Failing Makita DIY replacement Packs

[90Ninety]

With my hobbyist experience with electronics and machines( fixing various computers, devices cars) I tried my hand at making battery packs but, I am failing pretty miserably and trying to understand what I am doing wrong here .

I have access to an abundance of 18650 cells due many being disposed at work, many from unused packs that fail to charge ( large packs used for automation robots) . Many of these cells 'appear' fine often charging to 4.2 and discharging with 2k mah+ in capacity

The first batteries I have tried building are for makita tools, since I already have several Makita 18V LTX tools . There are many of the Makita LTX DIY kits on Aliexpress often with thousands or hundreds of reviews, so how bad can they be ?

Links for DIY purchased

https://www.aliexpress.com/item/1005008257505052.html

https://www.aliexpress.com/item/1005008915967676.html

To test the cells I have, I put them through three charge/ discharge cycles to check the actual capacity. Using this, charging the batteries to 4.2V then discharging to 2.9V and recording the capacity
Charger/ tester linked here

https://www.aliexpress.com/item/1005006422977231.html

I put the cells into groups of similar capacity within 100-200mah . Perhaps this is my first mistake, how vital is it to have cells of the same capacity here? Should they be within 10 or 50 or 100Mah difference?

Anyhow upon first use the all packs work fine, but many after a few charges usually one or more cells just die or become too low to charge . Some packs hold charge but seem just to cut off until reconnected


Im sure there is something glaringly obvious I am doing wrong here . Perhaps there is a pad or jump connection needed on the Charging circuit?

Perhaps the cells I am using are unsuitable.. However I tried some used but tested Samsung cells and getting the same result .

Does anyone care to point out what could be the problem?

 

[Matt Gruber]

I've been using used Makita packs made in 2011, since 2013.
you have to look up the exact cells you have and find out their max discharge rate.
if they are hi-amp than i'd try NOT using the BMS. I have never used a BMS.

 

[90Ninety]

I've been using used Makita packs made in 2011, since 2013.
you have to look up the exact cells you have and find out their max discharge rate.
if they are hi-amp than i'd try NOT using the BMS. I have never used a BMS.

Hmm that is interesting . I had a generic premade makita clone pack without BMS and that works fine.

I will try to find the cell information later

 

[j bjork]

Hmm that is interesting . I had a generic premade makita clone pack without BMS and that works fine.

I will try to find the cell information later

"without bms and that works fine"
If it doesnt have a bms it will work until something brakes (and maybe catch fire)
In your case the bms:es seem to have done what they should and prevented a possible fire that might have burned your hose down otherwise.
It sounds like you just mix whatever cells you have with similar capacity?
That will probably not work very good in that case

 

[90Ninety]

"without bms and that works fine"
If it doesnt have a bms it will work until something brakes (and maybe catch fire)
In your case the bms:es seem to have done what they should and prevented a possible fire that might have burned your hose down otherwise.
It sounds like you just mix whatever cells you have with similar capacity?
That will probably not work very good in that case

Ah my house is not burned down, a miracle huh ? Never mind the hundred of laptop and phone batteries I have replaced over the years. I think if the pack was bad, it would have gone already, had it several years.

No the cells are all matching, of the same model Samsung ICR 18650-22FU . The only difference was the discharge capacity varied a little maybe 200 mah tops

 

[Matt Gruber]

22fu is only rated 4.4a max discharge
the Sony cells in real Makita packs are 20 amps.
not enough juice!

 

[90Ninety]

22fu is only rated 4.4a max discharge
the Sony cells in real Makita packs are 20 amps.
not enough juice!

Thanks

Also I have a lot of Sanyo UR3657M2 cells ( AKA UR18650-ZM2 allegedly? ) These cells are doing a similar thing, just cutting out during use

 

[sternwake]

The BMS cutting out during use is what happens when one of the 5 parallel groups falls below a certain voltage threshold.
or perhaps if one group's voltage falls out of line with the other 4 parallel groups by a certain amount.

I would expect this to happen on a power tool battery built with older, used cells of varying resistances, even if their supposed capacity winds up similar when discharged at some super slow rate which is completely un representative of how the battery will perform at much higher loads which it will experience in a power tool.

I just Built a 5s2P DIY Makita kit, using New high quality, Low resistance power c21700 cells.
BAK45D which is Tabless and rated at 60 amps to 80C and can almost do 40 amps continuous with no temperature cut.

I used 0.3mm copper busbars below the Kit's 0.2mm nickel plated steel busbars.

This of course required a huge power level from my spot welder, ~ 83% of its 14.46kW rating

I fully get the desire to repurpose used cells, but I would not ever mix and match cells from different ages, or brands,in a power tool battery.
Maybe this can be OK in a device which draws very little power and is used intermittently, but a power tool needs a power cell meant for high amperage discharges without getting too hot. All the cells need to perform similarly under high loads or imbalances will occur.

Your BMS is doing its job, preventing your cells from going outside a safe voltage range, and perhaps temperature range.

I am actually glad to hear your report of the Makita DIY battery Kit's BMS doing its job properly.
That was my biggest fear of getting one of the DIY kits, that the BMS was just a POS which would not prevent overdischarge or shut down if a parallel group failed, or got too far out of whack with the other 4.

I still need to test mine for the low voltage disconnect.
I found my stock 4.0AH Makita packs will allow discharge below 12.5v. I am not sure by how much as I manually stopped it there, and now stop discharging within the tool when I notice it slowing.

Unfortunately, the stock 4.0ah makita battery, used in my leaf blower was basically noticeably losing power the whole time as its voltage declined.

I built the BAK45D 9 amp hour 5s2P battery just for the leaf blower.
Unfortunately it does not seem to add any more grunt when both batteries are fully charged, in my specific leaf blower

It will however hold maximum potential grunt for far longer than the stock Makita pack ever could.

I'd urge not using any cells rated for less than 20 amps continuous for a 5s2p power tool battery, and I think in a single stack 5s1P, no less than a 25 or 30 amp cell.

 

[j bjork]

Ah my house is not burned down, a miracle huh ? Never mind the hundred of laptop and phone batteries I have replaced over the years. I think if the pack was bad, it would have gone already, had it several years.

No the cells are all matching, of the same model Samsung ICR 18650-22FU . The only difference was the discharge capacity varied a little maybe 200 mah tops

If I got what you said right, you built packs where one cell group would die after a while?
Then your bms prevented the charger from charging the pack?
Lets say the pack is 5s, and the charger charges to 21V resulting in 4,2V/cellgroup.

Then one cell gets an internal short, so that group stays at 0V.
If you didnt have a bms and just plug in the charger the remaining 4 cellgroups would have to take the 21V, ending up with 5,25V each.
That would mean a high risk of fire.
I am just pointing out the risk of "it would work if it didnt have a bms"

 

[Matt Gruber]

What happens in the real world, as my 2011 packs, 10 of them have proved, is that 1 cell pair loses capacity, and a BMS is not going to balance it anyway because it has low capacity. and yes, it will overcharge and the voltage will rise causing more damage as the remaining cells are undercharged. The Sony cells don't have a history of fires, not the ones from 2011. That is why I chose them in the first place . I use a 0-60 min timer and a voltmeter on the handlebar of my scooter. It is easy to spot a failing pair as the 10s2p voltage drops suddenly at full throttle. like from 34v to 30.5v so right there I see a failed pair. Then I patch the pair with a good cell from another pack. and balance by hand. and keep using the pack. Another clue of a weak pack is the charge time gradually drops from 60 min to 45 min and the voltage moves up too fast. So for me it is impossible to miss a problem.