diy solar

diy solar

Experiments with $40 'starter' lifepo4 for engine cranking

Vigo

Solar Addict
Joined
May 26, 2022
Messages
1,507
Location
San Antonio Texas
Recently I needed to replace an FLA battery in my little 600cc diesel Kubota tractor. I could have done the ~$34 thing and just bought another U1 riding mower battery from Walmart, but i wanted to turn it into an experiment instead. When i looked at lifepo4 options to use as an engine-cranking 'starter' battery, my prime suspect was a 12v 50ah model with a 50a bms for ~$100. But, I considered it dubious whether a 50a BMS would be reliable in terms of cranking an engine without tripping, and i was definitely not in the mood of spending $200 for a 100ah model and 100a bms that would have just, you know, solved the problem in one fell swoop. So instead of spending $200 for 100ah or $100 for 50ah i spent basically $100 on WAY LESS than 50ah of lifepo4 to test some concepts. Huh?

1702873360139.png
~$40 on amazon

So 'powersports' / motorcycle lifepo4 starting batteries on Amazon are getting down into the $40-60 range. They are 'rated' for 200-300 'cold cranking amps' in that price range, and are only 2-6ah of capacity! I assume they are simply some tiny but otherwise 'normal' lifepo4 cells hooked to a relatively monstrous BMS, and cycle life be damned (but still claim 2000-8000 because who will stop you!?). I have not opened one to verify. But I found a ~$40 2ah model rated for 200a and bought it. The model i bought, i found disappointing. My little tractor takes about 130-180a in current conditions (it was 70f today) and the battery would only crank it intermittently. I do not have an inrush meter but i know that if one were to scope a starter motor's current waveform it has an enormous but very brief spike before quickly dropping to 'steady' cranking amps. Unfortunately in this case, the bms electronics are fast enough to catch that and disconnect before the current drops off to the level the BMS is rated for. Even when the bms makes it across that jump and cranks the engine, I only get 1-2 good seconds of cranking before the BMS trips again. Put one way, it is WAY LESS of a cranking battery than the $37 lithium 'jump pack' i bought a couple of months ago. So, disappointing.

But i never really intended to put a 2ah battery alone in my tractor. I wanted a bit more margin for error than that, such as if i run the thing out of diesel and have to crank the engine to purge the air from the injection lines before it will start again. So I had the thought of paralleling a larger ah battery to the tiny motorcycle battery so that it would 'contribute' to some extent during cranking, but mostly 'add capacity' to the starter battery by transferring charge to it pretty rapidly, giving me many good long cranking attempts, only perhaps requiring some decent pauses in between.
1702873491528.png
Was ~$41 when i bought it
For this purpose i bought an 18ah battery for ~$41. I figured if this didn't work out it would go in one of the little electric cars for my 2 & 5 year old daughters, lol. I believe it has either a 15 or 20a continuous rated BMS. With this whole engine-cranking thing we are spending all our time not in the continuous rating, but in that 1s-5s duration rating that not many rock-bottom-money Amazon listings even mention. There is not even ROOM for a 200a bms in the 200a rated battery, so i believe the 'cranking amps' number these listings are using is really something at or under that 1-5 second rating we see specified on higher quality products. Anyway, I put a 100amp load tester to this 18ah battery and it consistently immediately tripped the bms. However, my theory for why the bms wouldnt trip during engine cranking was because my parallel connection to the other battery would be no larger than 12ga, which would 'throttle' the current sufficiently to stay in that 1-5 second rating region until the cranking was over. So, i hooked the 100a load tester through some 12ga wire and the 18ah battery did 75 amps for a solid 5+ seconds, and i lost my nerve and let off to see if my 12ga was getting too hot, before the bms actually tripped. So far so good!

