You're a legend. with simple tool and calculations you knew how to answer a largely asked question by many ppl who want to convert to lithium batteries for their ups. thank you
cant i just slap on a bms to mitigate the minimum discharge voltage of the 12v ups 4 cells pack as in the bms will shutoff the connection when the discharge voltage reaches 11v therefore keeping the battery safe
Some UPS motherboard are having some resistors around an OpAmp to set this 8.7v low discharged and 12.6v high charged (replaceable by a adjustable small resistor or 2) on you tube I seen once a video on Eaton Elipse UPS for modded Li Ion 12v/9Ah Lead case to 12v/14Ah battery I cannot find it anymore (you must have MB schematic diagram to identify resistors)... If someone find it... Or knows how...
Depending on what UPS you use, it may be possible to modify the charge voltage, and the cutoff voltage. If you're able to do so, you can utilize more of the cell's capacity.
I'm really missing the 48v one here... Especially since that's the most common one in rack mount form factor... Would be nice to know whether it's worth converting it.
I have a smart ups apc 3000 120v, that says it is constant voltage charging at 53v. The cutoff low voltage that I found it to switch off at (shutdown) is 37.18v. I have determined that using LIPO batteries I will need to be at 13S configuration to make sure that the the low cutoff isn’t lower than my cells ratings and the upper voltage will not overcharge them either. However, I see that I also will not be getting the full rates AH out of the cells I plan to use. The capacity of these particular cells is 2.5v (low) and 4.2v (full). So in a 13s set, on my ups, the cutoff voltage per cell is 2.85v and the upper charge will be 4.07v. so the capacity of my battery pack will not be whatever it’s rated at but a lower number as the range is lower than I would like. However since I am planning to use LIPO this is the safest way to do it as I see it. And it has the added benefit of longer life of the cells. The cells I plan to use are 8AH so I might get 6.5AH in this configuration. My only concern is the constant voltage charging, what will happen when the 53v meets the 37.8v how many c will that put into the cells? Specs say I can charge these cells at a 120a so I should be good but... I’d like to do the numbers first. SPIM08HP are the ones that I have to use in this. They are scary powerful but with a bms it should work fine
@@jgren4048 Should be fine. A 12v 12Ah SLA want less than 3.6A initial charging current so I'm certain it will be CC limited. As far as I understand Pb chargers, They run in 3 stages, CC then CV and end on 'float' or equaliser mode. So just like Lipo chargers, they start with preset current till the set voltage is reached, then switches to constant voltage and the battery accepts what it can at that voltage till its full. Lipos stop accepting current if the cv voltage is reached while Lead Acids begin to hydrolyse at the electrode and boil off the electrolyte which can help with desulfation and equalises the voltage in each cell. SLAs need to be kept fully charged at all times, Lipos will last much longer being charged below their upper threshold. The float charge mode of a Pb charger shouldn't be an issue for the Lipo if its below its max voltage. Lead Acids do not like to be charged quickly but they can disharge insanely hard so the charger will be gentle.
Neither of these configurations is good. Besides the charge bands described, the charging profile must be considered, and a lead acid charge profile is not compatible with Li-ion chemistries. Charge controllers for lead acid chemistries charge in three phases: bulk, absorption and float. Of these, the float phase is particularly problematic for lithium chemistries. Charge/discharge cycles from 4.0V to 3.0V are highly recommended to maximize the life of Li-ion chemistry batteries. Discharging below 2.5V severely compromises their life. Charging above 4.2V is downright hazardous, as well as life-reducing for Li-ion cells. Lastly, Li-ion cells don't like to remain at peak voltage for extended periods; the state where a lead acid charge controller will keep the battery. This, too, severely compromises Li-ion cell longevity.
Yup... Look up something called CC/CV chart.... says it all.... I have been looking at ways to convert my 24v ups to Li-ion. Perhaps if I could figure out how to find the charge unit and add a circuit between it and the battery.... but I'm still working on it. Maybe there is an external way such as a TP4056 for 7s xxp.... hmmm
@@aleksandersats9577 Unless you know some other way, the lead acid battery charger does not give cc/cv charge and in addition it gives a trickle charge which Li-ion cannot deal with.... over/under is important but not the only thing we want...
The 7S for 24V seems like a much better match if you ask me, because the UPS will float the cells at a little over storage charge. Would you really want a large li-ion pack sitting at 4.2V (or possibly 4.22V!!!) for months or years inside your house?? I sure as hell would not.
That is the conclusion I came to for a 24v UPS. For the 36V UPS, the charge voltage would be a little high using 10s so you could use 11s to be safe but you would only be using a little more than 50% of the cell capacity. However this would be great for the cell longevity and if you are using this only as a backup solution then it would be the best option.
I will test mine, I have one 12 v and one 24v. I think the LifePo4 batteries would make a better fit if my voltages turn out close to yours. I don't know yet how to set up bms or anything but those batteries are more forgiving than li-ion
I made a lithium ion battery for an ultima LCD 2KVA EURO ups The battery was as follows: 30 cell bak high drain 4.2 v 3 A 5. The batteries were placed in 5 columns and 6 rows and thick nickel stripe was used for spot welding. I used BMS 6S.The problem is that in few seconds, not more than a minute, the battery gets extremely hot and smoke begins to appear. Please explain the error.
what about car batteries. 1 for 12v, 2 in series for 24v, and 3 in series for 36v. do you know if that work too? and IF yes, do you think there is a limit for paralell conections?
Thank you, You have done the hard work for me. Im going to build a 7s for my ups. I been trying to find a use for some 1.3ah high discharge 18650s. and I also need this used ups I bought.
Isn't this issue exactly what a Balanced BMS is for (i.e. eliminates the potential overcharge in the 12V UPS)? Please correct this newbie if I'm wrong due to misunderstanding.
Thanks for this extensive test! Please can you test and clarify if the same conclusions apply also to battery packs including an internal BMS like the one from ecoworthy? These packs are sold as drop in replacement, but maybe the they can have the same problem of raw cells with UPS voltage. Many thanks for your work!
