The care and feeding of a123-based packs...

[GGoodrum]

Okay, the new clip springs seem to be just the labor-saving ticket I was looking for, and they seem to be able to provide a better mechanical contact. What I did was use both the smaller and larger springs together. This allows the smaller spring to be centered on the center of the cell, and will be used for the negative/button end of the cell. The larger spring is then used for the positive end. It is a bit of a tight fit, with the two together, and then with the copper "hanger" tape in the middle, but it makes for a very solid and positive mechanical connection.

Here's a few shots of whatthe new configuration looks like:

I'm going to try making up a new 16s5p pack, using this new method, and see how it goes, but so far, it looks very promising.

-- Gary

 

[Doctorbass]

Okay, the new clip springs seem to be just the labor-saving ticket I was looking for, and they seem to be able to provide a better mechanical contact. What I did was use both the smaller and larger springs together. This allows the smaller spring to be centered on the center of the cell, and will be used for the negative/button end of the cell. The larger spring is then used for the positive end. It is a bit of a tight fit, with the two together, and then with the copper "hanger" tape in the middle, but it makes for a very solid and positive mechanical connection.

Here's a few shots of whatthe new configuration looks like:

I'm going to try making up a new 16s5p pack, using this new method, and see how it goes, but so far, it looks very promising.

-- Gary

What is great is that you can remove every A123 cells from the last that used spring and put these cells in another pack desing.. without the need of unsolder or unspotweld.. that is nice!.. just like AA battery in a old RC futaba remote control using tons of them..

Doc

 

[Roy Von Rogers]

You may also be able to put a dab of solder on the copper and clip, to keep it on place.

Roy

 

[swbluto]

Can someone explain to me how you would take 3 dewalt pack's worth of a123 cells, put them into 3-cell parallel groups, and then string the 3-cell parallel groups in series while using the original housing(So, get 10s3p)? It seems that the original configuration was meant for a string of 10 cells in series, so I'm not sure how you're getting each cell into a parallel configuration with a unique single cell from each pack.

 

[dnmun]

5P

 

[swbluto]

Yes, I can see Goodrum's 5P handiness with the original case. Is that the only known parallel count that's easily adaptable? Ideally, I want 3.

 

[MitchJi]

Hi,

Yes, I can see Goodrum's 5P handiness with the original case. Is that the only known parallel count that's easily adaptable? Ideally, I want 3.

I think you could easily get 4p if you used the end caps from the 28V Dewalt packs. They have 8 cells. That would only require an extra 10 cells but you might have a hard time finding the 28v end caps (maybe Doc). If the extra space isn't an issue you could use 8 cell end caps and only put in 3p worth of cells and have a pack that's easily expandable to 8p. You could do the same thing with 10 cell end caps (only put six cells in each 10 cell cap). I like the idea of an easily expandable pack.

I don't think it would be that hard to trim the 10 cell caps down to 6 (just cut the ends off). You could use tape or cable ties to keep the cells from falling out the cut-off end if necessary.

You could also figure out a way to put nine cells (3 x 3 x 3) in the 10 cell caps. It would be a little more time consuming but if you are only building one pack...

Someone recommended using a Dremel to cut the tabs to separate the cells. Is a Dremel the easiest way to cut the tabs?

Is it best to remove the tabs or just cut them to separate the cells? If removal is better is a Dremel the easiest way to do that?

Thanks!

Mitch

 

[CGameProgrammer]

The cells in the 8-cell pack are 18mm diameter, not 26mm diameter. They're 18650 cells and are only 1.1 Ah. So you can't use those end caps for the 26650 cells.

10s3p can be done with the original plastic holders, by taking three 10s packs and running 11 insulated wires between them to connect the cells in parallel. I did this originally, though without the plastic holders. The advantage is you keep the tabbed batteries connected together and don't have to do so much tedious wiring. But it's not as 'elegant'.

Mitch, never remove the tabs! You can cut them to separate cells, using a Dremel (it's very quick and easy), but you definitely don't want to actually remove them because it's them that you solder to.

 

[webfootguy]

The cells in the 8-cell pack are 18mm diameter, not 26mm diameter. They're 18650 cells and are only 1.1 Ah. So you can't use those end caps for the 26650 cells.

Actually, this is false. The Dewalt DC9280 pack (28 Volt) uses quantity 8 A123 M1 cells which are 26mm in diameter (same as DC9360). I should know, I have 56 DC9280 packs I bought for building my A123 based pack. You may have been thinking of the 18V pack (DC9180) which has 12 1.1 Ah A123 cells.

 

[MitchJi]

Hi,

The cells in the 8-cell pack are 18mm diameter, not 26mm diameter. They're 18650 cells and are only 1.1 Ah. So you can't use those end caps for the 26650 cells.

As posted by Webfootguy the 28V packs use 8 M1 cells and the 18v packs use 18650 cells (I think it might be 10, not 12).

