A123 20ah Lithium Install

[retepsnikrep]

Well my 52 x 20ah A123 lithium pouch cells arrived today! Quite a small packet and weighed about the same when lifting the parcel as an oem pack. So i'll use this thread to document the install. These will be going in my 240,000 mile car, which has had the unsatisfactory 8ah lifebatt cells taken out and is currently running a refurbed nimh stock pack.

The BMS for the lithium is still installed in the car so just need to build up the pack and bung it in. So easy to say, lots more difficult to actually do. The brief for this project is they must fit in the oem battery space, i'm confident they will.

I have a BCM fooler built up but won't install a BCM interceptor straight away. I'll see how it goes. probably will need one though. My yellow car with the 40ah cells has been on 19 bars for the last year LOL!

[MN Driver]

Great to hear that they've arrived. Sorry to hear about those lifebatt cells not working out, I was thinking we could have had a fairly affordable solution with 38120 cells but it's unfortunate they they weren't able to take the regen current. It seems charging at high currents is a fairly large challenge for lithium of any variety. These larger cells with, what seems to be, lower internal resistance should do very well, I think the smaller 2.3Ah cylindricals are rated at 5C, so if these followed suit, you are only putting in half that amount.

It will be interesting to see how you work them together in the OEM space. Are you still planning to use some sort of clip to fasten them like you said before or do you have other plans now?

For the 19 bars, is it in positive recal constantly? I don't remember exactly how everything went together regarding the fooler and interceptor but I thought you had it where you would cause it to recal every once and awhile, I didn't think you worked it out to never leave the top of the gauge. I know you were worked out the CAN codes so you could try to make the SOC gauge accurate but I thought that was separate to what the BCM thought.

[retepsnikrep]

Just a couple of pics to show how 50 cells should fit into the case with a bit of work to cut the top out of it. All the connections will be at the fan (rear) end so you can terminate them. Each cell and tabs will be seperated by a thin sheet of plastic insulation. I'll probably solder the tabs together with some low temperature solder. The BCM and MCM may be able to remain in postition or raised slightly with the aluminium support bracket adjusted!

The 50 cells will be a snug fit which is good and to compress them I may pass threaded rod throught the case and have a clamp system to ensure they are tightly secured.

The two spares on top are just to give you an idea of size etc. The cells when checked are all within mv of each other and the stack is in parallel in case you were wondering. I may charge them all in parallel first to ensure they are all charged and balanced to the same voltage.

www.solarvan.co.uk/A123/A123Cells003.jpg

www.solarvan.co.uk/A123/A123Cells004.jpg

[retepsnikrep]

MN

After a pos recal the gauge always goes to 19 bars on standard cars. My interceptor just fakes 19 bars SOC all the time. I have my own BMS screen to tell me the SOC of my cells. I just want the car to think it has a nice full battery which it does

If you fake 20 bars then regen is reduced so 19 bars allows full regen and full assist so long as the battery temp is acceptable, and in fact I fake that as well telling the MCM via the interceptor that the cells are always at 20C. Lithium arent really affected by temp like nimh so it doesnt really matter, again my own BMS screen tells me the actual cell temp.

This has worked well in my phev for over a year now. I may eventually drive the stock soc gauge directly with the real Soc but at the moment i don't need too.

[retepsnikrep]

Anyone got any recommendations for some good low temp solder paste?

I might make a clamp device to solder the cell tabs with a couple of 25-50w aluminium clad resistors bolted to some copper tab size tongs to grip the tabs and melt solder paste between them. If that makes sense.

[retepsnikrep]

A quick pic of the 50 x 20ah cells in the stock case. It's a nice fit and when the case is tidied up and insulating sheets between cells then it will be a nice snug fit.

They stand proud by about 2cm which means you could use spacers to stand off the BCM and MCM mountings and do away with the stock ally bracket or make a new one.

Or I could cut the bottom (middle section) out of the case as well and fit a thin aluminium sheet on the base, this would drop them down the 2cm and allow the stock bracket and bcm/mcm mounting to work. The aluminium sheet on the base would help the case retain it's strength and support the cells. Hmm?

www.solarvan.co.uk/A123/A123Cells005.jpg

[MN Driver]

I don't have any idea on using solder paste, I'm only familiar with the standard roll of solder on a spool. I think some of the other guys using these were talking about solder but I don't remember what it was that they were using, or if they even specifically said which one they chose.

I thought it would take more of a shoehorn operation to get them in but it looks as if they were designed to fit the length and width of this box. Far better energy density and power density than the NiMh batteries for sure. 6.5Ah cells where only 60% or so of the capacity is being used before versus 20Ah in roughly the same space, although slightly taller, with roughly the same weight and far more power capability.

[Mike Dabrowski 2000]

Peter,
I was surprised to see the flatpack cells, I was thinking you were getting A123 large cylindrical cells.
How will you cool the cells, it looks like the packs just sit stacked one on the other with no room for air cooling?
The low temp solder used on peltier thermopiles is Indium
Solder and Solder Paste by Indium Corporation

Solder paste is sold in small tubes, it consist of flux and tiny solder balls.

What are the soldering specs of the cells?
I would use a big soldering iron to solder the tabs, with a clamped on thermal sink near where the tabs go in to the package to prevent that seal from getting overheated.
A big iron will store enough heat in the tip so the soldering will only take a second or two as the stored heat rapidly heats the tabs, rather than using a smaller heat source, which will have to generate the heat while heating the tab.
Another battery company that makes that style cell.
**** KOKAM ****
Good luck with the project, I hope the cells work out for you.

[retepsnikrep]

Cooling is not envisaged as a problem but we shall see. The testing these cells have undergone on the ES forum has been upto 600A continious drain until nearly exhausted. Even then they did not heat to dangerous levels. We will only be seeing 100A max for short periods and 50A charge ditto for short periods. Have to see how it all heats up or not as the case may be. Thanks for the solder tips. I'm away shortly for a couple of weeks so i don't expect any real progress for a while.

[retepsnikrep]

I'm working on special version of my BMS multi slave pcb for this project. it will fit into the right hand end of the case where the normal orange cell end connector board would go. I will use two boards capable of monitoring 26 cells each giving a theoretical maximum of 52 cells although this number of cells would be a very tight squeeze into the case! 50 is a good snug fit for these 20ah a123 cells and is the same as I have in the PHEV. The rally insight runs with 52 cells but that uses 15ah lifebatt cells.

I think the voltage drop under load will be lower with the A123 cells than the lifebatt due to the construction, and testing on the endless sphere site has shown incredible performance. So 50 is around the optimum i think.

Endless-sphere.com • View topic - Testing the big 15 and 20Ah LiFePO4 cells is tough! *Pics*

Hopefully pcb layout and production will get done while I'm away and i can build up the boards on my return. I have some low temp solder and flux so will play around with some soldering of aluminium takeaway food containers and cell tab offcuts to see how it works.