Once the two batteries were in parallel in the tractor with 12ga wiring in between, it DID work out basically as i thought, with the caveat that the 2ah bms still occasionally tripped. When it didn't, the batteries cranked the engine fast and plenty long enough, and i didn't have the patience to see how many 'good long cranks' i could get out of it but it was definitely enough. I measured both batteries together to crank the engine fast at 180a with 75a of that coming from the 18ah, and the 2ah alone (18ah disconnected) cranks the engine a bit sluggishly at 133a.
1702873633939.jpeg1702873672333.jpeg1702873732076.jpeg

I consider my idea to be viable, with the exception that the 2ah 'starter' battery (really just its internal bms) is slightly undersized for this task. Im returning it (hey, it says 200 cranking amps and won't consistently do 130 without tripping) in favor of a slightly larger 300a rated 5ah model for ~$12 more, and keeping the 18ah. At that point i expect the thing to be fine. I didn't save any money over buying a $100 50ah model that might have worked out fine and had more capacity, but i verified some thinking about how this whole parallel balance between two different size BMS's and ah capacities would work out.

1702872867181.jpeg
Theres a pic of the two tiny batteries sitting in the battery tray of the Kubota B6100. Don't hate on my Renogy there, it's the cheapest ($16) SCC that is specifically negative ground. It's being fed by a 5w solar panel and the AC dynamo of the engine (after passing through that little rectifier block there). It's a whopping FIVE amp charging system at redline! Below half throttle it doesn't charge at all! Finally, if anyone is worried about low temp charging protection, I can assure you i live in South Texas and i will not be getting on this tractor when it is below freezing!:ROFLMAO:


Here is the replacement starter battery i am waiting on:
This one's rated 320a...
1702917332467.png
 
Last edited:
I don't think 2000 cranks of 5 seconds each counts as 2000 cycles, at 200A it is only 14% DoD of a 2 Ah battery. 2000 cranks is 280 equivalent 100% cycles.

Probably battery heating isn't excessive during that brief cranking either. FET would be, unless sized for it. But you can get FETs rated for hundreds of amps. I used one rated 90 amps for low voltage drop in a 2A circuit.

I wouldn't use lithium for my '64 Superhawk with unmodified circuit; it uses FLA battery as shunt regulator. I built an electronic shunt regulator and switched to AGM. Would want to be even more cautions for LiFePO4.
 
Thanks. This tractor has a super primitive charging system which i modified to use a PWM scc, so this pair of batteries would be getting charged by a ~0-28v ac dynamo going through a bridge rectifier (still somewhat lumpy DC, i assume) and then through the pwm scc to the batteries. I do have a ~$13 adjustable step down converter which could limit the rectifier output to 14.6v if i were to be convinced that the existing system was a problem. In all reality it's likely that the majority of charging would be from the 5w solar panel on board when the tractor is parked (solar and dynamo are paralleled to scc input). This is because while the stock charging system maxes at about 5amps when at max revs, when under half revs it does almost nothing and that's where i spend the majority of my time operating the tractor because i mostly use it for front end loader and box blade work that don't need the revs. There is a mower on it in this pic but i very rarely use it to mow either.

Im open to hearing whether my pwm scc is going to kill my lifepo4s.
1702917633710.png
 
I thought PWM shouldn't be used with lithium, but people here say they do it.

 
I seem to recall reading that even the extremely brief pulses of a too-high voltage (say 17v from my solar panel?) would gradually degrade the lifepo4. I guess the question then is 'how gradually?'. Also, is there 'leakage current' across the fets of the PWM even when they are supposed to be totally open circuit? I could see either having some effect over time. I have done some fairly high-current PWM charging of lifepo4 from a higher voltage source, but ultimately i havent even been playing with lifepo4 long enough for any of my experiments to result in a 'dead' battery from 'cycle loss', so in the absence of me running capacity or IR tests i have no data showing whether anything i've done has been detrimental at all.

I admit that one of the major appeals of lifepo4 for me is that it makes it possible to have so much 'excess' longevity that you can be 'sloppy' and not have your system fall apart in short order. As long as you stay within some limits (most of which are enforced by BMS's anyway) it's hard to really screw it up that badly. I own.. 13 pre-built lifepo4s at this point from 25wh (this tiny starter batt) to 3500wh (12v 280ah) and i haven't 'killed' a single one!

