UPS upgrade - who's done it?

TeddyandJ

TeddyandJ

Member
Joined
Feb 8, 2018
Messages
71
Howdy all,

My area is notorious for micro blackouts. Usually just a few minutes long. It's one of the reasons I started on the power wall journey.

Since my wall is going in the van instead of the house, I thought I'd try and beef up a couple of old UPS units I have lying around (likely only 12v).

Question: Do I bother investigating a lithium upgrade or just strap an old AGM lead battery to it and be done?
 
I have 2 1500VA (24V) APC ups's running some base loads in my house running off of 7s Lithium. Works great as a Peak offset power wall where I turn the input to the ups off in the morning when the costs go up and On again when they get cheaper. The UPS charges the wall overnight (with some help from an external charger)
 
Using 12v UPS, I would recommend LiFePO4 as it will match perfectly. Or LTO.
 
Pretty much what I thought. Can't justify $$ to buy LiFePO4 :(.

Until one of the commercial UPS units from work hit the ewaste pile, I'll give the old AGM a crack. Only has to work a few times to be worthwhile. Gotta last longer than the 7ah SLA.
 
Charging lithium from an UPS is not recommended unless you can change values. Charging 12V LiIon wont work at all.. LifePoe4 is only that works for 12v but still not optimal unless you can change values

Most people here charge their ups batteries with external charger instead.
 
Don't remember where you're from, but if you are in the US, you could probably use eBay and get a server UPS unit. There are some fairly inexpensive units available on eBay. I think you can get a 3000 W unit for just around $100 or less. You just have to make sure that they do not ship batteries with the unit so that you can save on the shipping
 
5s LTO cells work fairly well with 12v UPS arrangements, they even match the charge voltages output by default on most UPSs

My initial testing has shown that if the cells are balanced to start with they mostly stay balanced during charging....mostly. They do tend to creep so they will eventually go out of balance.

I have just purchased a stack of these to use to manage this:
https://www.ebay.com.au/itm/263810452104?var=563156047608

The bypass current is quite low, only 1A, but this is only relevant when charging is almost complete and overall charge current should be lower any.

They only arrived yesterday so experiments haven't started yet, I'll let you know if the smoke comes out of my theory.
 
The easiest answer is just slam a large 12v car battery, deep cycle, or AGM and let it ride especially if it is 12v. LifePO4 is great too but sometimes the easiest and most cost effective route (Lead) really does make the most sense.

That said, I have a 24v and 48v UPS, I plan on doing a lithium conversion, but it is somewhat low on my priority list at the moment. Need to do more voltage monitoring at both low and top end before committing to the project.
 
I have been looking for while to convert ups from AGM to li-ion but I am not finding a solution to float charge.
li-ion should/must not be float chargee.

Is there an inexpensive import BMS that it is also a over-under voltage protection board good enough?

Some APC ups's do have float voltage adjustment, but not to be disabled.
 
marcin said:
I have been looking for while to convert ups from AGM to li-ion but I am not finding a solution to float charge.
li-ion should/must not be float chargee.

Is there an inexpensive import BMS that it is also a over-under voltage protection board good enough?

Some APC ups's do have float voltage adjustment, but not to be disabled.
Click to expand...

What really is float charging to a lithium battery...

Personally I don't see an issue with it, unless you do notexceed the max voltage of the cell. If your "float" voltage does not exceed 4.0v per cell, are you really doing any damage? It's not being over charged, and it's not accepting current as the voltages have reached an equilibrium.

My question has to do with what is the max voltage the cells will see across the entire charge cycle when the APC is doing its thing.
 
Floating Li chemistry cells at a sensible voltage isn't an issue, adjusting the charge algorithm used by equipment usually designed for other chemistry cells is the issue.

A list of UPSs that are proven to allow float voltage setting to a sensible voltage seems to be in demand.
 
I have a data logging multimeter and I intend to find that out myself for the equipment I have... long to-do list.. but it is on the list.
 
OK so the first test is complete, the units work as advertised, however for some reason I decided to purchase the 3.0v versionwhich starts activating at 2.95v. Thisis a little too high for the cells and for practical use with 14.0v charge voltage of Lead Acid UPS.

