Arkmundi's saga making & maintaining an A123 AMP20 battery

ohh n OUCH I cant imagine what a hevily bruised hip would be like except majorly crap. Glad you are okay and stoopid Dumb-ass driver

EVEN MORE THO... way to crash test the battery :twisted: Top construction mate well done.

Joe
 
winkinatcha said:
That don't mean that you wont be seen by ANYONE or NEVER given due courtesy, it happens all the time, but at no stage can you rely on it, and when you are lying on the pavement, your "right of way" is small consolation.
Joe
Click to expand...

So sad but so true... :(
 
Bummed out cell
Ok, to sum up some posts elsewhere, my pack died! The evidence is twofold:
One, My smart charger wont' go into charge mode - the red LED flashes at me, telling me its a no go; do some Voltsphreak single cell charging on the low cells & try again; lo' the smart charger starts, cycles through and completes a full charge! I'm thinking great, I'm good to go.
Two, I go, and at first all is OK, but the motor chugs out and I've got no juice. What the frellingfrack is that?
Bummer, yea, so out with the voltmeter to check out what cells are OK and which are not. Turns out there is just one bad guy, in position 6 on the pack (the middle).

Here's where the story turns out good. All my choices are proven in this scenario. Agnius' kit allows not only for easy construction, but also easy de-construction. Also, when I bought my cells from Victpower, I got 14, so 2 extra in my 12S pack. I was able to pull out just the one bad cell, without having to remove any others, and insert a replacement. Smooth as a wet pussy and feeling just as nice.
cell6.png

So, one more piece of information on that bad cell. It simply won't hold a charge. I can get it up to around 3V, but as soon as I remove charge current, it takes a dive, until its more or less fully discharged.

Poll - what happenned?
If you're into the A123 AMP20 pack line, I'd like your opinion on what went wrong.
  • A. The cell was bad from the beginning, just a casualty of the salvage that put these half-tab cells on the market through Victpower.
  • B. Without a BMS, and through usage, the cell became dangerously low and lost its ability to hold charge.
  • C. With the accident (hit by car and knoked to the pavement), the jar in fact caused damage to the pack - fortunately just the one cell.
  • D. With the bike in the shop for repair, and since I inadvertently left the controller on and Watts Up meter connected, the Watts Up drained the cell and killed it in the process.
  • E. An alien being that lives on nanophosphate li-ion fluff invaded and sucked it dry.
  • F. Something other
 
Yo Bro my 2-12s packs are well balance be one cell has a lag. I charge pack as 2-12s on my hiperion 1420 as 2-12s packs I charge the packs one at a time I have one cell lagging a little. I don't worry as I have 20ah pack and only take 12ah out. One cell afte charged will pulled after being off the charger is down to 3.54v to 3.48v or so the the rest sag to 3.54-3.359v. out of 24 so that one is showing signs.
 
I did the DRUTLEDGE pack build in the big moose thread and if careful it can be taken apart without viporizing a tab or a corner of one. So be careful because you will find out how live these cells are. Make sure it is of the same voltage. Do you have 6s balance wires ? There's a cheap 6s charger in the used thread.
 
A. Cells went bad.

Lessons learned, don't buy cheap cells. You may think you're getting a good deal but you're just wasting your time and money in the end. :roll:
 
i doubt if there was anything wrong with the cells you received. if the accident caused the cell to be damaged by impact or if something penetrated the pouch that might account for the problem. but if there was no way to stop the discharge as the controller and wattmeter ate up the juice, then it woulda dragged the cell all the way to zero. this is what the LVC function of the BMS does to protect the battery from accidental over discharge. lucky tommy did the same thing and lost several of his headway cells.

i have never had a cell go all the way to zero so i have no experience with that, but i have had headways at .25V and .41V from self discharge sitting unused in the box for 2 years. they came back to life in a second. have seen more in the .75V range and they were all ok, but now suspect and marked as such on the case.

you did not mention over charging, this is the only way i have ever killed a headway cell, from overcharging with a cell phone charger. for me it was overnight so there was a long time spent overcharged. i have also accidently charged a cell that got to 4.6V before i caught it, and it seems to be ok, but that was just seconds. also to 4.2V for seconds on another one, and it seems ok. but all now marked on the case so i know what the previous history was if the cell ever fails later.