The difficulty of using the 3S config is: 1. You absolutely must have a BMS. While the low side disconnect is below that of the UPS shut off, the high side will absolutely cook a 3 s without it. The BMS should disconnect charging... is the changeover quick enough for UPS?
HOLY COW!! I just stumbled onto your video!! Been researching for the past week on the tripplite and you had one tested. Looks like 10S would work next to perfect for the 36V. Thanks for the video and testing!! The other 2 can be fixed or even the 36V can be fixed with a BoostBuck converter connected for charging. You would need to disconnect the charging the UPS does, itself or just never use the ac plug. Great work
@njfulwider5 Funny you mention a buck converter but choose to add a boost. I was thinking a DC to DC step down on the charge side of the 12v 3s, and 24v 6s configurations just before the battery connection. By stepping down the overcharge you can charge your lithium to full, and the UPS will reach min before your batteries, and you can still use the original charging system. Im just starting to learn a little electronics after being in IT for years. If there is some reason I'm missing which lead you to a step up like you did please share, but I thought it was more logical to use a stepdown as you retain the original charging and the UPS doesn't;t need 13.64v/27.03v to run, those are just the charge voltages for the original led acid batteries, the UPS will be quite happy at the 12.6v/25.2v. The only downside I see in this, is you are leaving some power in your batteries when the UPS cuts out. However, as a fix for the issue of not being able to use lithium (I keep saying lithium because LiPo batteries have the same nominal voltage and 18650s and it could be cleaner and cheaper to buy a pair of 6000mAh 3s LiPo's than a pack of 10 18650's and would open up a whole new world of applications for your battery with a HUGE increase in discharge output, you could say for instance boost a dead car with a couple 3s LiPo's in parallel, and I like the capacity of LiFe which is a lower nominal Voltage and you need 4s to hit 12v, but that's not an issue with what I learned in this video, I now know how to spec it out) for your 12v and 24v UPS's (which I assume most people have on their home computers anyway) it was worth bringing up a modification to your fix.
@@joeshedler6496 using a boostbuck converter, you can fine tune the charging, no matter the voltage coming in. A boostbuck are great when using a ups that uses 12v or 24v or even 36 if the boostbuck can go high enough. A boostbuck can take the minimum voltage difference that a boost or buck converter can’t do by themselves ( usually 1.1v). Hope that made sense...lol
@@njfulwider5 My fisrt experience with a buck converter was a boost. Im using a 3-30/3-30 to bring the 5v provided by USB up to 12v to run 12v led strips off USB power, well actually over driving them at 13v, lots of open air no heat issues and it gives them just a bit more oumph. The only thing with the converter, because its a boost, my output has to be 2v higher than than the input. I assumed the step down would work the same way with the range of input and a dedicated set output so the charger can do what it wanted, and you provide your battery with, ohhhh shit, I just saw the 2v difference needed, which means you only charge your lithium to 11.6, how sad. Damn, how gullible was I to think my 1 week deep dive into electronics was gunna win me some street cred on youtube with my brilliant suggestion. :P Damnit, if only the gap was wider like 2.1v, I think a buck converter step down (in product listings I see then called both buckconverter and dc-dc step up/down converters same product, wide range of input, and I think its an adjustable pot that lets you choose your output) would be the way to go.
@@joeshedler6496 you have to be real careful with the sellers out there. They’ll sneak in the word boostbuck or step up/step down in the title, just to get more sales. I fell in that trap a few times, and ended up returning the item because of that. There are 3 types of converter. A boost, a buck and a boost buck. Most of my projects I prefer a boost converter, because those usually handle amperage better than buck. Unfortunately the boostbuck converter handles amperage the same as a buck, because of the buck side of the boost buck converter. The boost buck converter takes away the limitations on the voltage side, making them great for ups. Have you been to the Facebook sites like (18650 battery) or (DIY powerwall). Great site I’m apart of that is a wealth of knowledge, when it comes to this. I’m apart of many many others and have my own channel on RUclips. Hope to see you there, you’ll love it if you like projects like this👍👍👍👍
@@njfulwider5 So, I'm a little unsure of what the advantage of a boostbuck over a variable on the fly input and user selected voltage on the output side dc-dc up/down converter is. What voltage limitations are there? Is it more than just a variable voltage on the input? If not, my dc-dc step up converters seem to be doing the same job your describing getting out of a boostbuck. Sorry, I wasn;t kidding when I said I just started in electronics. I know lots about power and some basic electronics llingo, which some of I actually understand, but I don't even know if Im at a highschool level. Im defiantly all theory, and no hands on at this point. An acute benign tremor in my hands kept me from finer solder work, so I pushed off electronics. But I have more time on my hands now and a few days ago I said I should build a drill press like machine to hold my soldering iron and fine points wont be a problem. then like 30 seconds after opening my mouth I engaged my brain and decided that was way over engineered, and I could get the same results from 2 boards, a hinge, and a spring. LOL. so yeah, Im finally doing it haha.
Leaf Xpack thanks for the information i have a 24v ups the same one you showed in the video an i need to get some batteries when you say 7 cells that would be for one if the battery or both ?
@@Shiv.141 You could just hook them up and as long as they are well balanced, it will work. It would be advisable to use a BMS if you think of using the UPS in regular service. I recommend the daly BMS
32650s are going to be slightly difficult if you try to fit it in the same space. 26650s may be better for you. If building an external pack, then it doesn't matter.
There are drop-in replacement LiFePO4 batteries with bms but it is critical that you get quality batteries. Many cheap Chinese batteries of Amazon have a bad bms and other issues.
BTW, you can do something else DO NOT plug in the UPS into the AC plug, BUT use DC power supply at 12.6v and connect it to the li-ion batteries and the wires between the batteries and the UPS.
@@jmd.759 Not exactly. If you put in parallel the DC source with the battery then the UPS take the current from the DC source. If the DC source is absent (power cut) then the UPS take the current only from the batteries.
As long as the voltage per cell remains within the appropriate range for li-ion batteries, then it is fine. This is why I did this video, to determine the maximum and minim voltages. Unfortunately, most UPS do not have a voltage range which is appropriate for li-ion. Lithium iron phosphate batteries work better.