10s3p can be done with the original plastic holders, by taking three 10s packs and running 11 insulated wires between them to connect the cells in parallel. I did this originally, though without the plastic holders. The advantage is you keep the tabbed batteries connected together and don't have to do so much tedious wiring. But it's not as 'elegant'.

I didn't realize that would work. Good idea. Thanks!

Mitch, never remove the tabs! You can cut them to separate cells, using a Dremel (it's very quick and easy), but you definitely don't want to actually remove them because it's them that you solder to.

Thanks for the Dremel vote. I'm not planning to solder the cells or the tabs. Planning to use the end caps in a very similar way to Gary's excellent idea. I thought the springs/clips might make a better connection without the tabs (directly on the end of the cells).

 

[dnmun]

the "end of the cell" is probably a metal cap and the battery cell tabs are attached to a disc underneath that end cap. in physical contact from the forces inside the the cells. when you peel the tabs off, each of the welds has to tear off and it pulls that cap up and away from the disc underneath. plus you have to worry that the weak spot where the welds contacted the end cap would leak. lotta pressure inside i would expect.

 

[MitchJi]

Hi,

the "end of the cell" is probably a metal cap and the battery cell tabs are attached to a disc underneath that end cap. in physical contact from the forces inside the the cells. when you peel the tabs off, each of the welds has to tear off and it pulls that cap up and away from the disc underneath. plus you have to worry that the weak spot where the welds contacted the end cap would leak. lotta pressure inside i would expect.

Thanks for the reply!

Someone told me its not hard to use a dremel on the welds to remove the tabs (no pulling) and smooth off the surface of the cell.

My questions are is there an easier way and is it worth the time?

Since (so far) everyone thinks its a bad idea, I gather leaving the tabs in place, has not been a problem so its probably not worth the time. I will probably test this on a dead cell just to see how much time it takes and how well it works.

Thanks Again!

Mitch

 

[leamcorp]

I've pulled all my tabs off and they are fine - all 150 of 'em. The end plate on the positive side do come up or get lifted a bit but I'd just lightly tab it back in with screw drive handle.

 

[GGoodrum]

I've done both, left them on, and peeled them off. I prefer off. What I do to get them off is use a pair of needlenose pliers, grab a "corner" and just roll it off. Kinda like opening an old can of sardines.

-- Gary

 

[MitchJi]

Hi,

I've pulled all my tabs off and they are fine - all 150 of 'em. The end plate on the positive side do come up or get lifted a bit but I'd just lightly tab it back in with screw drive handle.

Thanks leamcorp! But I really don't like the idea of putting that much pressure on the cell end cap. I'd rather just leave the tabs on. Thats about $1k worth of cells that have potential problems.

I've done both, left them on, and peeled them off. I prefer off. What I do to get them off is use a pair of needlenose pliers, grab a "corner" and just roll it off. Kinda like opening an old can of sardines.

-- Gary

Thanks Gary! Quick and easy and very little if any tension on the cell cap.

 

[stator]

Here is some interesting reading for anyone considering building A123 packs.

http://www.evworld.com/library/prius_fire_forensics.pdf

In particular, the "Battery Pack Design Safety Guidelines" from A123 systems in Appendix D recommend not building packs as described in this thread.

Never use a clamping force at the top and bottom of the cell or hold cells together,
end to end, in a way that restricts the cell rupture vents at the ends of the cells. If
the vents are blocked, the gas can’t exit the cell in case of cell failure.

Regards,
--Bill

 

[GGoodrum]

I agree, "hard-clamping" them is bad, but they are fine if battery springs are used, especially on the positive end, which is where the vent is. I recently had a pack where one 4-cell parallel block of cells had three of the four cells vent. I had left off my LVC board and had this block not charged all the way (new pack...). I then proceeded to try and pull about 100A of it and eventually the 3 cells vented. when they vent, it is on the outer portion of the cell's positive end.

 

[Doctorbass]

Here is some interesting reading for anyone considering building A123 packs.

http://www.evworld.com/library/prius_fire_forensics.pdf

In particular, the "Battery Pack Design Safety Guidelines" from A123 systems in Appendix D recommend not building packs as described in this thread.

Never use a clamping force at the top and bottom of the cell or hold cells together,
end to end, in a way that restricts the cell rupture vents at the ends of the cells. If
the vents are blocked, the gas can’t exit the cell in case of cell failure.

Regards,
--Bill

Thanks for the article!! VERY INTERESTING !!.. it include all great recommandation to be by the law with A123 requirement for safety... very great find!!

Doc

 

[stator]

In addition to the clamping issue (and even the spring force may increase the amount of internal pressure required to rupture the vent according to the report), there is the issue of no fuse links between paralleled cells. From the report:

In parallel assemblies, a short from one cell to another cell in
the next series string causes the energy of all the paralleled cells to flow through the one
cell and into the short. Cell-by-cell fusing prevents that from happening by separating the
shorted cell from the rest in the group.

Just things to be aware of...

--Bill

 

[leamcorp]

Anybody not using LVC or BMS with these cell?

Because of my lazyness and space requirement, I've been running it without any protection other than CA. Granted, I only use 5ah out of 18ah each day.

