• Howdy! we're looking for donations to finish custom knowledgebase software for this forum. Please see our Funding drive thread

Headway Pack Builders thread

dak664 said:
[
My case has holes drilled above each battery screw for insertion of voltmeter probes, or with the top and bottom covers off I can access the ends of the batteries directly.
Otherwise you could connect voltmeter + to a small wire and buzz it up the BMC connector, with the negative voltmeter probe connected to battery minus for safety. 56 volts won't give any protection against overvoltage if one of the cells is near full charge and the others are at 3.3

I have also taped one of each pair of fat blacks, since I don't draw more than 30 amps..
Click to expand...

Thanks. Couldn't get to anything last night. I'll report in tomorrow.

Thanks again,

Preston
 
dnmun said:
before you do anything to your pack you should make sure your charger is putting out a minimum of 58.4-58.8V or you may be chasing a non existent problem. you should verify what the voltages are on each cell when the pack is still at the peak of balancing, that may account for your problem if the charger doesn't charge up the low cell to 100% at 3.65V.
Click to expand...

dnum:

Thanks. I have been using an "Ebay" SLA charger up to now. I have a new LifePo4 charger coming today from Kenalten:

http://www.kenalten.com/chargers.html

I will check to make sure it is putting out the correct voltage (58.4 to 58.8).

When you say "peak of balancing" does this mean at the END of the charge cycle?

Thanks again,

Preston
 
yes, the BMS will shut down the charge when the high cell reaches 3.9V. i think that is the spec. but if the low cell doesn't get full, then it just goes more out of balance. this variation in final voltage is a function of the variation in the resistors in the bridge with the zener that switches the shunt transistor. so you need some extra voltage if the BMS doesn't have the tight range of balanced voltages. that is why they test them and adjust by adding the little parallel balancing resistors at the end. none of the other low priced BMS are tuned this way. this is a feature of higher price units. travis just got us a good deal on the BMS when he did the group buy. i think normally they would sell for over $100 for the 36V and $175 for the 72V through retail outlets.

the 4 headway BMS (3-48V & 72V) i have are my major stash. i am using the ping v1 signalab for the 4 36V packs i am building. they are less but not too much less. but a known design and reliable.
 
Just to add, the 20 ohm headway shunt resistors can handle ~180 milliamps so if the charger puts out less than that no cell will ever go to 3.9 volts. The resistors get hot at that current so once the first bypass becomes active I limit the charger current to ~100 milliamps using a series resistor. At that current that charger can stay on indefinitely and low cells will eventually reach their 3.65 clamp voltage.
 
Hi after a recent disaster with a duct tape lifepo4 that shorted inside and set alight I think the best option is to build my own lifepo4 from headway 38120s cells http://www.evcomponents.com/ProductDetails.asp?ProductCode=38120S I am thinking of 8 (perhaps paralleled with another 8 ) in series to start with and adding other groups of 8 later. I was thinking of the 24v KAE bms here http://www.kenalten.com/bms.html . I am presuming from the help on this board that I can at a later date add as many batteries as I like in parallel with the original 8 or 16 with some rewiring continuing with the KAE bms. It is to power a cyclone 360 motor. I am just checking with you technical wizards that I have the right idea. Thanks
Rob
 
it is actually cheaper to just buy a pack already assembled at the prices that evcomponents.com is asking for a completed pack. you can estimate the costs for cells, BMS, and connectors, then pay yourself 25 cents an hour to build it, and it is still cheaper, and the girls that build them do a better job to boot, imho.

i like the one that comes with the rack, but the best thing is that an assembled pack will get warranty protection in case something happens. if you wanna build something into the frame, then that is a different situation and building from scratch is how you have to go.

i attempted to use the piece of ground wire i found in my junk pile and the 10G lugs would not fit. turns out the ground wire is 6AWG solid, 4mm diameter, way fat.

i ended up buying a one foot joint of 8AWG solid bare ground wire and used that for my short (25mm) jumpers between cells since it fits the 12-10G lugs with a little slack that i can crimp before soldering.
 