Perhaps I missed it but you *really* need a BMS to monitor the voltage across each li-ion cells. Connecting up li-on batteries in series like this is potentially dangerous - dangerous in that you could have one cell have a thermal run-away event and cook off (catch on fire) and ignite the others. Sooner or laters li-on connected like this WILL get un-balanced (their individual voltages will begin to drift). Do a quick youtube search of have scary even one of these cells cooking off like that. The BMS board will protect against the individual cells from getting un-balanced - but its not a cure all for all risks. Even WITH a BMS board if I was doing this I would have the batteries in a NATO style 7.62MM ammo can - to contain the fire. And even than I would want to have that can located away from anything that might ignite.
Do 10S BMS's exist? Can you actually replace 3 x 12v Lead Acid batteries with 10 x 3.7v Li-Ion Cells in series this way and even charge it with the UPS built-in power supply. Does it need some extra protection like a PMMT controller?
Yes 10s BMS's exist, just do a google search for daly 10s BMS, you can get them from banggood. You wouldn't need any additional protection as long as the UPS doesn't charge to much over 42v and your cells are well balanced, you should be fine.
Now i have a Server Rack UPS with 48v battery voltage Lead Acid voltage range you could argue that Li-Ion 13S has an even better & tighter margin with a range of 35,75v - 55,25v compared to fit 38,8v - 55,2v used by the UPS. The BMS protects the Li-ION being overcharged with UPS Trickle charging. Why didn't you include a 48v UPS test?
Hi, Is there any reason why companies are not producing UPS with Lithium Battery instead of lead acid battery? Is there any issues for building large number of units?
$$$ I would assume. Lead-acid is much cheaper and handles the intended use-case for UPS systems: maintaining power output during short mains power outages. I want a UPS that can run for hours but most of the time, when the power goes out, it's only for several minutes at the longest.
@@Megatog615 You are correct, lead-acid batteries are dirt cheap. Also they like being kept at 100% charge, whereas li-ion do not. Manufacturers could design a ups to maintain a li-ion pack at 80% which would be ideal but again... $$$
Well explained really good job enjoy your video.... I have a 12-volt APC ups and hoping to do 4S 10 P with a BMS to provide low voltage protection and charge balance
If you have BMS better try 3s not 4s, becasue for 4s you will charge up to 13.64v and this mean 13.64/4=3.41v which is basicaly around 0% of the 18650, see this table lygte-info.dk/pic/BatteryChargePercent/BatteryChargePercent.png
Yes it should work! At least when you charge your 18650 battery in a power bank with USB you use 5V from the USB and a BMS for the batteries and it works ….
I have the same 24v APC backup as in the 2nd example. I just wish I knew if or how to covert it to Lithium. Maybe someone could make a video: Converting for Dummies? Then, I might know how to create and install the Lithium batteries into the old housing of the Lead Acid battery packs. I dont even know if Lithium actually works in such a conversion? Is a bms neccesary? What other parts are needed? So many questions, so little information.😟🤦♂️🤷♂️
Even without the apparent expertise of Mr Coloe who commented, somehow I don't think it's as simple as what's presented. Lithium-Ion batteries have been generally available for almost two decades now. Since they have, what, 7x the energy density over lead-acid, why aren't lithium-ion UPS's flooding the market now? Makes an enquiring mind kind of suspicious.
Simple, lead-acid batteries are cheaper than li-ion. Furthermore, li-ion batteries need a BMS to be safe which makes them even more expensive. But probably most importantly, lead-acid batteries die after about 5 years and most people simply buy a new UPS.
I will just say this is how li battery fires get started. The charger circuit is not for li batteries. Your test are not enough to determine how safe it is to swap.
Agreed. It might be wrong to assume the top charge voltage with that method shown in the video because old batteries will have increasing resistance as the battery health decreases which will keep the voltage lower with some charging designs especially if they are undersized to keep manufacturing cost down. It's possible there is a bulk charge rate of anywhere from 14.2-15v per battery.
The other issue with lithium in series is that it needs to have cells balanced especially under high load with the application of a UPS and there is no balancing circuit in a lead acid UPS. If lithium cells come out of balance you will have an issue that will lead to fire.
how about connecting 2 diodes to the charging cable to decrease cell charging voltage by around ~1-1.4V, and 1 diode in the opposite direction not to lose much when discharging?
LiFePO4 is most suitable and works easily with all 12V, 24V and 36V. But also it must be paired with CC/CV dedicated charger (for example from separate wall/PSU connection) which also includes BMS for discharge balancing. These are not expensive.
If you do it assure that you have a HIGH QUALITY charger as charging Lithium Ion batteries can be very dangerous. The lithium iron you suggest is a much better suggestion. Lithium ion thermal runaway is no joke and improperly charged or discharged lithium ion is asking for disaster-just look at the Tesla fires. I'm waiting to see the havoc when the Power Walls start to ignite as they are one of the few not using LiFePO4 but rather go with lithium ion.
@@TVMADoc Yes, for other than mobile use (where weight/power density is important) the LiFePO4 is better choice. In future solid-state litium batteries could be the optimal alternative. For example few days ago Nissan informed than their solid-state tech is moving forward and the larger pilot factory will be ready 2024. They said that 2028 the price should be 75 USD /kWh and few years later drop to 65 USD /kWh. This sounds promising, though we need to wait and see. Of course even larger players - like Toyota - are working with solid-state batteries as well and they said them to be ready even earlier. But for those I don't remember seeing specific price estimations.
For the 12v scenario, did you consider using 3 batteries in a series added to two parallel? You would end up with 9.625 / 14.7 if you really wanted to use your old 12v UPS...
I have a CyberPower CP1000PFCLCD UPS and noticed the life is pretty terrible. I have been thinking about doing a Lithium Iron Phosphate replacement, but the ready-to-replace batteries are crazy expensive. I found some IFR26650PC batteries which are very reasonable (max min 2v, max 4.2, recommended 2.5v & 3.65v tho) but don't know the min/max of my Cyberpower. I guess I need to buy a voltmeter and figure out how to test my unit with it.