I've just re-made the single large pack into 2 smaller packs and while it was loose, i've check each cells and they still match the original voltage. Nice.

 

[biohazardman]

Anybody not using LVC or BMS with these cell?

Because of my lazyness and space requirement, I've been running it without any protection other than CA. Granted, I only use 5ah out of 18ah each day.

I've just re-made the single large pack into 2 smaller packs and while it was loose, i've check each cells and they still match the original voltage. Nice.

I don't have mine hooked up yet but might not have to worry about it now as I just checked on the charge and its about 4.8v per cell! That's a 24 cell pack @ 68V with my 48v sla charger!!!!!!!!! Lucky day for Murphy. Not so lucky for me as I'm sure the pack has lost some capacity. :^( Update; Hooked up the BMS to the charger so no more overcharges for me. ;^)

 

[ejonesss]

you should be ok but you are taking a big risk of ruining them.

Anybody not using LVC or BMS with these cell?

Because of my lazyness and space requirement, I've been running it without any protection other than CA. Granted, I only use 5ah out of 18ah each day.

I've just re-made the single large pack into 2 smaller packs and while it was loose, i've check each cells and they still match the original voltage. Nice.

 

[oatnet]

Given his low utilization and CA I think the risk is pretty small, plus a123 is pretty resiliant. I've run 1000s of cycles on LiFe, once I baseline how many wh I can take out of a pack it is a no-brainer to use less. I did lose 2 cells when two of my voltphreaks single-cell chargers overheated and died, so now I make sure my voltage off the charger is correct.

If you dont have a CA, then I'd agree a BMS is necessary. However, If you aren't tracking your wh on a CA anyhow, you are just begging for a quad-busting hump home when your BMS cuts out on you.

-JD

you should be ok but you are taking a big risk of ruining them.

Anybody not using LVC or BMS with these cell?

 

[swbluto]

Just to confirm, a123 cells can die. I haven't diagnosed the charger though it appears to be working, but it appears that I've officially killed two a123 cells. I was hearing reports about how you can take down a123 cells to zero volts without failure so I thought if I caught the cells' voltage fall early enough, I wouldn't have problems. Diminished capacity, possibly, but not completely lost cells.

Well, it appears that a cell group probably went into reverse voltage from being overdischarged and now the cells seem to be soaking up 3A of recharge current with no indication of it actually charging (the voltage isn't increasing - it remains at ".1" volts). I checked the rest of the cells, and they seem to be retaining 3.9 volts without problems so they don't appear to be damaged at all.

It looks like I've killed a cell group of 2.

So, just to mention, it'd probably be a wise idea to run these things with a BMS or low-voltage detector at a minimum.

 

[Lock]

Just wished to file this away somewhere on ES... It's a report from June 2008 detailing the forensics that went into a A123 pack added to a Toyota Prius to make it a plug-in, which let the magic smoke out:
http://file.sunshinepress.org:54445/toyota-prius-a123-car-fire-investigation-report-2008.pdf

One quote:
"As a current-carrying joint becomes loose, the resistance of the connection increases. The power dissipated in heat by a resistor is given by P=I2R. Therefore, a doubling of the resistance increases the power dissipated by heat by a factor of two. The battery pack will experience a maximum discharge current somewhere near 100A during normal driving. Given that the power dissipated by heat is proportional to the square of this current, an increase in resistance to even a few Ohms will result in extreme heat generation. Experience shows that loose connections can lead to temperatures over 250°F."

and:
"It should be noted that there was likely an additional source of heat generation. Due to the placement of the connecting tab at the edge of the Nickel sheet and the layout of the 10 cells in each parallel group, the cell closest to the tab would experience a higher current throughput than the rest of its parallel group mates. This additional current would tend to make this cell run warmer than the rest."

...and the Summary:
"A Toyota Prius converted to a Plug-In Hybrid Vehicle configuration was destroyed by fire. The likely cause of the fire is improper assembly of bolted joints with electrical lugs. These joints became loose causing excessive heating which led to the rupture of individual battery cells which resulted in significant short circuiting of the battery ultimately resulting in the vehicle fire.

The batteries were designed so that a spacer should have been placed between the fuse blade and the electrical lug so that the plastic battery cover would not be in compression. Compressed plastic has a tendency to creep which would result in loss of clamp load on the fastener. Inspection shows that the brass washers that were intended to be the spacers were instead installed underneath the bolt head (between the bolt and the lug which Report of Investigation NRECA PHEV Fire.doc 24 served no purpose). Additionally, there was no locking mechanism (lock washer, safety wire, adhesive, etc.) that prevented the bolt from backing out of the nut. It is likely that the bolted joint became loose over time and during regular vehicle operation which resulted in a high-resistance connection causing those components to increase in temperature.
A123 Systems’ design guidelines appear to be violated in several areas which may have contributed to the severity of this incident."

Edit: Gotta love the part where they describe a failing cell as cutting through steel like a plasma torch...
Tks
llok