I'm building my 30 AH bat now. Going to use GGoodrum bms, I have a diagram with no explanation. Anyone here care to give me an English translation on how to put it together(the battery to bms connections)?
 
I'm done testing the two cells that were in the cell group that triggered the LVC. One of them tested 100% good, I charged it up with my single cell charger, and then hooked it to an array of five 10W 1 ohm resistors in parallel and got 10Ah out of it, and the other one's looking sort of OK -- I got 8.5 Ah out of it before it dropped to about 2.2v, at which point I stopped the test. Might need a little more "break-in" (is this a real LiFePO4 effect, or is it a myth that came from imbalanced packs?), or maybe it's just a weak cell, but it's definitely not a complete dud.

Edited to say -- I must have made a mistake, taken that 8.5Ah cell off the charger early or something -- the 'weak' cell isn't weak, further testing showed it working just as well as any of the others. Not sure what happened there.

So... I don't get it. Why did the LVC trip when it did? How is it possible I only got ~9.5Ah out of a 20Ah pack, when the cell group triggering the LVC had a combined capacity of at least 18.5Ah, and the pack had been through at least 6 20-30% charge cycles?

I figure even if the cells were horribly imbalanced when I first built the pack, at least one cell in every pair would have to be fully charged before the pack voltage would reach 58.4 volts, where my charger shuts off. Yes?

And I would think after six 20-30% discharge/charge cycles even the lower-charged cell of each pair should have picked up several amp-hours, both during each charge cycle and just from the stronger cell charging the weaker one. So how did I trigger the LVC at ~9.5Ah? I don't think there's anything wrong with the LVC circuit -- the cell pair I tested were both pretty low when I removed them from the pack. Am I missing something?
 
My reasoning is that I could make a tidier pack with 1 bms and charger that is easy to check for damaged cells/connections etc. and could be added to over time with some rewiring. I know it would take more time but it does not seem too complicated with the screw top headways. Presumably to parallel the headway packs would need diodes. http://www.evcomponents.com/ProductDetails.asp?ProductCode=H2410P Would each pack need to be charged separately or could I make up a lead to connect the charging inputs of each pack to a common input? Also if I did get a dead cell in the pack it would need to extracted and welded back in which would be difficult.
I am a bit put off with packs because I now have a 6 month burnt out pack that I am not confident I will get my money back or repaired under a guarantee. It is Chinese holiday at the moment so though I have started to explain the problem to the vendor I am not getting any further forward till next week.

Rob
>>it is actually cheaper to just buy a pack already assembled at the prices that evcomponents.com is asking for a completed >>pack. you can estimate the costs for cells, BMS, and connectors, then pay yourself 25 cents an hour to build it, and it is >>still cheaper, and the girls that build them do a better job to boot, imho.
 
julesa said:
I figure even if the cells were horribly imbalanced when I first built the pack, at least one cell in every pair would have to be fully charged before the pack voltage would reach 58.4 volts, where my charger shuts off. Yes?

And I would think after six 20-30% discharge/charge cycles even the lower-charged cell of each pair should have picked up several amp-hours, both during each charge cycle and just from the stronger cell charging the weaker one. So how did I trigger the LVC at ~9.5Ah? I don't think there's anything wrong with the LVC circuit -- the cell pair I tested were both pretty low when I removed them from the pack. Am I missing something?
Click to expand...
Can you measure the voltages while charging in series? With the flat voltage/charge curve for LiFePO4 I wouldn't assume the cells in a pair are equally charged until ~3.5 volts when the voltage/charge rise is steeper. When the cells in a pair are fully charged the BMS will bypass all the current it can but if the charging current is higher than that the BMS will cut off regardless of the voltages on the other cells, which would then trigger the charger voltage cutoff. I made a pack with one fully charged cell and had to charge at <100ma to keep it from tripping the disconnect, until I clipped a 10 ohm resistor across it which allowed increasing the charging current to 400 ma.