Larger starter battery is supposed to arrive today. Pending that turning out to be true, and me having time to install and do a bit more testing, i should have an update on that soon.

Not mentioned so far in this thread: I put a 'normal' 12v 120ah lifepo4 in my conversion van as a starter battery and have had NO issues so far even with fairly long cranking times (fuel pressure slow to rise..).
 
Im planning to eventually mount a full size car alternator out on the front of the tractor frame. There is a shaft from the crank pulley out to the hydraulic pump for the front end loader there which i believe i can add a pulley to and mount an alternator above or below the hydraulic pump. That would let me use the tractor as a potentially 1000-2000w diesel generator at a moments notice through my 12v system, but as far as the actual need to charge the onboard batteries, it spends so much time parked that i think the 5w solar panel on the tractor does more watt-hours of charging than the actual dynamo system does!
 
Interesting parallel battery setup. I wonder if the tiny battery even has a BMS? Have you looked into super-capacitor modules? You can get something that's like a 16V, 83 Farad module for $20-30. They're made for heavy bursts of current.
It would be kinda amusing if you could start the tractor with something containing 2Wh of energy.
 
The 2ah battery definitely DOES have a BMS because.. it trips! It goes to 0v and doesn't recover until you disconnect from it. In the tractor, cycling the key off and on would let it 'come back'.

I got the new starter battery, so maybe ill have time to install and test and update today? We'll see..
 
Ok, this project is complete. I received the $56 lifepo4 starter battery and installed and tested it.

So i have developed a fondness for this cheap little 100a load tester which can be had for $16 with a coupon.
At first id been skeptical of its usefulness but ive used a variety of larger adjustable carbon pile load testers on a bajillion FLA car batteries and ive never had the little one disagree with the big ones.

So this $56 battery handled the 100a load tester for ~5 seconds just fine ( i let off, the bms did not trip) and voltage dropped to the ‘200cca’ region on the gauge, which i can say from experience is the same region a $35 U1 fla ‘riding mower’ battery tests at. I have cranked this very tractor with those and it cranked very similarly, which is to say ‘a little slow’. Before embarking on this lifepo4 upgrade this tractor had a U1R in the battery tray and since it was ‘marginal’ in my estimation and i came into a newish ‘16’ powersports battery, i mounted that to the loader frame and paralleled it to the U1R. In ideal conditions either could start the tractor but i wanted more margin than that.
11264AD2-EC34-46F3-B231-C421413EB64D.jpegEB7381DC-BB2B-435B-AA60-110A414B6531.jpeg


Anyway, this $56 battery acted equivalently to a $30 U1 but hopefully with many times the lifespan. It could start the tractor on its own but it cranked much more comfortably with the 18ah paralleled in, which is the final setup.

I tested the pair of lifepo4 at 192 amps. Not EXACT same conditions as prior $40 battery attempt, but close and a slightly better result. Of that, 43 amps came from the 18ah battery, a significantly smaller proportion then before. When i disconnected the 18ah, the $56 battery alone cranked the engine at 168 amps.


7B21A1EB-CE80-4269-A495-8066E29163E8.jpeg
0956FDD6-5175-4B51-B49F-D8065EE8EC91.jpeg
BAB0CB6C-E5F3-4C58-B407-0B0B11780D72.jpeg

So, conclusions:
  • A $40 lifepo4 starter battery will not comfortably crank a 600cc diesel, but its pretty nifty for what it is.
  • A $56 lifepo4 starter battery will ALMOST comfortably crank said diesel. I would consider it viable but i wanted more margin, so i ran another 18ah non-starter batt in parallel.
  • A $40 18ah non-starter lifepo4 with a 20a bms would trip instantly at 100a but lasted a good 5 seconds at 75a and bms never tripped while assisting the other batt at 43-76a during engine cranking.
  • The 600cc diesel cranked slow @130a, acceptable @160a, fast @190a. From my personal experience as an ASE Master Tech this overlaps with the vast majority of gas 4cyl engines in cars, so my results would translate to those apps. Caveat is that letting a car alternator charge these small batteries would be an abusive C rate long term, plus the ‘regular’ downsides of lifepo4 in cars, ie caring more about low temp charging protection, overheating alternators at idle, etc.
  • Assuming the duty cycle is short enough to avoid overheating the wiring, you can totally ‘throttle’ the proportion of current flowed by two paralleled batteries into the same load, by the sizing and arrangement of the wiring between them.
Thanks for coming along.
A97DFBE6-CE32-4AB3-A2A1-CA95B23C2E62.jpeg
 