For descriptive terms the pack is configured + c1 c2 c3 c4 c5 -

UGlea1ADlVw2gLMG7Qs3fz6bL9T8j3FtbCBMTizMNdKjgvfuZajntUtPS3zzXCCHiBHjlIxyO0QMSUMLJ1JCP8YTlmZstLMSIJFhzTyUTr6fi1oS7brw7SOY7aR3YkUlEw9TvJWrtGrKmqoahOf1quGRReCuuwRXsJtgPrdpZZwJ5RLk_6PsCyo3iX1MDjm7ypotcrILAojwZk1ONvRYoJ6UFG4n0usyEBvfrZ57lGPbS4kR-sOLO_xjeidBLkYvhnyC3ulbalLRg5NWEvkuAwmcvQNRi5unZjWXLK7yznzuy18wje5-NXAHSUPW7WMYVQHjZ0dq7G-UTs-gmMICndHUbkIaFHed_XqE4yzN5skzQmH4GdAppnTnuD2dhWOTt4WvZ5F-FnCDiqmeR6o45jzh81I1KgA555rt3VpQuY292C71JimhP0F8AXlNORwQra4ivaS9n7n3Wc9PICnEzODAqD0zywuY75q3qAHeKF-rly3pVs1GdQX_mn1HVWeYELzjiEYnXUqjy-AKvBAT9-8FK3JB7chqQRnc_3gU0ba1bPGjDPiXDqYk9ZukOP_OFTKkmetWpFpTfjVU2PIOif2ZswEDBofJf1WS9zi2faIWA_mZKZqZzuzZoV80Vl0ZN2VuYcu33WMjxXDbaRM-e7naJP10ftD40TszAMaFDAMQ9MBe9YaHEyYKRw=w515-h1058-no


I had to dial the input voltage up to 14.9v to get c3up to full charge

FED8FtIQ4ttGaxTkth5Dr3Q_7K5YTWDkjLAcn-5a6kdgNBLZ-Gb2aGcUERkIREouDUC4mTcDSKViGEW3gdQP_pPqQVQrZ8f57g8HcfSyYXRJiQEmQKQDEK_vGEzyQLspcOiDChWiMTc_fDTV4jvswY-OAjfEdOpSsSgKJd4R554uByePMdf4S6TaXcg1pE8I2y4VDf3lxspLaRitMoj34DV249ejTLx4RHTH_Vl3F-3CBt7GjksDfNzSVtIW9YStSi2DO0oWYI8jIKDVPI29Q7e6444ABaeg0OHFxYGVYvuC-Xc0FbYQWXJR1Mx6tsk-1ZolRxTx_UlGKzAG1wISar1eTOPD9xKWeAVDgH7VnyB1Q64nbOyHmc0SlksZxRs2e71LWNjVvXhe1x5N7BYQE97vftGY9KOra_TPPabQclWe2Jc-W_eYxReT9UXI8OuUIa--v1S5EOVEdJGnu95G1l9mojhBwlZa0EDuFbTXYyNzXRdh273Zabas6gIK9d_2dNNEFnEZHhEjG21qc5waC17RJnkbN6V0cK_o1mTORkB-iHfuhJRhft5xCMxcUjOPSnUZ05arJGW396SsYy6fkqgfbNMkyesYgW0GtLe7XUwMNtGFUqySXeasVsa3k0eb=w515-h1058-no
4Oy9o-6rWqgQes9G_Ty2ys-jD7xz9pANxHEgdtnI8okfwWqapfUOzRHPfSO761zIIHg9qmtDvJNJmrrd9cciKl14o55Jqhg9EUMCqiznFXkVEvgKaICjDgYr-ZKEGakZ-tBTbP6CRq3mRt9VoyH6AZBVEX8UeXzQRJJ_pI-S56r79c_j1SUJize_HF2RCXkWVe52LZ05Tn0LkzV50z9SL4irFTnpGygWrol4HTjl73idLV8RQ9VOQM2OwYTNFLVgjKE57UHqTOnOuRuzX77257m8XHX1Kxp1j1_meU80iV1VuFe0VdEwaWEAeXfmgQCmUnzWkKFaAVm68m0U9KSfaxxQopyQ1MYzFQZzpHyPAbuqttY1JccIBfegxD9lDOFZCquld78Erx4UfC6XaEb7BZrky8UM4V5UgS_CH7fAzf1po4Cd_pv1s21rBBOpJ_WToV-NqUDJ-96q6x_SrAInP777n8k94NPQLm836PfbALOFcx5wLyLYMMQhsI-zT1l2jFM4Z6G6noYrJ6EknPkl4x-Ey_e3ulx5HKzri4UJlY2mXsbXqn56zxCvyts2Se9XGKzTy6PbuLtB_SFvRdPJosl66luPpbPhB0n4or5FGqtw9XIy8FKeMfLPIttQ72q0kAnvK2U4l_lkFvCT7nJl-FSyYdzV7LMdiurggSQu4vC89b3VSPTuJgkB8Q=w515-h1058-no


At 14.4v c3 topped out at 2.45 and c4 at 2.65, at this plateau the balance boardshad deactivated and no charge current was being drawn from the PSU.

Once the balance board activated the cells were managed quite well, all holding at 3.020-3.050v, once the PSU was turned off the balance board burned off theadditional voltage until it deactivated.

