards draw, and we all know what happens to them with weak 18650's. I'll stick with rc lipo rated for at least 200A at 10ah.
I also think the charger is limited to 29.2V max, which is perfect for 7s. But I'll monitor it to be sure it doesn't overcharge.
I also think the charger is limited to 29.2V max, which is perfect for 7s. But I'll monitor it to be sure it doesn't overcharge.
I've had used A123's running 4s on a 12V UPS for many years and those work OK. I think they're near 10 years old and the elevated temperature (35C/95F with no load) in the UPS has finally worn them down.
Lipos really don't like sitting around at full charge and they'll age fast that way. A UPS might limit the around 28V or 4V/cell which would reducing the aging a lot compared to 4.2V/cell. If your UPS also elevates the temperature, that will accelerate their decline.
You might also want to think about some sort of balancing circuit. If one cell leaks just a little more than the rest, then its share of the voltage will end up on some other cell and that can bring on a fire a lot quicker. The A123's I've used seem to limit that just a little - they seem to get very slighly leaky over 3.5V which keeps them in check as long as all of the cells are in good condition, but that could change over time. I've checked mine periodically and found one cell up around 3.55-3.60V and the rest in the 3.45-3.50 range. I've often thought that a 1mA balancer would be useful in long-term situations like a UPS, but lower current doesn't really simply the design of a balancer any.
I really like the safety that the LiFePO4 formulation gives. You might also consider Headway cells. They'll probably both be more expensive than lipos, but they may last more years.
Good luck!
Lipos really don't like sitting around at full charge and they'll age fast that way. A UPS might limit the around 28V or 4V/cell which would reducing the aging a lot compared to 4.2V/cell. If your UPS also elevates the temperature, that will accelerate their decline.
You might also want to think about some sort of balancing circuit. If one cell leaks just a little more than the rest, then its share of the voltage will end up on some other cell and that can bring on a fire a lot quicker. The A123's I've used seem to limit that just a little - they seem to get very slighly leaky over 3.5V which keeps them in check as long as all of the cells are in good condition, but that could change over time. I've checked mine periodically and found one cell up around 3.55-3.60V and the rest in the 3.45-3.50 range. I've often thought that a 1mA balancer would be useful in long-term situations like a UPS, but lower current doesn't really simply the design of a balancer any.
I really like the safety that the LiFePO4 formulation gives. You might also consider Headway cells. They'll probably both be more expensive than lipos, but they may last more years.
Good luck!
wesnewell
100 GW
As it turned out, the ups I was going to test this on had actually just crapped out with a constant replace battery alarm. It was 13 years old, so I just bought a new one. A new 1500VA was only $140. Didn't feel like wasting time with it. Maybe later.
wesnewell
100 GW
As it turned out the ups actually needed new batteries. 
Anyway been running some test on a n old APC BK400 with a 4s 5000mah rc lipo pack. Had to take the voltage down to 13.5V for it to work, and it crapped out in seconds under a 100W load when voltage dropped to 10V when I unplugged the power, but that was expected. What I was really concerned with was what voltage it would charge back to. added a volt meter and it went to 13.6V and steadied off there. The power analyzer on the balance plug reads 13.701V. From this I'm going to assume that a 24V system won't charge over 28V or less. which sounds great for a 7s lipo pack keeping per cell charge to 4V or less. Don't know when or if I'll get around to testing this out, but I'll follow up here if/when I do.
Anyway been running some test on a n old APC BK400 with a 4s 5000mah rc lipo pack. Had to take the voltage down to 13.5V for it to work, and it crapped out in seconds under a 100W load when voltage dropped to 10V when I unplugged the power, but that was expected. What I was really concerned with was what voltage it would charge back to. added a volt meter and it went to 13.6V and steadied off there. The power analyzer on the balance plug reads 13.701V. From this I'm going to assume that a 24V system won't charge over 28V or less. which sounds great for a 7s lipo pack keeping per cell charge to 4V or less. Don't know when or if I'll get around to testing this out, but I'll follow up here if/when I do.
parabellum
1 MW
I use 3s 18650 packs for 2 UPS's. I use protection circuits (will call it BMS) and when they cut off the battery UPS trickles up to somewhere between 15.5 to 16.5V, very unstable, fluctuates over time and differs on both UPS's (being same model). I encounter 2 problems, both seemingly related with BMS:
1) At power shut down BMS does not come in, I believe it switches ON to slow and UPS errors, but it only happens 1 time of 10. UPS off/on and it works again.
