Don't read if you like Headway Cells!

[Spacey]

This pack marks the end of me EVER using bloody Headway Cells again!

Edit to add: This pack has been cycled 8 times!! Had next to no use.

I built a 16 cell 48v Headway Pack for a mate, this pack uses 16Ah cells and a BMS from Headway Headquarters... link below...

http://stores.headway-headquarters....-bms,-16-cell/Detail.bok?category=BMS/PCM/PCB

upgraded to: SignaLab 30A Nominal 60A Cutoff PCM/BMS W/Led's

The controller is a standard 12 Fet Infineon from Cell Man along with the nice Mac Motor with the stronger gears.

Problem 1:

First the charging FET on the Signlab BMS was getting so hot that it was melting all the solder off, I changed it to a fet from one of the many broken BMS's that I had kicking about. That seems to work now, no heat at all when charging. Maybe the 12A charger was too much for the BMS? But it was boiling the solder at 5A charge as well?

Problem 2:

This is the main problem, the bike would run for no more than 2 miles before it would stop, pulsing the throttle would give you maybe 2 secs of power before cutting out. So the BMS was shutting off the pack for some reason.

Testing:

So I test each BMS sense wire, yep all working all voltages around 3.3 on rest. Problem is with LifePo4 cells you never really know how far along the discharge the cell is due to the voltage holding flat until discharged.

Anyhow I decided to take the whole pack apart and discharge each cell from full charge to 2.1V using an iCharger 1010B+. This charger can only discharge at a max of 7A even on a single cell.

Observations:

None....not one of the cells go past 15.2Ah on a lowly 7A discharge, this is lower than 0.5C. The Internal Resistance measured ranged from 27mohms to 33mohms.

One Cell which is Cell 13 only gives just over 13Ah at a 7A discharge! Question is would this cell at 35A be sagging down low enough to trigger the BMS?

Here is the Graph for Cell 13 form the iCharger 1010B+



Here is one of the better cells...remember these are 16Ah at 1C and these were discharged at 0.5C

 

[Spacey]

Cell 13 used to take a long long time ...over 8hours to come up to the same voltage as the others when balance. The LED light on the Signalab BMS would take over 8 hours for the LED to come on with that Cell and it was always the first to go out when the charger was disconnected.

Not impressed to be honest.

Is it possible that this cell would trigger the BMS at 35A or is it more possible that the BMS is broke?

How on earth do I discharge single cells at a higher rate than 7A and measure...not possible I suppose with RC chargers.

With regard to the 12A Charger, on Headway Headquarters website they sell the Headway 12Ah Battery pack with Signalab BMS...but offer a 12A Charger upgrade so the BMS can be charged at that allegedly.

 

[miuan]

The good thing about Headway is how easy they are to work on. Don't mess with the dud cell anymore. I suggest you buy a new cell and try to revive the battery. The others seem to be okay. 15Ah is still plenty. I don't even get 9Ah out of my 10Ah zippies. As long as they age evenly I don't care.
My mate has a 12Ah Headway pack and he is very happy with it, keeps going with zero maintenance or issues.

 

[Spacey]

But something is cutting out the BMS after less than 2Ah of use?

Trying to figure out if it Cell 13 that only puts out 13Ah @7A discharge....goodness knows what 35A would do to it?

 

[miuan]

I have a small device from HK called Turnigy Cell Diagnostics Meter
It can sense voltage lows and store them for review after ride.
Go for a ride and see how deep this cell sags.

 

[cassschr1]

Put a voltage meter on the bike ,( cell) and go for a ride. Compare with a known good cell also. Cheap easy and would work, Ya think?

 

[jdb]

This pack marks the end of me EVER using bloody Headway Cells again!


The Internal Resistance measured ranged from 27mohms to 33mohms.

At 3.0 V/cell nominal, a 30 mOhm/cell internal resistance would limit the short-circuit current to only 100A. That's only a pinch over 6C for a 16 amp-hr battery. I thought they were rated to a peak discharge current of 10C?

 

[Spacey]

This must be about the 12th LifePo 4 Headway based pack I have built and the 18th Ebike built to date, just really curious at why the pack is shutting off after not even a couple of Ah's.

Bit pissed off at the 15Ah's from a 16Ah cell at only 0.5C, yep they reckon 10C from these cells my arse. Couldn't imagine pulling 160amps from this cell lol.

Discharging the rest of the cells to see if I can find anymore duds...have ordered a spare 16Ah.

Then will put my Cell Log 8's onto the BMS sensor wires and see what is what, will bypass the BMS though first.

 

[999zip999]

Sounds like #13 is the runt cell kicking in the bms. looks like an extra cell is a must when building a pack, just like lipo from HK.