Wouldn't be able to sell you overpriced motorcycle batteries then.... Also due to the high cost, lead acid is still probably the way to go for a lot of people. However in my case I can get some slightly fire-hazardous batteries for free, so I can get my dumpster dive acquired backup battery running for less than $20 and have insane battery specification that is completely and utterly overkill on power output and will be limited by the unit itself. Seems like a good deal, other than I really need to see if there is anything that can stop graphene fires...
@Leaf Xpack, I just got a haircut for you, do you like it? LOL. So you mention that as long as we have a max voltage larger than the charge voltage of the UPS and a lower min voltage under the UPS's cut off we are fine to use the UPS to charge. Ok, I follow, and from the zoomed out look, that might seem fine, but when you look closer, you remember you had to build a 10s battery to pair up with a 36v UPS.... How do you address regulating cell voltages across the 10s pack? I wouldn't trust the cells to not fall out of perfect alignment voltage wise for many cycles. If it were me, I'd be buying, or building if I knew how, a 10s balance charge circuit that I would power with the UPS's regular DC charge power. Oh, and if Im adding components like that, I'd add a DC to DC step down converter on the charge voltage of the 12v and 24v UPS's so they can run on 3s and 6s packs :). HINT for anyone who does not want to build the 18650 battery packs you can sub in the same size premade (incl. balance board and connector) LiPo battery pack (same nominal voltage, look up min/max like shown here) which will give your battery a much wider range of use because LiPo batteries have a massive discharge current. Everytime I look, the LiPo's come out pre made and looking all pretty and cheaper than even just the 18650's never mind anything else you need to buy to make your battery, case, connectors, solder, heat shrink. Ok some of those costs are pretty nominal, but I added them just to nail that point down a little more. :)
To keep your cells in balance you'd use a BMS or active cell balancer. I don't think using a DC-DC converter would be wise as it would need to be bi-directional and would require high amerage on discharge. I don't even know if that would exists, and if so, it would be more expensive than the UPS.
lithium ion batteries will not work for a 12v systems efficiently. but it i will work perfect for a 24 volt systems and 36 volt systems. 3/4 not good for 12v system, but 7s is perfect for a 24 volt system and 10s is perfect for a 36 volt system. on the other hand LiFePO4 cells will work perfectly on a 12 volt system since it has 3.2v nominal so 4s will give a nice 12.8v.
@@LeafXpack18650 , i think they are making more money from selling lead acid batteries. Ive bought 3 already for one of my ups. Not to mention the non tech people who would just buy a new ups than bother buying a battery and then replacing the one inside.
You need to lookup the manufacturer's recommendations but in general: Li-Ion full charge 4.2v, discharge under load 2.8v (most people, including myself, use 3v as a lower limit, especially when discharging slowly) LiFePo4 full charge 3.6v, discharge under load 2.5v For these A123 cells, the manufacturer permits a discharge to 2v but from the discharge chart, you can see that the cell voltage drops like a cliff past 2.5v. lib.store.yahoo.net/lib/buddipole/ANR26650M1A-Datasheet-APRIL-2009.pdf
That assumes the backup unit is easily modified circuit or that a beginner is competent enough to reverse engineer the electrical traces and schematic. This one is probably a safer bet for most people, but yes if you can change those parameters and the inverter is still happy with it then that would be ideal.
The concept is all wrong, the charge profile was based on lead batteries but if you remove them the li batteries will change the charge discharge design. You also need betier than a multi meter to see the power profile. You only see the DC value .
You're a legend. with simple tool and calculations you knew how to answer a largely asked question by many ppl who want to convert to lithium batteries for their ups. thank you
Thanks!
Thank you for doing all this, I'm sure you have saved me from either expensive mistakes or even a battery fire!
same here
cant i just slap on a bms to mitigate the minimum discharge voltage of the 12v ups 4 cells pack
as in the bms will shutoff the connection when the discharge voltage reaches 11v therefore keeping the battery safe
Some UPS motherboard are having some resistors around an OpAmp to set this 8.7v low discharged and 12.6v high charged (replaceable by a adjustable small resistor or 2) on you tube I seen once a video on Eaton Elipse UPS for modded Li Ion 12v/9Ah Lead case to 12v/14Ah battery I cannot find it anymore (you must have MB schematic diagram to identify resistors)... If someone find it... Or knows how...
Depending on what UPS you use, it may be possible to modify the charge voltage, and the cutoff voltage. If you're able to do so, you can utilize more of the cell's capacity.
You can change the charging profiles... You just need to change a few things on the UPS mother board..
Or use less of the cells capacity, which is better for its overall Longevity
I'm really missing the 48v one here... Especially since that's the most common one in rack mount form factor...
Would be nice to know whether it's worth converting it.
I have a smart ups apc 3000 120v, that says it is constant voltage charging at 53v.
The cutoff low voltage that I found it to switch off at (shutdown) is 37.18v.
I have determined that using LIPO batteries I will need to be at 13S configuration to make sure that the the low cutoff isn’t lower than my cells ratings and the upper voltage will not overcharge them either. However, I see that I also will not be getting the full rates AH out of the cells I plan to use. The capacity of these particular cells is 2.5v (low) and 4.2v (full).
So in a 13s set, on my ups, the cutoff voltage per cell is 2.85v and the upper charge will be 4.07v. so the capacity of my battery pack will not be whatever it’s rated at but a lower number as the range is lower than I would like. However since I am planning to use LIPO this is the safest way to do it as I see it. And it has the added benefit of longer life of the cells. The cells I plan to use are 8AH so I might get 6.5AH in this configuration. My only concern is the constant voltage charging, what will happen when the 53v meets the 37.8v how many c will that put into the cells? Specs say I can charge these cells at a 120a so I should be good but... I’d like to do the numbers first.
SPIM08HP are the ones that I have to use in this. They are scary powerful but with a bms it should work fine
@@jgren4048 Should be fine. A 12v 12Ah SLA want less than 3.6A initial charging current so I'm certain it will be CC limited.