2) When the battery is fully discharged BMS shuts down and UPS does not detect battery when power returns, and does not send any charging current. It is only solved getting some voltage on UPS leads to reactivate charging and open protection circuit. I use external PSU from LineIN to wake them.
1) At power shut down BMS does not come in, I believe it switches ON to slow and UPS errors, but it only happens 1 time of 10. UPS off/on and it works again.
2) When the battery is fully discharged BMS shuts down and UPS does not detect battery when power returns, and does not send any charging current. It is only solved getting some voltage on UPS leads to reactivate charging and open protection circuit. I use external PSU from LineIN to wake them.
wesnewell
100 GW
Yeah, I know that rc lipo won't work with 12V system because of the voltage differences, but it should work great with 7s for 24V systems, which should charge to ~28V-29V, and never more than 29V.
parabellum
1 MW
Those UPS's are made for dumb VRLA batteries where no precise LVC and HVC is needed. As I described, my units LVC randomly at 9-10V (2.7V/cell is my BMS LVC) and trickle randomly at 15.5-16.5V, is corresponds to 18V-33V DOD for 24V system.
Not sure I could sleep well without some kind of BMS, protecting 7S LiPo in such a system.
Not sure I could sleep well without some kind of BMS, protecting 7S LiPo in such a system.
dmwahl
10 kW
I've used some old LiFePO4 with the crappy single cell "balancers" with a UPS and it works great. The voltage comes up to about 13.6V (3.4V/cell) and the current drops to zero. The charging circuit doesn't force a trickle current into the battery once it's at the maximum voltage, it just holds it there with zero current. I know some have fears about float charging LFP, but most manufacturers spec a float voltage around 3.40-3.45V as being fine for the cells. That said, I wouldn't use RC lipo in any situation where you're not watching it.
Keep in mind also, that many home UPS products will overheat if run longer than the stock batteries would let you. So if there's a 10Ah SLA to begin with (yielding maybe 5Ah under load) and you put a 10Ah lithium in it, actual run time will roughly double and you may burn up the UPS. I made the mistake of doing this with a 100Ah LFP pack, and came back to the UPS stinking like burnt electronics and with a tripped circuit breaker. That was with a roughly 100W load from my desktop PC.
Keep in mind also, that many home UPS products will overheat if run longer than the stock batteries would let you. So if there's a 10Ah SLA to begin with (yielding maybe 5Ah under load) and you put a 10Ah lithium in it, actual run time will roughly double and you may burn up the UPS. I made the mistake of doing this with a 100Ah LFP pack, and came back to the UPS stinking like burnt electronics and with a tripped circuit breaker. That was with a roughly 100W load from my desktop PC.
wesnewell
100 GW
The ups's I use all have a setting for when to shut down the ups at a certain percentage of battery soc. Set it to the percentage that allow the ups to run on battery power for just so long and that shouldn't be a problem. Of course this assumes one has setup the controlling software correctly to begin with and knows how to change the settings. I have and do.
wesnewell
100 GW
Finally able to remove a cell from a 4s pack and have a 7s pack. As you can see, it won't over charge a 7s lipo pack. Max battery voltage I saw was 28.3V reported for a couple of seconds. In any case, 7s would work fine.