 

[jdb]

Do all of the cells fall in that 27-33 mOhm resistance range, or is it only the runt? And that is per cell, right, not per pack?

 

[neptronix]

Sad...
Let me tell you about one important variable though .. if you got a newer 12fet controller from cell_man and it came from the same batch as mine.. you could be drawing more like 55 amps. That's the peak & hill climbing amps i saw on mine.. end every wire along the current path ( battery cable, phase wire from controller, phase wire on motor ) got pretty damn hot..

Half C discharge rate and you're not even getting 16ah.. that is sad..

As for the discharge rating it's a joke.. at least ev-assemble has the decency to call these 5C cells.. in reality they don't have any lower resistance than other 1C lifepo4, they can just shed more heat because they're cylindrical and have a large surface mass.. that's the ONLY reason they are rated higher than 1-2C.

 

[Alan B]

New cells often require a few cycles to "wake up" to their full capacity.

It appears that you may have one bad cell. Better diagnostics are needed to verify this. The easiest thing might be to install a spare cell in the questionable unit's place and retest the pack. One weak cell is all it takes to make the whole pack look bad.

 

[Spacey]

That's the next step for sure, I do not think though that the 16Ah @ 1C discharge Headway Cells are magically going to find an extra 1Ah.

If they are hitting 15.1Ah out of 16Ah on average at 0.5C then how bad are they at 30 to 50A?

Neptronix: That is something I must check out, could be that it is tripping the runt cell with stupid amounts of amps causing it to dip below 2.4v and trip the BMS. It 90% feels like the BMS is tripping as you can pulse on and off....or wait 10 mins and get full flow again for 20secs before it trips again.

Going to hook up the HK Amp meter inline between the battery pack and controller...also run cell log 8s to see if they alarm. Unfortunately I didn't buy the ones that store data, have one on order though.

Will totally bypass the BMS to see what happens. 15Ah at 0.5C takes the piss, the amount of sag in these cells at anything above 1C should make it illegal to say they are anything more than 2C cells. 13Ah on the runt cell doesn't seem drastic enough of a reduction to be causing the BMS to trip...but this is at 7A and not the beasty amps that could be going into the controller.

 

[Doctorbass]

Spacey,

I think that the internal resistance measurement you made might be wrong due to the contact resiatance between the test lead and the cell tab that are often higher the internal resistance of the cell itself...

I have alot of headway from diffenrent sources and they are all between 8 and6 miliohm for the 10Ah model and 3.4 miliohm for the Power 8Ah model...

In other words, it's sinmilar exemple as if you try to power a 13A skillsaw thru a 100 foot 16 gauge extension cord.... the skill saw will not see the same voltage as the one that is present direct to the AC outlet... there is voltage loss in the extension cord... and i think that there is loss in your test leed between the Icharger and the cell... best is to use 4 wire kelvin measurement.. a seperate pair of wire to pull current thru the cells and the otehr pair of wire to measure the voltage across the cell ...
( as described here: http://www.cirris.com/testing/resistance/fourwire.html

25 to 30 miliohm is very surprizing... can you describe the way you connected the test lead to the cell ?

Doc

 

[Spacey]

Sure, I have 30cm of 12Awg Turnigy super flex RC cable going directly to the main + and - of the iCharger 1010B+ connected directly to the 16Ah Headway Cell using thick Copper lug connectors and screwed very tight.

Should be no loss of voltage, cables certainly do not get above room temp. They can't be off that much surely?

 

[Spacey]

New cells often require a few cycles to "wake up" to their full capacity.

It appears that you may have one bad cell. Better diagnostics are needed to verify this. The easiest thing might be to install a spare cell in the questionable unit's place and retest the pack. One weak cell is all it takes to make the whole pack look bad.

Just to add that I have spent the whole weekend individually charging up the cells then discharging at 7Amps the highest setting I can do ...and also documenting each cell by saving each graph which shows Ah's taken out of cell and Cell Voltage drop along with time.

Doing it as methodically as I can.

 

[oatnet]

I haven't messed with their 16ah cells, but 2.9v on a mere .5c discharge is way low for the 10ah cells I know very well. Logic would say you got a bad bunch of cells, or maybe there is a problem with your test setup. Here is Docs discharge curve, including the Headway at 3.5c @3.1v.

http://www.endless-sphere.com/forums/viewtopic.php?f=16&t=8926

-JD

 

[liveforphysics]

Sure, I have 30cm of 12Awg Turnigy super flex RC cable going directly to the main + and - of the iCharger 1010B+ connected directly to the 16Ah Headway Cell using thick Copper lug connectors and screwed very tight.