As far as I understand Pb chargers, They run in 3 stages, CC then CV and end on 'float' or equaliser mode. So just like Lipo chargers, they start with preset current till the set voltage is reached, then switches to constant voltage and the battery accepts what it can at that voltage till its full. Lipos stop accepting current if the cv voltage is reached while Lead Acids begin to hydrolyse at the electrode and boil off the electrolyte which can help with desulfation and equalises the voltage in each cell. SLAs need to be kept fully charged at all times, Lipos will last much longer being charged below their upper threshold. The float charge mode of a Pb charger shouldn't be an issue for the Lipo if its below its max voltage. Lead Acids do not like to be charged quickly but they can disharge insanely hard so the charger will be gentle.
What if we find the voltage divider which goes to microcontroller inside the ups for monitoring the voltage and change it then it will be fine
Most helpful lithium ups video I have seen.
Neither of these configurations is good. Besides the charge bands described, the charging profile must be considered, and a lead acid charge profile is not compatible with Li-ion chemistries. Charge controllers for lead acid chemistries charge in three phases: bulk, absorption and float. Of these, the float phase is particularly problematic for lithium chemistries.
Charge/discharge cycles from 4.0V to 3.0V are highly recommended to maximize the life of Li-ion chemistry batteries. Discharging below 2.5V severely compromises their life. Charging above 4.2V is downright hazardous, as well as life-reducing for Li-ion cells. Lastly, Li-ion cells don't like to remain at peak voltage for extended periods; the state where a lead acid charge controller will keep the battery. This, too, severely compromises Li-ion cell longevity.
Yup... Look up something called CC/CV chart.... says it all.... I have been looking at ways to convert my 24v ups to Li-ion. Perhaps if I could figure out how to find the charge unit and add a circuit between it and the battery.... but I'm still working on it. Maybe there is an external way such as a TP4056 for 7s xxp.... hmmm
would using a BMS solve the problem?
doesn't an over-charge and an over-discharge protection board solve this?
@@aleksandersats9577 Unless you know some other way, the lead acid battery charger does not give cc/cv charge and in addition it gives a trickle charge which Li-ion cannot deal with.... over/under is important but not the only thing we want...
@@KennyKenJm BMS usually just turn off and on..... not good for the battery
The 7S for 24V seems like a much better match if you ask me, because the UPS will float the cells at a little over storage charge. Would you really want a large li-ion pack sitting at 4.2V (or possibly 4.22V!!!) for months or years inside your house?? I sure as hell would not.
That is the conclusion I came to for a 24v UPS. For the 36V UPS, the charge voltage would be a little high using 10s so you could use 11s to be safe but you would only be using a little more than 50% of the cell capacity. However this would be great for the cell longevity and if you are using this only as a backup solution then it would be the best option.
I will test mine, I have one 12 v and one 24v. I think the LifePo4 batteries would make a better fit if my voltages turn out close to yours. I don't know yet how to set up bms or anything but those batteries are more forgiving than li-ion
Yes, it's is very often the case that LiFePO cell are a better fit from a voltage point of view.
I made a lithium ion battery for an ultima LCD
2KVA EURO ups The battery was as follows: 30 cell bak high drain 4.2 v 3 A 5. The batteries were placed in 5 columns and 6 rows and thick nickel stripe was used for spot welding. I used BMS 6S.The problem is that in few seconds, not more than a minute, the battery gets extremely hot and smoke begins to appear. Please explain the error.
Is 40 amp BMS enough for a 7S 24v UPS?
what about car batteries. 1 for 12v, 2 in series for 24v, and 3 in series for 36v. do you know if that work too? and IF yes, do you think there is a limit for paralell conections?
if u did not find the answer, car batteries will work but u must get ones that dont vent their gases out into the enviroment if ur using it inside.
a good BMS should stop any overages or even use protected cells?
Yes, but I don't have one.
@mwint1982 - During an Overcharge event a Common Port BMS will disconnect the battery. The UPS might not like that?
@@vtorsi610 UPS is pretty smart. I wouldnt worry.
@@LeafXpack18650 Buy one then...
I don't know if anybody's asked but did you measure charging current? An impossible what would that be
Thank you, You have done the hard work for me. Im going to build a 7s for my ups. I been trying to find a use for some 1.3ah high discharge 18650s. and I also need this used ups I bought.
Great video! Thanks! You did a video using LifePo4 instead of li-on ?
Isn't this issue exactly what a Balanced BMS is for (i.e. eliminates the potential overcharge in the 12V UPS)? Please correct this newbie if I'm wrong due to misunderstanding.
Thanks for this extensive test! Please can you test and clarify if the same conclusions apply also to battery packs including an internal BMS like the one from ecoworthy? These packs are sold as drop in replacement, but maybe the they can have the same problem of raw cells with UPS voltage. Many thanks for your work!
The difficulty of using the 3S config is: 1. You absolutely must have a BMS. While the low side disconnect is below that of the UPS shut off, the high side will absolutely cook a 3 s without it.
The BMS should disconnect charging... is the changeover quick enough for UPS?
Hi could you test a 60 volt and 72 volt please.
Can i use 29.4 for 24 v ups no damage in the board??
HOLY COW!! I just stumbled onto your video!! Been researching for the past week on the tripplite and you had one tested. Looks like 10S would work next to perfect for the 36V. Thanks for the video and testing!! The other 2 can be fixed or even the 36V can be fixed with a BoostBuck converter connected for charging. You would need to disconnect the charging the UPS does, itself or just never use the ac plug. Great work
@njfulwider5 Funny you mention a buck converter but choose to add a boost. I was thinking a DC to DC step down on the charge side of the 12v 3s, and 24v 6s configurations just before the battery connection. By stepping down the overcharge you can charge your lithium to full, and the UPS will reach min before your batteries, and you can still use the original charging system. Im just starting to learn a little electronics after being in IT for years. If there is some reason I'm missing which lead you to a step up like you did please share, but I thought it was more logical to use a stepdown as you retain the original charging and the UPS doesn't;t need 13.64v/27.03v to run, those are just the charge voltages for the original led acid batteries, the UPS will be quite happy at the 12.6v/25.2v. The only downside I see in this, is you are leaving some power in your batteries when the UPS cuts out. However, as a fix for the issue of not being able to use lithium (I keep saying lithium because LiPo batteries have the same nominal voltage and 18650s and it could be cleaner and cheaper to buy a pair of 6000mAh 3s LiPo's than a pack of 10 18650's and would open up a whole new world of applications for your battery with a HUGE increase in discharge output, you could say for instance boost a dead car with a couple 3s LiPo's in parallel, and I like the capacity of LiFe which is a lower nominal Voltage and you need 4s to hit 12v, but that's not an issue with what I learned in this video, I now know how to spec it out) for your 12v and 24v UPS's (which I assume most people have on their home computers anyway) it was worth bringing up a modification to your fix.