‘Market Watch’ on this silly little topic:

So powersports batteries have way too many letters in their group size/part number, but to keep it simple, i started with a group 4 battery for $40 (smallest powersports battery afaik), replaced it with a group 7 battery for $56, and just now bought this group 9 battery for $48:

320cca $48

I actually dont have a plan for it yet, but i may try paralleling this with the group 7 i have in the tractor and trying to crank some bigger engines with them to see if $100 of motorcycle batteries in parallel will reliably start a v8, or something like that.

After that there are sizes 12 and 14, and best price i see is $66 for a group 14 by Weize with a crapload of reviews (21,000!!!), highly rated:

360cca $66

It seems fairly certain that something like that $66 battery would have done fine on my little tractor without the slight added expense and wiring of the 2nd battery, but i would still be giving up a bunch of total Ah capacity which may come in handy some day, so i intend to leave the tractor setup alone other than perhaps borrowing the batteries temporarily to run other experiments.
 
Last edited:
So i received the $48 battery. I havent installed it in anything yet but i did run the 100a tester on it and it landed in the ‘600 cca’ region, basically ‘full size car battery’ territory! I held it for what to me was an uncomfortably long time (keep in mind im holding a tiny 1200w heater with no active cooling) and it didnt trip the BMS.

So i think at this point we can say the cheapest way to get ‘600 cca’ levels of cranking amps is NOT to buy a 600cca car battery for $100+, it’s actually to buy a 300-odd amp rated lithium motorcycle battery for 40-something bucks!! I mean imagine, two of these in parallel is cheaper than basically any lead acid car battery AND would crank harder than any of them, AND weigh less, AND potentially last longer, AND have more redundancy if a cell failed!

966ADF48-E742-4C56-9C4A-D41FFB60E159.jpeg
Granted, that doesnt address the issues of low temp charging, managing the alternator charge rate so it’s not a crazy 5-10c rate at basically all times, etc.

Still, going from a $40 battery that tests like a riding mower battery, to a $48 battery that tests like a FULL size car battery.. that’s pretty dang cool. And it certainly could have handled starting my little tractor all by itself.
 
I did some very abusive testing to the little $48 battery. I had it in parallel with two other large lead acids and hooked to a running car through jumper cables, repeatedly cranking on an old diesel tractor that hasnt run in years. The battery got so warm that the case inflated slightly.

After i disconnected the battery from the lead acids it was at 12.6v. Later, i put 48wh (out of ~60wh rated capacity) into it via a pwm scc, and it tests the exact same on the 100a load tester. So, i think i can say it survived the abuse. ?
 
Recently I needed to replace an FLA battery in my little 600cc diesel Kubota tractor. I could have done the ~$34 thing and just bought another U1 riding mower battery from Walmart, but i wanted to turn it into an experiment instead. When i looked at lifepo4 options to use as an engine-cranking 'starter' battery, my prime suspect was a 12v 50ah model with a 50a bms for ~$100. But, I considered it dubious whether a 50a BMS would be reliable in terms of cranking an engine without tripping, and i was definitely not in the mood of spending $200 for a 100ah model and 100a bms that would have just, you know, solved the problem in one fell swoop. So instead of spending $200 for 100ah or $100 for 50ah i spent basically $100 on WAY LESS than 50ah of lifepo4 to test some concepts. Huh?