Should be no loss of voltage, cables certainly do not get above room temp. They can't be off that much surely?

Sadly, no RC charger I've ever used is remotely worth a damn at measuring an Ri below ~100mOhm. They simply don't have the hardware to make it possible.

 

[Spacey]

The iCharger I use has always been pretty accurate with the Ah's it puts into a pack. Seemed accurate with my Lipo's...did need the volts re calibrating on a couple of the balance channels.

I am basically just connecting a single 16Ah 3.3v nominal Headway Cell to 12AWG Turnigy wire with copper lug connectors straight to the main charge socket on the iCharger. Not much to go wrong really.

15Ah at 7A discharge is pretty weak to be honest and I just wonder if the cell that is only giving out 13Ah is Runty enough to dip below the cut off point on the BMS when 35A+ is drawn from it. Starting to doubt it could be the BMS at fault...although it did blow a charging Fet at 12 A charge...well within spec.

I haven't messed with their 16ah cells, but 2.9v on a mere .5c discharge is way low for the 10ah cells I know very well. Logic would say you got a bad bunch of cells, or maybe there is a problem with your test setup. Here is Docs discharge curve, including the Headway at 3.5c @3.1v.

http://www.endless-sphere.com/forums/viewtopic.php?f=16&t=8926

-JD

 

[liveforphysics]

The iCharger I use has always been pretty accurate with the Ah's it puts into a pack.

They are very accurate with Ah, far more accurate than you would ever think considering the cost of the device.

The only thing they can't do well is Ri measurements.

 

[auraslip]

I did a ballpark measurement of nominal Ri of a 4s lipo hardpack today by dividing the voltage drop by current. I used a 50w mr16 halogen 12v bulb.
As in .04v drop / 4.5a / 4s = 2.2mohm per cell
Right around what it should be. did it with a cheap multimeter too. not real accurate, but can probably pick out a dud cell real easy.

 

[John in CR]

LFP aptly named them Deadway's, so I wouldn't touch them unless they were free. Take the hit and sell them to a Deadway believer and get some true performing cells. Don't waste any more of your valuable time. You sound technically proficient, so forget the BMS, Battery Murdering System, in favor of a cell level low voltage alarm, using the BMS only on the charge leads if you must, or just invest in some extra capacity instead and run the pack more conservatively so no alarms are needed with occasional balance charging. That removes numerous potential points of failure, and I believe it's better in the long run. I'm convinced that as long as you don't just run your pack dead and monitor usage, BMS's kill far more cells and create far more other headaches than they're worth.

 

[auraslip]

I'd say the bms is the problem. headways are fine for ebike with more than a 30 minute run time.
Does anyone know who makes the signalab BMS. I'm thinking they're of decent quality and would be worth ordering some, if I can find them at prices equivalent to what I can get other, no name brand BMS.

 

[Doctorbass]

Thereis a way to modify most of the RC charger to accurately measure RI

Usually what make the voltage sensing is a little trace on the PCB of teh charger that goes to each positive and negative banana plug connector on teh PCB..

Just seperate these two voltage sensing trace and connect them thru another pair of probe that you DIRECTLY and INDEPENDENTLY connect to the cell tab.

I also recommand to connect a 100 ohm resistor i parallel between the sense and the current probe for each positive and negative

I wrote a thread about that in 2007-2008.. about my MEgapower 12s cahrger that i modified and that worked perfect for the RI emasurement.. and it overide the resistance measurement of the probe and just measure the true resistanc eof the cell..

 

[chvidgov.bc.ca]

I've got two packs of Headways, one with a Signalab 2.5 BMS (from Pingbattery) that I put on it, and one with the BMS from BMSBattery. I think the non-Signal one has better cooling on it. I wouldn't call my BMSes "battery murdering" so far, and they have both been fine for over 50 cycles. I also wouldn't say that the packs themselves are better in any way than my other two Ping packs, performance-wise - if anything, the voltage sag on equal loads is greater than on the Pings. I do like them for their modularity, and its nice to just charge and go, no hassles, no worries. On the right bikes, like a trike or a recumbent with low under-the-seat carrying racks, the extra weight over LIPO is not a problem. So far I've been satisfied, but not ecstatic, as I was expecting better performance C-Rate-wise. The BMSbattery BMS seems to balance the cells better than the Signalab - that's probably because the Signalab "likes" 61 volts on the charger, rather than the 58.4 I'm using with it, in order to balance really well. So, I agree that they are over-hyped, but still a nice cell for worry-free packs that can be easily repaired if a cell goes bad. On the other hand I've only been doing less than 50% DOD, so I've been avoiding taxing them, by running them for short-in city commuting or using them in parallel with a Ping on my trike.