so it could be from over charge or over discharge, no way to know which. don't know how high the voltage got when you finally recharged it.

i was not referring to you in particular about the people who refuse to use a BMS. i actually consider you more of a victim of misinformation from self styled authorities on batteries who have no real idea of what they are talking about.

i consider the thread someone started to deliberately mislead newbies into thinking they could manage the charging and discharging of their pack better than a BMS as the haven for those guys. but none of them really understand how a BMS works so it is just easier for them to be ignorant and play out their own testosterone poisoning in an environment that supports it.

you asked about which BMS i would recommend. the A123 cells can push so much current that it would be hard for the ping signalab mosfets to handle if high current is what you need. the old headway BMS would handle that much current and was designed for it. the new headway BMS also can handle that kinda current. the difference is in the balancing.

the shunt resistors on the old headway were 20 ohm so the shunt current would be about 190mA when the cells were at 3.8V and the shunt transistors were totally turned on. the new headway has 84 ohm shunt resistors so the shunt current is only about 45mA so they take 4 times as long to balance. but in normal use, the BMS does not usually have to spend a long time in balancing once all the cells are balanced to each other. they all charge up to the same point together if they have the same internal resistance which they do when new. so the balancing current is not so crucial.

the shunt balancing current is important when the pack is new and the cells have self discharged enuff that the new cells have differing amounts of charge remaining when assembled into a pack. this is why you have to take all the cell to be used in a new pack build and short them all together so that each cell in the batch has the exact same voltage as the others in the group. otherwise it takes forever to get the BMS to bring them into balance, so this step is crucial as the first step in assembling a pack. some people put the group of cells tied together on a charger and charge them together, but that is not really needed if you are gonna bulk charge.

if you need a higher discharge rate than the 40-60A that the ping signalabs can do, i would recommend the headway BMS. if you can live with the lower discharge rate, i would recommend the older v1 signalab from ping. it balances the pack to a slightly higher voltage/cell than the newer v2.5 signalab, and the shunt current is about 180mA versus the 60 mA of the v2.5 so the balancing can go faster and may be needed if your cells now have a large dispersion in the internal resistance.

the ping is cheaper than the headway BMS, but the headway will balance the cells to exactly 3.65V and the ping is not as exact, maybe 3.64-3.68V spread, which is still perfectly functional. the new v2.5 will leave the cells at around 3.6V after balancing when the leds turn off.

we are talking half the price. a lot of us have used signalabs too and you could put out a call for to see if you could find one cheaper than ping. shipping costs money too.

if power is your game, then consider the headway from jimmyD or you can actually use the ping for balancing and then discharge the pack without running the current through the BMS and burning up the mosfets. in that case you would hack up a solution where the output mosfets on the ping would sink the current from an LED on your dashboard that would be a simple warning indicator to let you know when the BMS had determined it was unsafe to continue discharging your pack.

it would not disconnect the power of the pack from the controller unless you decided to do another hack also to cause the mosfets to ground the throttle or brake signal to shut off the controller in the same fashion that richard's BMS designs do. this means you would have to know to shut down when the green led on your dash shuts off. it also would not protect against the accidental overdischarge from leaving the load attached to the pack. but that is one way to get a cheap balancing solution and cover the most important part of exercising the battery to the limits of its capacity safely.
 
dnmun said:
i doubt if there was anything wrong with the cells you received....
if you need a higher discharge rate than the 40-60A that the ping signalabs can do, i would recommend the headway BMS. if you can live with the lower discharge rate, i would recommend the older v1 signalab from ping.... the ping is cheaper than the headway BMS, but the headway will balance the cells to exactly 3.65V and the ping is not as exact, maybe 3.64-3.68V spread, which is still perfectly functional.
Click to expand...
Many thanks for your reply :!:
 
Bummed out cell numer 7
arkmundi said:
So, one more piece of information on that bad cell. It simply won't hold a charge. I can get it up to around 3V, but as soon as I remove charge current, it takes a dive, until its more or less fully discharged.
Click to expand...
What happenned for cell 6, happenned again for cell 7 - just won't hold a charge. So I repeated the remove & replace operation with my last remaining extra. The pack is back to holding a full charge at 42.94 volts, an average 3.58V per cell. But the question remains what happenned. After a charge on these two cells and with voltmeter probes on, you can visibly see the self-discharge as the cells capacity evaporates. Eventually they go to close to zero. What can cause this for the AMP20 cells? Anyone? Thanks.
 