APC : 001,046,1161
DATE : 2016-09-15 10:51:43 -0500
HOSTNAME : mythfe0
VERSION : 3.14.10 (13 September 2011) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2016-09-14 23:59:49 -0500
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 118.0 Volts
LOADPCT : 0.0 Percent Load Capacity
BCHARGE : 100.0 Percent
TIMELEFT : 71.7 Minutes
MBATTCHG : 25 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Medium
DWAKE : 000 Seconds
DSHUTD : 000 Seconds
LOTRANS : 100.0 Volts
HITRANS : 136.0 Volts
RETPCT : 000.0 Percent
ITEMP : 29.2 C Internal
ALARMDEL : 30 seconds
BATTV : 27.9 Volts
LINEFREQ : 60.0 Hz
NUMXFERS : 2
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
APC : 001,046,1161
DATE : 2016-09-15 10:51:43 -0500
HOSTNAME : mythfe0
VERSION : 3.14.10 (13 September 2011) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2016-09-14 23:59:49 -0500
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 118.0 Volts
LOADPCT : 0.0 Percent Load Capacity
BCHARGE : 100.0 Percent
TIMELEFT : 71.7 Minutes
MBATTCHG : 25 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Medium
DWAKE : 000 Seconds
DSHUTD : 000 Seconds
LOTRANS : 100.0 Volts
HITRANS : 136.0 Volts
RETPCT : 000.0 Percent
ITEMP : 29.2 C Internal
ALARMDEL : 30 seconds
BATTV : 27.9 Volts
LINEFREQ : 60.0 Hz
NUMXFERS : 2
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
wesnewell
100 GW
Ran more and final test today using 5ah 7s rc lipo cells. 24V Conext (APC) 750VA ups, 100W load. Ran it down until the ups actually shut itself off. Monitoring the cells, the lowest one got to 3.49V just before shutdown. It recovered to 3.52V before I plugged it back in. Or approx. 3-4% soc. These are cells over 4 years old. When fully charged, software was reporting total voltage of 28.1V, which averages out to 4.01V per cell. A check of cells with a medic showed them at ~3.95V per cell. but either way, well within the max charge of 4.2V per cell. So my conclusion is 7s rc lipo on a 24V ups system not only works good, but will be far superior to sla batteries. As a disclaimer, it is not recommended by the manufacturer, and will possibly void the warranty, which will probably be up by the time the original batteries die anyway. Also the risk of lipo batteries catching fire will be greater, so do so at your own risk. I'm not worried about it.
wesnewell
100 GW
Been running the ups live now for a few months without a problem and have come to the conclusion it shouldn't be a problem. I've run a calibration cycle on the ups using the 5ah pack and this is what I get after it charged back up.
APC : 001,045,1087
DATE : 2016-11-27 12:32:37 -0600
HOSTNAME : mythfe0
VERSION : 3.14.10 (13 September 2011) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2016-11-26 00:29:02 -0600
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 123.0 Volts
LOADPCT : 11.0 Percent Load Capacity
BCHARGE : 100.0 Percent
TIMELEFT : 35.8 Minutes
MBATTCHG : 0 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Medium
DWAKE : 000 Seconds
DSHUTD : 000 Seconds
LOTRANS : 103.0 Volts
HITRANS : 136.0 Volts
RETPCT : 000.0 Percent
ITEMP : 29.2 C Internal
ALARMDEL : 30 seconds
BATTV : 27.9 Volts
LINEFREQ : 60.0 Hz
LASTXFER : Low line voltage
NUMXFERS : 0
TONBATT : 0 seconds
CUMONBATT: 0 seconds
XOFFBATT : N/A
SELFTEST : NO
STESTI : None
STATFLAG : 0x07000008 Status Flag
MANDATE : 2003-10-13
SERIALNO : QC0342230498
BATTDATE : 2016-07-20
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
FIRMWARE : 811.r1b.D USB FW:r1
END APC : 2016-11-27 12:32:44 -0600
APC : 001,045,1087
DATE : 2016-11-27 12:32:37 -0600
HOSTNAME : mythfe0
VERSION : 3.14.10 (13 September 2011) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2016-11-26 00:29:02 -0600
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 123.0 Volts
LOADPCT : 11.0 Percent Load Capacity
BCHARGE : 100.0 Percent
TIMELEFT : 35.8 Minutes
MBATTCHG : 0 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Medium
DWAKE : 000 Seconds
DSHUTD : 000 Seconds
LOTRANS : 103.0 Volts
HITRANS : 136.0 Volts
RETPCT : 000.0 Percent
ITEMP : 29.2 C Internal
ALARMDEL : 30 seconds
BATTV : 27.9 Volts
LINEFREQ : 60.0 Hz
LASTXFER : Low line voltage
NUMXFERS : 0
TONBATT : 0 seconds
CUMONBATT: 0 seconds
XOFFBATT : N/A
SELFTEST : NO
STESTI : None
STATFLAG : 0x07000008 Status Flag
MANDATE : 2003-10-13
SERIALNO : QC0342230498
BATTDATE : 2016-07-20
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
FIRMWARE : 811.r1b.D USB FW:r1
END APC : 2016-11-27 12:32:44 -0600
wesnewell
100 GW
Been over 2 years now (July 2016) that I replaced my sla batteries in my 24V ups with some 5 year old 7s lipo. Just though I'd update this for anyone interested.
