My headway LiFePO4 cells keep dying

Which BMS did you get with the pack? We are finding that some of the 3rd party BMS systems Headway packages with their cells are junk and quit selling them too. The cells are all fine. The BMS boards are killing them. I also do not like any of the packaging I have seen directly from Headway. However the cells are still testing to my liking :-)

Mike

I have found it wise to periodically TEST the BMS (battery management system) to see if any section is discharging a cell that should not be discharging-that is, if a cell with equal, or lower, voltage to the other cells is being discharged, there is a BMS problem! How do you do this?--Simply use a voltmeter, and check for voltage at each cell balancing load resistor-these resistors are provided for each cell, and are normally of a larger size and lower resistance than the others on the board. Typical values are in the 3 to 75 ohm range, depending on the size cells the board is designed for. (Bigger cells, lower resistance, usually.)-I had an "Inductive Current Spike" cause a BMS to fail, and continue to draw down a lithium polymer cell, even though the cell was not at excess voltage level. Luckily I found the problem before the cell was destroyed.-You may need to add supression of such short duration, high-frequency voltage spikes to your assembly, doing so is a good idea anyway, as it reduces the possibility of Radio Interference both to anf from your circuitry. It is easy for most D.C. circuits to momentarily generate voltage spikes that are at least 2 or 3 times battery voltage, whenbswitching, pulsing, or disconnecting a load. Use of proper "Snubbing Networks" overvoltage varistors, zener diodes, current limiting resistors, bypass capacitors, RFI shielding, and suppression chokes/inductors can improve circuit reliability. (Don't be a "Runaway Toyota"!)-Bob Curry

I have had 12 of 48V20AH Headway packs for a few weeks now waiting to go into a car. From the time they were made until now is about 8 weeks and I measured every pack when I got them and about 3 days ago to ensure they were at least showing reasonable terminal voltage with no load.
All 12 measured 53.4 volts.
The BMS with them is a Headway BMS WL086-05.

So far so good.
The cells are fine but it sounds like your BMS is discharging and destroying them.

Did you resolve your BMS issues? I have been using the Signalab BMS V2 (SMD LEDS) with my 16s2p pack. Works flawlessly with about 75 cycles to date. I balanced the cells to 3.3 volts first (put all in parallel with a single cell charger). All cells balance within 1-2 minutes (LED's go (shunt) on and then off). I think I only paid $80 for the BMS.

Just like to add here that Ive used 10amphr Headway cells for 2 years now (without a BMS) on a few bikes and they are all fine.
I like Headway cells as they are easy to bolt together. I series charge and they dont get that out of balance without a BMS, and it saves all those wires, ocassionally I do manual individual cell checks to make sure they are level.

I add this here because if I was converting a Vectrix to Lifepo4 I think I would use Headway cells , they are only $18 each and so 40 would cost about $700.

p.s. Tip, heres a trick, cut the heads of 6mm bolts and use them to connect the headways in lines with no gaps or wires inbetween.
For 48v system have 4 lines of 4, or if you have the room 2 lines of 8. Put these lines in alternating diections and then you only need one short copper bar to connect the strips!

[>>>>>>>> +ve
[<<<<<<<< -ve

I got mine from hightekbikes. Ping may sell them as well, he claims to make them. I'd never run without a bms, the curves at both ends are way too steep to chance it. Besides, if you're canstantly monitoring and worrying obuot each cell your not running without a bms...YOU are the bms.

http://www.hightekbikes.com/index.html

I like the series connect idea...4 would be like one long12v cell.

Yes, when I first bought my cell's I was hoping to go "bare back" as well, but after bench testing I found that during charging the headway cells could go from 3.6 to over 4 volts in seconds....even after matching them prior to. The first 4-5 times I charged I'd watch them all blink into shunt and then turn off my charger. Been over 70 cycles now and no issues....does its job, cheap and you don't have to build it. Theres a schematic floating around too, some have upgraded the output FETS and lowered the shunts R to increase its 30 amp output to 80amps.