Earlier you said the cell voltages were all 3.3 after charging...that should be more like 3.5.
 
Yeah I've bought three chargers now, two from ecitypower and one from Ping, and none of them seem to have a CV mode. All of them SAY they have a CV mode, but they just shut off the second they hit 58.4 volts... for example with the Ping charger, the current drops from 5A straight to zero. Maybe with the Signalab BMS it will behave differently,

I wasn't assuming they were equally charged, in fact with no CV mode I had assumed they were not equal. I just thought I shuold have been able to get something more than 10-Ah out of a 20Ah pack that had reached 58.4 volts. I hadn't measured cell group voltages while charging yet. Worth a look, thanks for the suggestion. Hopefully the BMS wasn't letting any cell group exceed 3.65V... :shock:
 
the BMS does not use 3.65V as the shut down voltage. it is usually 3.9V and the BMS will allow current to pass through the charging FET as long as every cell remains below that voltage but when any cell reaches that 3.9V cutoff then the charge is shut down by turning off the charging FET on the BMS.

so what happened to the chargers?
 
Whaaat? Jack at ecity told me this BMS would prevent cell voltage from exceeding 3.65v...

I still have (and plan to continue using) the ecity chargers. I will probably sell the 5A Ping charger once I'm done troubleshooting and have everything working properly on both my packs.

I already bought one signalab BMS from Ping to replace the DOA ecity BMS, but I haven't installed it yet. Maybe I need to order another one. Or maybe I should decommission all the BMS I bought, buy two Goodrum/Fechter BMS kits and be done with it. All this shipping from China is killing me, at least If I did that I think I could be confident everything would work properly and the cells would last.
 
i just got 2 different 5A chargers from ping. one is his type B with the plastic case and the fan, the fan which is big and noisy becasue it is so big. that charger went right to 60V and stayed there while charging my 5SLA test bank.

i also got the KP series of charger from ping which he is now selling too i think.

to relate ping's KP series to the ecitypower charger, this is the same charger that ecitypower passes off as their own but i think ping has modified it to continue charging and not shut off as in a lipo cycle which he talked about on the sphere earlier, but it is really the king power charger. mine is set to 60V exactly and is really very nice. very nice. i bet this will be the lux option battery charger. but still only 5A. i promise not to hack it up. yet.

maybe we can get your ping charger to work right for you.
 
Yeah, it's the one with the aluminum case... it is a nice charger. Yours goes into CV mode, eh? I'll try the Signalab BMS next... maybe that'll do the trick. Thanks
 
I think it is important to discuss surface voltage and charging voltage here. Anyone qualified to discuss the aspects of these concepts (coz i ain't).

I've found that the Chinese LifePO4 chargers are fairly inconsistent, had to return 2 before i found one that seems to work.

Another issue is some of the BMS's that we are using start shunting too late. Mine for example begins shunting at 3.7volts (this presumably is ~OK~ for the batteries with a max charging current of 3.65+-.05V) but because the charger is 5amps there is no way the shunts have enough time too bleed off the power. Therefore, the highest charged cell continues to rise. Then as said by others the BMS trips the charge as soon as a single cell reaches 3.9volts. Which is presumably the surface charge of the battery with a charge current coming to it (is this ok? im not convinced but would like to hear it from a battery chemist). I've noticed that once the BMS trips these cells come down pretty fast but i am still concerned with damaging my cells. and because i have a few high cells after everything balances I am at about 3.4volts per cell, which is not a full charge.

So, I think the best option is described by Dak664 earlier in this thread. By connecting a ammeter and several resistors in series with the charger you can limit the current flowing into the battery, thus giving the balance shunts enough time to bleed off the excess voltage of the highest cells.

It would be nice to have a BMS that starts shunting at 3.5 or even 3.4 or lower, right? that way the lower cells can come up to charge before the high cells trip the BMS.
It would also be nice to have a charger with an ammeter and pot so you could tune down the amps when you need to charge an unbalanced pack. It might take longer but it would give me peace of mind.