wesnewell
100 GW
Lipos don't get hear as warm as the sla's did, which normally had a ups temp of 40C. It now stays at a constant 29.2C. Latest readings
APC : 001,045,1033
DATE : 2019-01-15 11:01:04 -0600
HOSTNAME : mythfe0
VERSION : 3.14.14 (31 May 2016) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2019-01-14 10:01:31 -0600
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 116.0 Volts
LOADPCT : 3.0 Percent
BCHARGE : 100.0 Percent
TIMELEFT : 71.8 Minutes
MBATTCHG : 0 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Low
DWAKE : 0 Seconds
DSHUTD : 0 Seconds
LOTRANS : 96.0 Volts
HITRANS : 136.0 Volts
RETPCT : 0.0 Percent
ITEMP : 29.2 C
ALARMDEL : 30 Seconds
BATTV : 28.3 Volts
LINEFREQ : 60.0 Hz
LASTXFER : High line voltage
NUMXFERS : 0
TONBATT : 0 Seconds
CUMONBATT: 0 Seconds
XOFFBATT : N/A
SELFTEST : NO
STESTI : None
STATFLAG : 0x05000008
MANDATE : 2003-10-13
SERIALNO : QC0342230498
BATTDATE : 2016-07-20
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
FIRMWARE : 811.r1b.D USB FW:r1
END APC : 2019-01-15 11:01:05 -0600
APC : 001,045,1033
DATE : 2019-01-15 11:01:04 -0600
HOSTNAME : mythfe0
VERSION : 3.14.14 (31 May 2016) debian
UPSNAME : APCUPS
CABLE : USB Cable
DRIVER : USB UPS Driver
UPSMODE : Stand Alone
STARTTIME: 2019-01-14 10:01:31 -0600
MODEL : Conext CNB750
STATUS : ONLINE
LINEV : 116.0 Volts
LOADPCT : 3.0 Percent
BCHARGE : 100.0 Percent
TIMELEFT : 71.8 Minutes
MBATTCHG : 0 Percent
MINTIMEL : 3 Minutes
MAXTIME : 0 Seconds
OUTPUTV : 120.0 Volts
SENSE : Low
DWAKE : 0 Seconds
DSHUTD : 0 Seconds
LOTRANS : 96.0 Volts
HITRANS : 136.0 Volts
RETPCT : 0.0 Percent
ITEMP : 29.2 C
ALARMDEL : 30 Seconds
BATTV : 28.3 Volts
LINEFREQ : 60.0 Hz
LASTXFER : High line voltage
NUMXFERS : 0
TONBATT : 0 Seconds
CUMONBATT: 0 Seconds
XOFFBATT : N/A
SELFTEST : NO
STESTI : None
STATFLAG : 0x05000008
MANDATE : 2003-10-13
SERIALNO : QC0342230498
BATTDATE : 2016-07-20
NOMOUTV : 120 Volts
NOMINV : 120 Volts
NOMBATTV : 24.0 Volts
NOMPOWER : 440 Watts
FIRMWARE : 811.r1b.D USB FW:r1
END APC : 2019-01-15 11:01:05 -0600
wesnewell
100 GW
Well this experiment is over. The 8 year old lipo batteries I had in the ups finally got to where they had very little capacity and rather than spend time building up a new pack I just went back to sla. Actually, agm this time. Anyway after almost 3 years of use, without ever checking them, cell voltages ranged no higher than 1 at 4.20V, with other down to 3.5V. Of course the batteries I used were almost 5 years old to begin with and all or most had been puffed and depuffed before using them in the ups.