View attachment 183617
~$40 on amazon

So 'powersports' / motorcycle lifepo4 starting batteries on Amazon are getting down into the $40-60 range. They are 'rated' for 200-300 'cold cranking amps' in that price range, and are only 2-6ah of capacity! I assume they are simply some tiny but otherwise 'normal' lifepo4 cells hooked to a relatively monstrous BMS, and cycle life be damned (but still claim 2000-8000 because who will stop you!?). I have not opened one to verify. But I found a ~$40 2ah model rated for 200a and bought it. The model i bought, i found disappointing. My little tractor takes about 130-180a in current conditions (it was 70f today) and the battery would only crank it intermittently. I do not have an inrush meter but i know that if one were to scope a starter motor's current waveform it has an enormous but very brief spike before quickly dropping to 'steady' cranking amps. Unfortunately in this case, the bms electronics are fast enough to catch that and disconnect before the current drops off to the level the BMS is rated for. Even when the bms makes it across that jump and cranks the engine, I only get 1-2 good seconds of cranking before the BMS trips again. Put one way, it is WAY LESS of a cranking battery than the $37 lithium 'jump pack' i bought a couple of months ago. So, disappointing.

But i never really intended to put a 2ah battery alone in my tractor. I wanted a bit more margin for error than that, such as if i run the thing out of diesel and have to crank the engine to purge the air from the injection lines before it will start again. So I had the thought of paralleling a larger ah battery to the tiny motorcycle battery so that it would 'contribute' to some extent during cranking, but mostly 'add capacity' to the starter battery by transferring charge to it pretty rapidly, giving me many good long cranking attempts, only perhaps requiring some decent pauses in between.
View attachment 183618
Was ~$41 when i bought it
For this purpose i bought an 18ah battery for ~$41. I figured if this didn't work out it would go in one of the little electric cars for my 2 & 5 year old daughters, lol. I believe it has either a 15 or 20a continuous rated BMS. With this whole engine-cranking thing we are spending all our time not in the continuous rating, but in that 1s-5s duration rating that not many rock-bottom-money Amazon listings even mention. There is not even ROOM for a 200a bms in the 200a rated battery, so i believe the 'cranking amps' number these listings are using is really something at or under that 1-5 second rating we see specified on higher quality products. Anyway, I put a 100amp load tester to this 18ah battery and it consistently immediately tripped the bms. However, my theory for why the bms wouldnt trip during engine cranking was because my parallel connection to the other battery would be no larger than 12ga, which would 'throttle' the current sufficiently to stay in that 1-5 second rating region until the cranking was over. So, i hooked the 100a load tester through some 12ga wire and the 18ah battery did 75 amps for a solid 5+ seconds, and i lost my nerve and let off to see if my 12ga was getting too hot, before the bms actually tripped. So far so good!

Once the two batteries were in parallel in the tractor with 12ga wiring in between, it DID work out basically as i thought, with the caveat that the 2ah bms still occasionally tripped. When it didn't, the batteries cranked the engine fast and plenty long enough, and i didn't have the patience to see how many 'good long cranks' i could get out of it but it was definitely enough. I measured both batteries together to crank the engine fast at 180a with 75a of that coming from the 18ah, and the 2ah alone (18ah disconnected) cranks the engine a bit sluggishly at 133a.
View attachment 183620View attachment 183621View attachment 183622

I consider my idea to be viable, with the exception that the 2ah 'starter' battery (really just its internal bms) is slightly undersized for this task. Im returning it (hey, it says 200 cranking amps and won't consistently do 130 without tripping) in favor of a slightly larger 300a rated 5ah model for ~$12 more, and keeping the 18ah. At that point i expect the thing to be fine. I didn't save any money over buying a $100 50ah model that might have worked out fine and had more capacity, but i verified some thinking about how this whole parallel balance between two different size BMS's and ah capacities would work out.