Somebody call China, or better yet, why can't we start building this stuff here
 
how do you know what you just said? how do you know the shunt transistor doesn't switch until 3.7V?

what BMS do you have?

why would you not fully charge the battery, that doesn't make sense.

why did you send the chargers back? what was wrong with them?
 
dnmun said:
how do you know what you just said? how do you know the shunt transistor doesn't switch until 3.7V?

what BMS do you have?

why would you not fully charge the battery, that doesn't make sense.

why did you send the chargers back? what was wrong with them?
Click to expand...

1: Asked my supplier he said he believed BMS Shunts around 3.7v (which makes sense and is similar to the 3.69v of the Goodrum/Fetcher BMS), I also checked individual cell voltage and felt the shunts for warmth.
2: I have BMS from KAE
3: I did not fully charge the battery because 2 cells were reaching voltage of 4.1v, I do not want to damage the cells so I stopped the charge, which makes perfect sense to me.
4: I have a 36volt system so chargers should be set for a top voltage of 43.8. One of the chargers read ~78volts while charging the other was at ~28 volts, because of issues with these chargers KAE is no longer carrying them from this supplier, He quickly replaced with a different model which seems to work fine. It is an alluminum enclosure and looks similar to the ecitypower chargers
 
A good way to measure the shunt trigger voltage is to put ~50 milliamps through a fully charged pack. All shunts should be operating effectively at that point so the cell voltage will be the shunt trigger voltage. On the 24 channels of my two 36 volt BMSs I think two read 3.70; all were within 50 millivolts of 3.65. (my meters have 10 millivolt precision and I calibrate them to 5.00 against a bunch of LM7805CT regulators which all read the same although they are nominally +- 200 millivolts).

I assume the Headway BMS just switches the resistor in at the trigger voltage. You could use more charging current if the shunt activated at a lower voltage so it would operate over a longer time. My ATTiny's progressively shunt at 10-30-50 ma.

If you can limit the charging current with a series resistor you can also clip a resistor across each cell as it comes up; for the first pack I balanced at ~350 milliamps and added 10 ohms across the charged cells which let the Tiny's handle any excess.

Sixteen 250-300 WH cycles later they are still balanced. I charge at 2 amps and the BMS has never kicked in since the charger stops at 41.8 volts, around 3.5 volts/cell.
 
Heres a good example, logged through an ATTiny infrared link I just got working. Three cells only, that's all the IR led's I have at present. It shows the cell voltage going above the BMS setpoint at 500ma, then dropping to 3.65 when the current is reduced to 150ma. As the other cells came up the current dropped to 50ma, where I raised it to 100ma to speed up the charge of the last cell.

The ATTiny's also have 50ma shunt capability and try to clamp at 3.5 volts, which is the reason for the blue drop and yellow plateau until the charge current was increased from 50 to 100ma.
headwaybmsgraph1.gif
 
please,
will someone make this clear to me ?
i have two 48v,10a headway packs,and i have been charging to 58.6.
i monitor the current myself,its a cc/cv -bench supply -
60v, 12a capability.
i havent used any diodes to parallel the pack-pair.
and i may change to a 16s2p ,when i find out what i
am missing..

in the first twenty cycles,i could notice the "balance" triggers as
charge - current "interruptions".
one bms still does this.the other ,slowly draws less and less .
eventually that pack's chg current goes to zero,or 10 ma...\.
i did notice the performance change,recently.
i dont know if i am doing anything wrong.
i hear of the bms having a "sweet spot" at 58.6.
then i hear the bms may balance fully when run at a voltage high enough to
cause "current shunting".
plz
"it sure works great"
 
I'm back and still having problems (I suspect) with my BMS. Again, my configuration is:

Headway cells (16s, 48V, 10AH)
Headway BMS from May group buy
Charger from KAE (4 Amps)
I have about 5 cycles on the batteries (all approximately 30 to 40% discharge ... 3 to 4 AH)