For an update, I've lost 3 cells in my pack, all pretty much the same observation - they won't hold a charge. After charging, they just disipate and supprisingly quickly. I took a
Poll - what happenned? said:
  • A. The cell was bad from the beginning, just a casualty of the salvage that put these half-tab cells on the market through Victpower.
  • B. Without a BMS, and through usage, the cell became dangerously low and lost its ability to hold charge.
  • C. With the accident (hit by car and knoked to the pavement), the jar in fact caused damage to the pack - fortunately just the one cell.
  • D. With the bike in the shop for repair, and since I inadvertently left the controller on and Watts Up meter connected, the Watts Up drained the cell and killed it in the process.
  • E. An alien being that lives on nanophosphate li-ion fluff invaded and sucked it dry.
  • F. Something other
Click to expand...
I read this, a good article on li-ion battery safety: http://www.ainonline.com/aviation-n...hium-fires-generate-myths-and-misinformation.
The lack of charge protection circuitry in lithium-ion batteries for the experimental market is a problem, he added, because if a battery is over-discharged, then trying to charge it generates heat and can cause thermal runaway. And if a lithium-ion battery gets below 5-percent charge, it is probably damaged and won’t be recoverable. So, for example, if a pilot leaves the master switch on and kills the lithium-ion battery, it will have to be replaced.
Click to expand...
That along with the many friendly comments on my not using a BMS. So I'm goint with B. & D. above. 10 days with that Watt's Up meter, which draws a charge, albeit negligible and no BMS on the battery. The result was taking down some cells and not others, but below some threshold of cell tollerance.
 
i think that refers to lipo.

lifepo4 seems to withstand the discharge to low levels if the current draw is not very high and the cell does not end up reversed. current draw at low levels seems to be key. but like i said, .25V is the lowest i have ever seen them and that was only from self discharge.
 
Yea sorry to hear that bud. IMO the life cells are a lot harder to watch via a voltmeter or cell log device compared to lipo.

Lipo is more predictable and less sag , but with my liefpo4 cells I have a hard time telling where the pack is by just looking
at a voltage reading.. probably my in experience but I do like having that BMS to watch that for me.

you leaning any certain direction on the BMS ? how many amps do you need out of your pack ?
 
he should buy one of those used headway BMSs from you. surprised you would sell them. you can build a larger pack using two of the 12S headway BMSs in series, and you can do the same with the ping signalabs too. but i cannot hack a series build with the new digital 16S headway BMS. just the old ones.
 
dnmun said:
he should buy one of those used headway BMSs from you. surprised you would sell them. you can build a larger pack using two of the 12S headway BMSs in series, and you can do the same with the ping signalabs too. but i cannot hack a series build with the new digital 16S headway BMS. just the old ones.
Click to expand...
You're referring to ohzee having used BMS's? So, true? Yea, I'd be interested if they're in working order. Thanks for that.
 
Yea I already offered so did not wanna feel like I was forcing them down his throat.. Id let em go pretty cheap.
I ran them in series for a while with a nice diode from Ping to get 24s. I just got tired of all the wiring/connections.

I have ping sending me a 4s BMS now. Going to use 12 headway cells in 4s3p as my accessories lighting. I just love the BMS with lights even
tho they only really light up when charged.

I like my eva BMS also.. it's done exactly as needed.. I have only cycled that 24s pack about 4 times now so I only unplugged it the previous charge
so hopefully no damage to the cells. I did 40 miles on that pack one day with a good amount of pedal ling.. Got tired of waitin for the
BMS to cut off and switched to my 20s pack because I did not want to drain a cell too far. I am probably just paranoid.
 
you don't need a 24S BMS to run your 24 S pack as one battery. the diodes are not needed, you can use the two 36V chargers and the HVC for each section will control them and then you can use the CA to control for the over current during discharge if you want.

i have even figured out how to hack the headway BMS to less than the full series complement.

i just built a ping pack from parts that is 22S with a 12S section on a 12S signalab and a 10S section on a 16S signalab that i hacked down to 10S.

so you can use the two 12S headway BMSs you have now for the 24S pack.
 
How bad was the impact on the battery and case in the accident. I know all my batteries could be better protected. Like the one hanging on the rear rack.
Still care about you to Arkmundi's.
 