View attachment 183616
Theres a pic of the two tiny batteries sitting in the battery tray of the Kubota B6100. Don't hate on my Renogy there, it's the cheapest ($16) SCC that is specifically negative ground. It's being fed by a 5w solar panel and the AC dynamo of the engine (after passing through that little rectifier block there). It's a whopping FIVE amp charging system at redline! Below half throttle it doesn't charge at all! Finally, if anyone is worried about low temp charging protection, I can assure you i live in South Texas and i will not be getting on this tractor when it is below freezing!:ROFLMAO:


Here is the replacement starter battery i am waiting on:
This one's rated 320a...
View attachment 183709
Do you still have that 200 CCA battery you started this with? I am trying to find out what the max charging amps is for that. I want to combine it with a 16ah deep cycle to get under the battery limit on a school project (https://diysolarforum.com/threads/unique-application.74214/)
 
Hey there. The rule of thumb for charrge current is based on C rate, which is based on Ah of capacity.

So, for a 2ah battery, a charge rate of 2 amps would be 1c. 1 amp would be 0.5C. Generally charging above 1c is not recommended for lifepo4, but as far as i know the cost in accelerated degradation is pretty mild. Like, probably still a longer lifespan out of an ‘aggressively charged’ lifepo4 than a ‘pampered’ lead acid. So my .02 is simply not to worry about it in your application. The two batteries will split your 30amp supply and while a small 2-5ah battery would be getting charged pretty aggressively by its share of the 30a, it will probably have no effect in the lifetime you are expecting to get from it.
 
Hey there. The rule of thumb for charrge current is based on C rate, which is based on Ah of capacity.

So, for a 2ah battery, a charge rate of 2 amps would be 1c. 1 amp would be 0.5C. Generally charging above 1c is not recommended for lifepo4, but as far as i know the cost in accelerated degradation is pretty mild. Like, probably still a longer lifespan out of an ‘aggressively charged’ lifepo4 than a ‘pampered’ lead acid. So my .02 is simply not to worry about it in your application. The two batteries will split your 30amp supply and while a small 2-5ah battery would be getting charged pretty aggressively by its share of the 30a, it will probably have no effect in the lifetime you are expecting to get from it.
I wonder about the BMS shutting it down more than anything. We tried running lithium a couple years ago and I thought it was the older charging system just not wanting to work with newer lithium batteries with their higher voltages. Never really did figure out what the problem was but after a couple of years messing with this thing, I believe it was the BMS.
 
Got in our batteries
https://www.amazon.com/EBL-Motorcyc...C1VN5HS/ref=psdc_404722011_t5_B091NYZMDP?th=1
https://www.amazon.com/gp/product/B0B7MBRC69/ref=ewc_pr_img_1?smid=A2NKEN9O69YXYL&th=1

had them hooked up in parallel the in series to get 36v. The short time we were able to run, it seemed to run fine for the short time we were able to run it. We should be getting another kart from a different school to build on rather than our old half worn out one that we are running now so we will get battery arrangement swapped in and see what happens in the long run.
 
Nice, the price is right for sure. I am tired of going through lead acid batteries. They make them worse every year and the price keeps going up. Mild paranoia makes me feel like its being forced out.
Have you had any issues with the charge coil on lawn tractors messing with the BMS?
Perfect for summer use and remove for winter. The lack of cold tolerance isn't really an issue. I wonder what cells and BMS they use to handle the cranking.
If not lifepo4, I am considering a 6S LTO pack for my tractor. Just need to figure out how to make an old delco 10si play nice with it.
 
Got in our batteries
https://www.amazon.com/EBL-Motorcyc...C1VN5HS/ref=psdc_404722011_t5_B091NYZMDP?th=1
https://www.amazon.com/gp/product/B0B7MBRC69/ref=ewc_pr_img_1?smid=A2NKEN9O69YXYL&th=1

had them hooked up in parallel the in series to get 36v. The short time we were able to run, it seemed to run fine for the short time we were able to run it. We should be getting another kart from a different school to build on rather than our old half worn out one that we are running now so we will get battery arrangement swapped in and see what happens in the long run.
Had difficulty keeping batteries balanced and made mistake when trying to balance them resulting in a massive smokeshow and meltdown.

I had balanced them before several time but always inside our work area in the shade. When I did it again friday in the parking lot, I failed to disconnect from solar charger first and sent 30 amps to batteries.

They don't like that but nobody was hurt and I did learn what lithium batteries smell like just before they let out their magic smoke
 
Back
Top