Charging problems:

When I try to charge the cells through the BMS, the charger just shuts down (says it is in the "fully charged state"). So, I have bypassed the BMS to charge the cells while monitoring the individual cell voltages. At the end of the charge cycle (charger shuts down at about an hour), most cells are around 3.5 to 3.6 Volts, but the first cell peaks up at 4 Volts. After 10 minutes or so, the first cell drops to 3.6 to 3.8 Volts ... and after 60 minutes or so, all the cells are in the 3.3 to 3.6 range.

Interestingly, the section on the BMS just above the connector to the sense lines gets warm during this period.

I simply cannot get the charger to charge through the BMS.

Discharging problems:

Same problem that Lessss described. I read 50+ Volts on the discharge connector (BEFORE it is connected to the controller) ... but as soon as I connect it to the controller, the voltage drops to 6 to 12 volts or so (not even enough to activate the CA).

So .... I have bypassed the BMS for discharge as well.

I'm watching my overall voltage during discharge and have never let it get below 44 Volts (under 3C, 30Amp load). When I take the load off the battery, the voltage pops right back up to 52.8 Volts.

I'm ready to pop for another BMS .... but I'd sure like to be able to confirm the one I have is bad.

Any ideas?

Thanks,

Preston
 
My BMS is on it's way back to KAE. I put in a note on the inside cover stating...bubblewrap is not an anti-static packaging material.

Personally I think sellers would do well for themselves to solder on the connection wires to the BMS and enclose it in a small case so that they don't have to deal with static kills due to shipping or end user soldering.
 
dak664 said:
Can you measure the voltages while charging in series? With the flat voltage/charge curve for LiFePO4 I wouldn't assume the cells in a pair are equally charged until ~3.5 volts when the voltage/charge rise is steeper. When the cells in a pair are fully charged the BMS will bypass all the current it can but if the charging current is higher than that the BMS will cut off regardless of the voltages on the other cells, which would then trigger the charger voltage cutoff. I made a pack with one fully charged cell and had to charge at <100ma to keep it from tripping the disconnect, until I clipped a 10 ohm resistor across it which allowed increasing the charging current to 400 ma.

Earlier you said the cell voltages were all 3.3 after charging...that should be more like 3.5.
Click to expand...

Hi Dak,
You're right, they're not at 3.5 because they're not charging all the way. I thought all of them were at 3.3, now that I think about it I might have missed checking one of the pairs or something. I guess it's time I started writing all my measurements down, I can't even remember whether I checked them all.

Ping told me that if I upgraded the charging FET on the Signalab V1 BMS, that I should be able to use my 10A chargers with his BMS, so I assumed it must be doing PWM, like the Fechter/Goodrum E-S BMS. This is next on my list to try.

I've only had a little time to play with this... but like you said, the most charged-up pair are briefly getting as high as 3.9V before the ecity BMS cuts the charging FET off completely, even with the 5A Ping charger. (sigh... b-bye Headway warranty on half my cells :roll: ). I assumed that the ecity BMS (looks like the exact same one KAE is selling) would start doing PWM on the charging FET to limit current once one of the channels reached the shunt voltage threshold.

Just out of curiosity, am I understanding you correctly that you're not seeing any real balancing happening with your Headway BMS unless your charger is limiting itself to less than 150ma? Seems to me that's not a very good way of doing things -- there's no way for the charger to know whether the shunt is activated on any cell groups... Shouldn't the BMS be doing the current limiting?

If the Signalab/PingV1 BMS does its own current limiting, then it seems to me it's the only good alternative to the Goodrum/Fechter BMS available. I guess I could balance manually and turn the charger voltage down a little but considering what I've already spent on shipping and BMSs at this point, I should have just bought these:
http://www.endless-sphere.com/forums/viewtopic.php?f=9&t=10758
 
You must log in or register to reply here.
Back
Top