@@joeshedler6496 using a boostbuck converter, you can fine tune the charging, no matter the voltage coming in. A boostbuck are great when using a ups that uses 12v or 24v or even 36 if the boostbuck can go high enough. A boostbuck can take the minimum voltage difference that a boost or buck converter can’t do by themselves ( usually 1.1v). Hope that made sense...lol
@@njfulwider5 My fisrt experience with a buck converter was a boost. Im using a 3-30/3-30 to bring the 5v provided by USB up to 12v to run 12v led strips off USB power, well actually over driving them at 13v, lots of open air no heat issues and it gives them just a bit more oumph. The only thing with the converter, because its a boost, my output has to be 2v higher than than the input. I assumed the step down would work the same way with the range of input and a dedicated set output so the charger can do what it wanted, and you provide your battery with, ohhhh shit, I just saw the 2v difference needed, which means you only charge your lithium to 11.6, how sad. Damn, how gullible was I to think my 1 week deep dive into electronics was gunna win me some street cred on youtube with my brilliant suggestion. :P Damnit, if only the gap was wider like 2.1v, I think a buck converter step down (in product listings I see then called both buckconverter and dc-dc step up/down converters same product, wide range of input, and I think its an adjustable pot that lets you choose your output) would be the way to go.
@@joeshedler6496 you have to be real careful with the sellers out there. They’ll sneak in the word boostbuck or step up/step down in the title, just to get more sales. I fell in that trap a few times, and ended up returning the item because of that. There are 3 types of converter. A boost, a buck and a boost buck. Most of my projects I prefer a boost converter, because those usually handle amperage better than buck. Unfortunately the boostbuck converter handles amperage the same as a buck, because of the buck side of the boost buck converter. The boost buck converter takes away the limitations on the voltage side, making them great for ups. Have you been to the Facebook sites like (18650 battery) or (DIY powerwall). Great site I’m apart of that is a wealth of knowledge, when it comes to this. I’m apart of many many others and have my own channel on RUclips. Hope to see you there, you’ll love it if you like projects like this👍👍👍👍
@@njfulwider5 So, I'm a little unsure of what the advantage of a boostbuck over a variable on the fly input and user selected voltage on the output side dc-dc up/down converter is. What voltage limitations are there? Is it more than just a variable voltage on the input? If not, my dc-dc step up converters seem to be doing the same job your describing getting out of a boostbuck. Sorry, I wasn;t kidding when I said I just started in electronics. I know lots about power and some basic electronics llingo, which some of I actually understand, but I don't even know if Im at a highschool level. Im defiantly all theory, and no hands on at this point. An acute benign tremor in my hands kept me from finer solder work, so I pushed off electronics. But I have more time on my hands now and a few days ago I said I should build a drill press like machine to hold my soldering iron and fine points wont be a problem. then like 30 seconds after opening my mouth I engaged my brain and decided that was way over engineered, and I could get the same results from 2 boards, a hinge, and a spring. LOL. so yeah, Im finally doing it haha.
Why not use LIFePo4 battery? Seems have same voltage multiples.
Yes, LiFePO4 have an ideal votage range to replace lead acid.
Leaf Xpack thanks for the information i have a 24v ups the same one you showed in the video an i need to get some batteries when you say 7 cells that would be for one if the battery or both ?
7 cells in series will replace the two lead-acid batteries.
@@LeafXpack18650 oh ok thanks
@@LeafXpack18650 would I need anything to regulate the charging on the 7 cell or just hook it up on the ups?
@@Shiv.141 You could just hook them up and as long as they are well balanced, it will work. It would be advisable to use a BMS if you think of using the UPS in regular service. I recommend the daly BMS
@@LeafXpack18650 can I get that on amazon?
for 12V. . its better to use 32650 battery 4s config woth BMS. is that right?
Lithium Iron Phosphate (LiFePO4) in a 4s configuration is ideal, yes.
32650s are going to be slightly difficult if you try to fit it in the same space. 26650s may be better for you. If building an external pack, then it doesn't matter.
There are drop-in replacement LiFePO4 batteries with bms but it is critical that you get quality batteries. Many cheap Chinese batteries of Amazon have a bad bms and other issues.
BTW, you can do something else DO NOT plug in the UPS into the AC plug, BUT use DC power supply at 12.6v and connect it to the li-ion batteries and the wires between the batteries and the UPS.
Wouldn't it defeat the purpose of the UPS if you constantly need to manually charge it on DC every time it is discharged?
@@jmd.759 Not exactly. If you put in parallel the DC source with the battery then the UPS take the current from the DC source. If the DC source is absent (power cut) then the UPS take the current only from the batteries.
Thks I enjoyed/followed you video.
However I've heard lithium batteries can't handle trickle/float charging well.
?Is that the case?
As long as the voltage per cell remains within the appropriate range for li-ion batteries, then it is fine. This is why I did this video, to determine the maximum and minim voltages. Unfortunately, most UPS do not have a voltage range which is appropriate for li-ion. Lithium iron phosphate batteries work better.
Perhaps I missed it but you *really* need a BMS to monitor the voltage across each li-ion cells. Connecting up li-on batteries in series like this is potentially dangerous - dangerous in that you could have one cell have a thermal run-away event and cook off (catch on fire) and ignite the others. Sooner or laters li-on connected like this WILL get un-balanced (their individual voltages will begin to drift). Do a quick youtube search of have scary even one of these cells cooking off like that. The BMS board will protect against the individual cells from getting un-balanced - but its not a cure all for all risks. Even WITH a BMS board if I was doing this I would have the batteries in a NATO style 7.62MM ammo can - to contain the fire. And even than I would want to have that can located away from anything that might ignite.