I agree, Haedway without a BMS is a bad move. I recently charged my spare cells and while all went to 3.65 V in about the same time, one happily went to 4V in about the next minute while the other two (only have 3 spares at the moment) really wanted to be pushed to go above 3.7V.
Getting the odd cell to destructively high voltage is all too easy. Cells are 38120S.

Ahhhhh that explains it, Ive been wondering why some guys are having fried cells and Ive never had that problem, You're not giving your cells enough leeway.
On my 48v bike which has 16 HEADWAYS, I series charge to 57.4V - Thats about 3.59v per cell.
Maybe every 5 charge cycles or so, I use a meter to check each cell voltage. If there are any that are lower than 3.59v after charging light goes green, I give them an extra few minutes with a single cell charger until they reach that.

Never had a problem.

This obsession with driving the cells to 3.7v would give you at most 2% more energy than stopping at 3.59v, but totally removes your margin of error - as correctly pointed out the voltage raises sharply from 3.65v upwards.

Incidentally 57.4v charge end works out well too because its the max my old 48v SLA charger was set to - so it really is quite easy!

Your not frying your cells because of a combination of luck and manual monitoring. You've been lucky to catch the low ones by monitoring them and then manually bringing them into balance with the pack....if not the low ones would force the higher ones over 4v and into failure....your pack would have been dead by the 3rd of 4th cycle. Your not runnng without a BMS...you just choose to be a low tech BMS. I couldn't be bothered with opening my pack and checking 16 cells and then single cell charging....perfer they work just like the LI batts in my cell phone and laptops. Plug in the charger and walk away knowing the BMS is doing the job. I use a 48v, 4 amp scooter charger intended for AGMS and signalab V2.5 BMS.

Not quite, because one low cells deficit will be averaged out over 15 others.
E.G. say theres a low cell 3.3v (never had one that far out) but even then this .29v difference will be shared out amongst the other 15 cells which would then read 3.61v (no where near 4v)
As soon as the low cells topped up the next series recharge will leave the all back at 3.59v again.

My cells are easy to access, but as an experiment I wont manually rebalance any low cells for the next few weeks and Ill report back how far out of balance they go. I use and recharge the bike daily so will feedback in a couple of weeks!

Don't get me wrong, if you don't mind watching your cells and balancing them manualy thats just fine... with great care and some luck you may not kill any cells. Just doesn't change the fact that Li cells WILL go out of balance with normal use and WILL fail if overcharged or under discharged AND because of this they REQUIRE a BMS of somekind (even if its a human one with lots of free time). Some of those less festidious do/will kill expensive cells or packs going "bare-back". Soldering 14 extra little wires ONCE doesn't take that long in comparison. Fortunately there are some low cost and high quality BMS systems out there now for us DIY guys. Unfortunatly, some of the junk ones (like Headways 1st BMS) left a bad taste for Lifepo4 and BMS.

that's because of unbalance impedance,if every cell with same impedance,they even don't need BMS either !

Like our LFP,it is very balance impedance,less than 0.03m difference

Thats not true at all....impedance will be no help if you discharge below 2v long enough. And even cells with matched Z's will change over time and load.

I see you posting about CHL battery but were can they be bought? Website? Nothing but talk it seems? Also, if you want to impress people, show them a 4 or 5C discharge curve. A static Z reading means very little....for example A123's are rated at 7mohm...they are 35C cells. You say CHL is .31mohm? What is the C rating? See, static Z means very little.

I just bought 26 Calib 40ah's.....if I knew more maybe I would have bought CHL? If these batteries really do 6C they would be very useful for Motorcyles in the 40ah version.

My cells are easy to access, but as an experiment I wont manually rebalance any low cells for the next few weeks and Ill report back how far out of balance they go.

Just be very cautious. As they go out of balance the difference will accelerate with each charge.
It would be a shame to lose some cells just to prove a point.