CMHuGH0vWo2xsC5b-Kp_EjZ72-LkbtAbSQDIfu_QomydGIfb4II94B3UYkvmbIQ_uzdk-_RMNa2vk3jRfndIwAPiZggnTvzCumggOR-oR04-i5V1810UMrwUpyQPhn_D2eYUQj5FZV8fP5LArWBshSy3jdvPfN0TOYXVEFIrD3MJZS76qWTiImU1bRrN195mo4H4vfWIBuji18Be8xKbVWqz1wfwSZjht7mHFumxCJDyR-nr8bk6PYCoRPy_jeNPDTfFkQSZeyP3XgM6MOYBUTOpS1CIPtBcL-pdMwY0cZUawikw7m9b_kOVkWIXedao1jKF05ZmkSwDshLLEMQVKB1Qb5Z8LO1nj8Gx28kFLctG9PWz-hPh6vL9I-BQ_SjT4zDBrgBJJDt_K-dUwDHXqXc9C5kwjFqAGUv0O-4fcUuNw39F83I-D5pRrhhTiR8N0e5MtApUeHzga-Mth5tPY9fvFW0ngqfpRkVcJjvtX3VRancY9v1XhAB4F1h_6T1qA0SOoN1j6lnLcS7a4pB9G2EOi5SwAsvfjCo_p8Jqunx273vejIKJxEEzBPWzHLWUQuoMfPlV8hK_JdXEecD6Dkp4SptrRTb8vZNS_vGsDmCSaeexi31LK4fPgRXYc_4pkufiXcA46_jF_J0mngCvglqTxQxb5fsOXZ20uIGHe_jz5HdgECnDeZlHPA=w515-h1058-no

This was still charging c3 very slowly, slow enough that the remaining cells dropped to 2.995-3.010, deactivating the balance board.



I really should have purchased the 2.7v version, I will next time, I'll also have a fiddle with the resistor value to adjust the trigger voltage.
 
CrimpDaddy said:
The easiest answer is just slam a large 12v car battery, deep cycle, or AGM and let it ride especially if it is 12v. LifePO4 is great too but sometimes the easiest and most cost effective route (Lead) really does make the most sense.

That said, I have a 24v and 48v UPS, I plan on doing a lithium conversion, but it is somewhat low on my priority list at the moment. Need to do more voltage monitoring at both low and top end before committing to the project.
Click to expand...

Agreed. Also most 12v home/office grade UPS are not designed to run as inverters for extended periods of time. At the bare minimum extra cooling is required. Most often these units are designed around their battery capacity, and if there run time is increased they overheat.

Rebuilder.mike did a conversion on a 12v unit. Rev0 has a write-up on a 36v unit.
 
Geek said:
Agreed. Also most 12v home/office grade UPS are not designed to run as inverters for extended periods of time. At the bare minimum extra cooling is required. Most often these units are designed around their battery capacity, and if there run time is increased they overheat.
Click to expand...

I can concur. Last year I had an APC 450 unit I used during our power outtage that lasted over a week. I had the case off of it, and my cheapo DMM sitting on the pcb with the probes connected to the battery connections (it was here because the ups was on the counter and the blade terminals were easy for the probes).
Well, after the 2nd day, I had a nice hole melted in the back of the DMM. It still functioned, tho it had extra cooling :p
 
Korishan said:
Geek said:
Agreed. Also most 12v home/office grade UPS are not designed to run as inverters for extended periods of time. At the bare minimum extra cooling is required. Most often these units are designed around their battery capacity, and if there run time is increased they overheat.
Click to expand...

I can concur. Last year I had an APC 450 unit I used during our power outtage that lasted over a week. I had the case off of it, and my cheapo DMM sitting on the pcb with the probes connected to the battery connections (it was here because the ups was on the counter and the blade terminals were easy for the probes).
Well, after the 2nd day, I had a nice hole melted in the back of the DMM. It still functioned, tho it had extra cooling :p
Click to expand...

That's very scary. Interesting too.. What load did you have on it?
 
It was running a TV and a ceiling fan, led lights and charging my phone.

It's probably the fan that really hurt it, though. I'm sure it's a modified wave (cheap unit) and inductions don't like modified

Also to note that the transformer became a little loose. By the end of the week, the noise was so loud I had to close the door to the room it was in and I could still hear it and it was quite annoying. Not sure if some of the laminations came loose, or just loose from the pcb. I haven't studied it closer to find out.
 
I've got an APC950. I've had it run my 60" TV and the attached media centre and a NAS for about 6 minutes. Bet the battery didn't like that, but it didn't error out.

If the AGM can run all that happily for 20mins or more, I'll be happy.
 
You must log in or register to reply here.
Back
Top