You could wire a diode in serious and drop charge voltage that your batteries are seeing by one volt reliably. Could even do that with
can u test a 48v ? 4 battery of 12v? thanks
Unfortunately I currently don't have a 48v UPS. If I find one I will test it.
Do 10S BMS's exist? Can you actually replace 3 x 12v Lead Acid batteries with 10 x 3.7v Li-Ion Cells in series this way and even charge it with the UPS built-in power supply. Does it need some extra protection like a PMMT controller?
Yes 10s BMS's exist, just do a google search for daly 10s BMS, you can get them from banggood. You wouldn't need any additional protection as long as the UPS doesn't charge to much over 42v and your cells are well balanced, you should be fine.
What about the maximum current for a single 18650 is it Ok? is it smaller then the maximum current that go to the lead-acid batteries ?
If the Lead Acid is 12 Amp-Hours then divide by (approx) 2.5 to get 5 Li-Ion Cells in parallel, for equivalent energy
Now i have a Server Rack UPS with 48v battery voltage Lead Acid voltage range you could argue that Li-Ion 13S has an even better & tighter margin with a range of 35,75v - 55,25v compared to fit 38,8v - 55,2v used by the UPS. The BMS protects the Li-ION being overcharged with UPS Trickle charging. Why didn't you include a 48v UPS test?
I didn't have a 48v UPS to test with
Hi,
Is there any reason why companies are not producing UPS with Lithium Battery instead of lead acid battery?
Is there any issues for building large number of units?
$$$ I would assume. Lead-acid is much cheaper and handles the intended use-case for UPS systems: maintaining power output during short mains power outages. I want a UPS that can run for hours but most of the time, when the power goes out, it's only for several minutes at the longest.
@@Megatog615 You are correct, lead-acid batteries are dirt cheap. Also they like being kept at 100% charge, whereas li-ion do not. Manufacturers could design a ups to maintain a li-ion pack at 80% which would be ideal but again... $$$
Can we use LiFePo4 instead of Li-ion for 12v system
Yes, that would be best, with four cells in series, you obtain a voltage equivalent to a 12v lead acid battery.
@@LeafXpack18650 4S1P battery, ok then can we use 20A BMS ?
@@Bharani_0601 It depends on the size of your UPS and the load you will put on it, with 20A @ 12v you only get 240w.
Wait ..whi.not use 12v lipo battery pack
You don't need to worry about minimum voltage if the battery has an internal BMS with low voltage cut-off.
Well explained really good job enjoy your video.... I have a 12-volt APC ups and hoping to do 4S 10 P with a BMS to provide low voltage protection and charge balance
If you have BMS better try 3s not 4s, becasue for 4s you will charge up to 13.64v and this mean 13.64/4=3.41v which is basicaly around 0% of the 18650, see this table lygte-info.dk/pic/BatteryChargePercent/BatteryChargePercent.png
It's possible using bms for 18650
Yes it should work! At least when you charge your 18650 battery in a power bank with USB you use 5V from the USB and a BMS for the batteries and it works ….
@@andonrangelov8844 During an Overcharge event a Common Port BMS will disconnect the battery. The UPS might not like that?
I have the same 24v APC backup as in the 2nd example. I just wish I knew if or how to covert it to Lithium. Maybe someone could make a video: Converting for Dummies? Then, I might know how to create and install the Lithium batteries into the old housing of the Lead Acid battery packs. I dont even know if Lithium actually works in such a conversion? Is a bms neccesary? What other parts are needed? So many questions, so little information.😟🤦♂️🤷♂️
Using BMS may solve the problems
Aaaand this is why LiFePO4 cells are currently the best cells for drop in 12v replacements, with their 3.2v nominal voltage.
Thank you. This was great!
Even without the apparent expertise of Mr Coloe who commented, somehow I don't think it's as simple as what's presented. Lithium-Ion batteries have been generally available for almost two decades now. Since they have, what, 7x the energy density over lead-acid, why aren't lithium-ion UPS's flooding the market now? Makes an enquiring mind kind of suspicious.
Simple, lead-acid batteries are cheaper than li-ion. Furthermore, li-ion batteries need a BMS to be safe which makes them even more expensive. But probably most importantly, lead-acid batteries die after about 5 years and most people simply buy a new UPS.
They do make Lithium UPS, they are just much more expensive.
The lead battery dies because of old age not use. The li battery catches on fire because of poor charge design
Great explanations, i really wish i see a week earlier, as because i have just build 2 pcs of 3s for my ups. and find out problems.
I will just say this is how li battery fires get started. The charger circuit is not for li batteries. Your test are not enough to determine how safe it is to swap.
Agreed. It might be wrong to assume the top charge voltage with that method shown in the video because old batteries will have increasing resistance as the battery health decreases which will keep the voltage lower with some charging designs especially if they are undersized to keep manufacturing cost down. It's possible there is a bulk charge rate of anywhere from 14.2-15v per battery.
The other issue with lithium in series is that it needs to have cells balanced especially under high load with the application of a UPS and there is no balancing circuit in a lead acid UPS. If lithium cells come out of balance you will have an issue that will lead to fire.
That is, we need to use 3.2 Lifopo4 batteries for 12v UPSs.
how about connecting 2 diodes to the charging cable to decrease cell charging voltage by around ~1-1.4V, and 1 diode in the opposite direction not to lose much when discharging?
LiFePO4 is most suitable and works easily with all 12V, 24V and 36V. But also it must be paired with CC/CV dedicated charger (for example from separate wall/PSU connection) which also includes BMS for discharge balancing. These are not expensive.
If you do it assure that you have a HIGH QUALITY charger as charging Lithium Ion batteries can be very dangerous. The lithium iron you suggest is a much better suggestion. Lithium ion thermal runaway is no joke and improperly charged or discharged lithium ion is asking for disaster-just look at the Tesla fires. I'm waiting to see the havoc when the Power Walls start to ignite as they are one of the few not using LiFePO4 but rather go with lithium ion.