We did have a website,see www.china-huali.com

But my company are focus on china domestic market with a lot of government incentive policy on EV,especially that we are main supplier for biggest hybrid E-bus in Asia,Kinglong bus.

Export business just new start,so that our website don't have English vesion,pls be considerate.

We don't have machine to get 5C discharged curve,but our 40AH battery really under 5C big load of hybrid bus.And even 5kw to 8kw electric scooter testing,it is only very little drop on voltage when full throttle.

You welcome to start with sample testing !

@Johny - yes I will act with caution, will still monitor although I wont rebalance while Im doing the test. I think your web blog is very good, it has lots of clear photos - I want some of those battery terminal connector lugs and one of those mega crimpers! Whats the website for the Toolbox place you said you bought those lugs from?
I think I should return the words of caution with a warning about 600v systems - Dont you feel nervous working with voltages that can arc insulators??

@CHL, whats your price for 20ahr cells?
I bought some from Yesa last year and they are good (can do 5C without too much voltage sag) but they were quite expensive

We did have a website,see www.china-huali.com

But my company are focus on china domestic market with a lot of government incentive policy on EV,especially that we are main supplier for biggest hybrid E-bus in Asia,Kinglong bus.

Export business just new start,so that our website don't have English vesion,pls be considerate.

We don't have machine to get 5C discharged curve,but our 40AH battery really under 5C big load of hybrid bus.And even 5kw to 8kw electric scooter testing,it is only very little drop on voltage when full throttle.

You welcome to start with sample testing !

Sorry, but I can't believe a manufacture of batteries won't have 1. A website and 2. an active load capable of 5C. Does your company own a volt meter??

www.china-huali.com this is my website,why do you say no ?

If you don't believe,welcome to china or your friend welcome to china and confirm !

We have 1C discharge machine and I also post the curve on another topic. But very rare chinese lithium battery have 5C discharged machine,if you are professional enough,you can find the discharged rate from high power electric motorcycle which normally 4C or 5C on controller.

Ok, before you said CHL "did" have website. "did" is past tense, say "does". I have no problem with your english...as it is much better then my chinese!

I would buy (1) 40ah sample to try, you can PM me. But please also include price and delivery for qty 24. Right now Calib is best 40AH prismatic and cost is 1.25 per ah ($50 us dollar).

Whats the website for the Toolbox place you said you bought those lugs from?

http://www.thetoolboxshop.com/
They are in the UK and have pretty reasonable prices. Good from my point of view in Australia anyway.

I think I should return the words of caution with a warning about 600v systems - Dont you feel nervous working with voltages that can arc insulators??

Agreed. I never touch anything without disconnecting all the quick disconnectes. I'm VERY careful. I find that working on one pack's innards (48V) is way more stressful than the car itself.

Ok, before you said CHL "did" have website. "did" is past tense, say "does". I have no problem with your english...as it is much better then my chinese!

I would buy (1) 40ah sample to try, you can PM me. But please also include price and delivery for qty 24. Right now Calib is best 40AH prismatic and cost is 1.25 per ah ($50 us dollar).

Ok,to be a good start,Usd50/cell for one set (24cell) battery is not problem,where is your address ?

By air or by sea ?

Joe, lets wait until you post some 5c data. Also more info on your company, do you take credit card? how long for shipping, any referances? We want to see how your product and company compare with Calib & Winston? Calib now has a site in Ca, I ordered 24 cells last Wensday, they are due on tomorrow plus they take CC so it is mush safer then sending money to China.

You should start another tread...this was a Headway thread.

Hey reikiman,

While considering a DIY Headway battery project for my scooter after seeing yours in the original post, like you my main concern has been a reliable BMS.

Like many including you I have been following the Saga of the Fetcher/Goodrum BMS project for some time and I see that the long awaited final 'Zephyr' release is now complete and boards are available.

Your soldering iron hand getting itchy? Mine sure is.