999zip999 said:
How bad was the impact on the battery and case in the accident. I know all my batteries could be better protected. Like the one hanging on the rear rack.
Still care about you to Arkmundi's.
Click to expand...
A real impact - the car hit my left hip and threw me & the bike to the pavement, enough that I was dazed and injured. My understanding about the AMP20 cells is that they need protection, casing, BMS, etc. In my 2nd conversation with A123's VP for consumer products (everything except automotove and grid storage), as I discussed our trade, ebikes, etc. and wanting factory access to the AMP20 cells, she responded that there's no way we'd build batteries that would last, that they do a lot of R&D to build batteries the "right way" to last, and that there needs to be vibration support. The cells can apparently be damaged internally. I'm assuming the anode/cathode layers can become seperated, or the tabs can loosen, etc. Something to think about. Possible my cell failure was from the impact. Or it was the salvage operation and they were doomed to fail eventually, just waiting for an accident like mine to send them over the edge. Just speculation. Anyway, thanks & warm regards. :roll:
 
My next battery, should it come to that will likely be an A123 retail direct 12V start/stop battery placed in a series/parrallel configuration to yied 36V and 15AH. Because, from beginning to end its all A123. Pricey at current retail $1200. But I'm hoping as the market picks up, A123 succeeds and production goes way up, the price will come down. :p
 
Sorry to hear about your situation ark.
 
arkmundi said:
A real impact - the car hit my left hip and threw me & the bike to the pavement, enough that I was dazed and injured. My understanding about the AMP20 cells is that they need protection, casing, BMS, etc. In my 2nd conversation with A123's VP for consumer products (everything except automotove and grid storage), as I discussed our trade, ebikes, etc. and wanting factory access to the AMP20 cells, she responded that there's no way we'd build batteries that would last, that they do a lot of R&D to build batteries the "right way" to last, and that there needs to be vibration support. The cells can apparently be damaged internally. I'm assuming the anode/cathode layers can become separated, or the tabs can loosen, etc. Something to think about. Possible my cell failure was from the impact. Or it was the salvage operation and they were doomed to fail eventually, just waiting for an accident like mine to send them over the edge. Just speculation. Anyway, thanks & warm regards. :roll:
Click to expand...
Sounds like they are trying to back off people buying these cells from China again. Internally i don't think there is something special about the structure. The only way layers could be separated in my opinion is by loosing vacuum. The tabs i think are quiet robust. I tried to scrape of excess teflon from them and it was quiet some work.
I'd say that those cells need robust outside shell to protect them mechanically and perhaps something in between, like aluminium sheets.
I read and was told that these cells don't like to sit flat and need to be on a side or vertical. I didn't bothered with my pack as it is only 12S.
Have you thoroughly examined those cells for mechanical damage? Do they feel hard or soft?

EDIT: I was just checking your first post and found a manual on building a pack using my kit. Thanks for that, that should took some time to put together. Some remarks i want to make thou.
First as i emailed you about cell placement regarding putting cells when markings face each other. This removes stress from tabs and keeps them strait because this way it leaves 6mm gap between cells hence 6mm block size.
Second i don't know how your pack looks at the moment assembled but keeping it without some sort of outer shell is no good (last pgoto on the manual). Even 1mm aluminium sheets cut to size and placed on top and bottom, than whole pack wrapped with duck tape or heat shrink tube would greatly improve packs protection. Further more in my opinion battery pack has to be attached to the frame firmly this way bikes suspension eliminates vibration if that would be the case for failure which i doubt.
There are lots of videos on A123 assembling modules and except thermal management i don't see there something sophisticated that us can't make.
 
You must log in or register to reply here.

Similar threads

CUDAcores89
Found some A123 AMP20M1HD-A cells at a scrapyard today. How to take care of them?
Replies
9
Views
1,561
999zip999
999zip999
themadhatter106
Biggest Battery Fail Ever - A123 AMP20
Replies
15
Views
2,229
themadhatter106
themadhatter106
R
Making a Battery Pack using NCR18650
Replies
10
Views
2,965
999zip999
999zip999
themadhatter106
20S A123 AMP20 Battery Build
Replies
4
Views
931
themadhatter106
themadhatter106
granolaboy
My 4s2p A123 Amp20 Battery Build
Replies
24
Views
5,941
999zip999
999zip999
Back
Top