@@TVMADoc Yes, for other than mobile use (where weight/power density is important) the LiFePO4 is better choice. In future solid-state litium batteries could be the optimal alternative.
For example few days ago Nissan informed than their solid-state tech is moving forward and the larger pilot factory will be ready 2024. They said that 2028 the price should be 75 USD /kWh and few years later drop to 65 USD /kWh. This sounds promising, though we need to wait and see.
Of course even larger players - like Toyota - are working with solid-state batteries as well and they said them to be ready even earlier. But for those I don't remember seeing specific price estimations.
What if you use lifepo4 lithium chemistry?
lifepo4 cells have a voltage range which is ideal to replace lead-acid, you need 4 in series to be equivalent to a 12v battery
You do realize that you can use BMS to control under/overcharge?
Without the batteries when the BMS cuts off the UPS shuts off immediately. So unusable.
@@peterbalogh591 who said anything without batteries?)
1 wrong in reason, 24 v is better because not reaching abobe 85% on battery full charge makes Li-ion have more charge-discharge cycles .
For the 12v scenario, did you consider using 3 batteries in a series added to two parallel? You would end up with 9.625 / 14.7 if you really wanted to use your old 12v UPS...
I have a CyberPower CP1000PFCLCD UPS and noticed the life is pretty terrible. I have been thinking about doing a Lithium Iron Phosphate replacement, but the ready-to-replace batteries are crazy expensive. I found some IFR26650PC batteries which are very reasonable (max min 2v, max 4.2, recommended 2.5v & 3.65v tho) but don't know the min/max of my Cyberpower.
I guess I need to buy a voltmeter and figure out how to test my unit with it.
How about LiFePO4 ?
LiFePo4 is easy, you just need 4 in series to be equivalent to 12v. So 4s -12v, 8s - 24v, 12s - 36v, 16s -48v
Too bad APC doesn't just offer a drop in upgrade to Li Ion
Wouldn't be able to sell you overpriced motorcycle batteries then....
Also due to the high cost, lead acid is still probably the way to go for a lot of people. However in my case I can get some slightly fire-hazardous batteries for free, so I can get my dumpster dive acquired backup battery running for less than $20 and have insane battery specification that is completely and utterly overkill on power output and will be limited by the unit itself. Seems like a good deal, other than I really need to see if there is anything that can stop graphene fires...
BMS to te rescue
@Leaf Xpack, I just got a haircut for you, do you like it? LOL. So you mention that as long as we have a max voltage larger than the charge voltage of the UPS and a lower min voltage under the UPS's cut off we are fine to use the UPS to charge. Ok, I follow, and from the zoomed out look, that might seem fine, but when you look closer, you remember you had to build a 10s battery to pair up with a 36v UPS.... How do you address regulating cell voltages across the 10s pack? I wouldn't trust the cells to not fall out of perfect alignment voltage wise for many cycles. If it were me, I'd be buying, or building if I knew how, a 10s balance charge circuit that I would power with the UPS's regular DC charge power. Oh, and if Im adding components like that, I'd add a DC to DC step down converter on the charge voltage of the 12v and 24v UPS's so they can run on 3s and 6s packs :). HINT for anyone who does not want to build the 18650 battery packs you can sub in the same size premade (incl. balance board and connector) LiPo battery pack (same nominal voltage, look up min/max like shown here) which will give your battery a much wider range of use because LiPo batteries have a massive discharge current. Everytime I look, the LiPo's come out pre made and looking all pretty and cheaper than even just the 18650's never mind anything else you need to buy to make your battery, case, connectors, solder, heat shrink. Ok some of those costs are pretty nominal, but I added them just to nail that point down a little more. :)
To keep your cells in balance you'd use a BMS or active cell balancer. I don't think using a DC-DC converter would be wise as it would need to be bi-directional and would require high amerage on discharge. I don't even know if that would exists, and if so, it would be more expensive than the UPS.
Thank You
lithium ion batteries will not work for a 12v systems efficiently. but it i will work perfect for a 24 volt systems and 36 volt systems. 3/4 not good for 12v system, but 7s is perfect for a 24 volt system and 10s is perfect for a 36 volt system. on the other hand LiFePO4 cells will work perfectly on a 12 volt system since it has 3.2v nominal so 4s will give a nice 12.8v.
gooooood information!!
Why dont they just sell a lifepo4 ups replacement battery
Cost. Lead-acid batteries are dirt cheap compared to LiFePO4.
@@LeafXpack18650 , i think they are making more money from selling lead acid batteries. Ive bought 3 already for one of my ups.
Not to mention the non tech people who would just buy a new ups than bother buying a battery and then replacing the one inside.
I have 48v ups and 36v
You're supposed to subtract, not add (!) the 0,05V to the maximum allowable charge voltage.
lol, it says plus or minus. But ya he is losing a little bit of life hovering at the high voltage
Li-Ion = 2,75/3,75v, LiFePo4= 2,5/3,65v.?
You need to lookup the manufacturer's recommendations but in general:
Li-Ion full charge 4.2v, discharge under load 2.8v (most people, including myself, use 3v as a lower limit, especially when discharging slowly)
LiFePo4 full charge 3.6v, discharge under load 2.5v
For these A123 cells, the manufacturer permits a discharge to 2v but from the discharge chart, you can see that the cell voltage drops like a cliff past 2.5v.
lib.store.yahoo.net/lib/buddipole/ANR26650M1A-Datasheet-APRIL-2009.pdf
Just change the cutoff voltage and charge voltage and no problem.
Video is WRONG.
That assumes the backup unit is easily modified circuit or that a beginner is competent enough to reverse engineer the electrical traces and schematic. This one is probably a safer bet for most people, but yes if you can change those parameters and the inverter is still happy with it then that would be ideal.
The concept is all wrong, the charge profile was based on lead batteries but if you remove them the li batteries will change the charge discharge design. You also need betier than a multi meter to see the power profile. You only see the DC value .
Lifepo4 is a better replacement.
so much wrong and skipped entirely...
Haters will be haters. God damn troll..
Thank You