A123 pack build

curious

1 kW
Joined
Dec 29, 2007
Messages
445
Location
NJ, USA
Ok, I decided to share personal experience building the packs out of dewalt A123 cells. The goal is to have two 10s6p packs for my first e-bike conversion. I intend to later add a parallel / serial relay switch so I can run either as a 33V or 66V configuration.

I decided not to use original plastic cell endcaps to make parallel wiring of individual cells easier and packs a bit more compact. I plan to use sticky rubber strips to run along the pack edges before enclosing it in heat shrink.

For parallel buses I decided to use single strand bare tin-coated wire (McMaster is the only place I found that sells small pools) enclosed in sectioned high temperature sleeving between solder joints. This proved to be a right decision as soldering a pair of wires to the cell interconnect tab is more cumbersome than a section of continuous single strand wire. I ordered PTFE (teflon) sleeving for the wire (also from McMaster) which is fairly tough and high temperature resistant so you can solder right next to the sleeve (even touching it with soldering iron) without melting it. However if I would do that again - I'd go for a braided sleeving (coated fiberglass or kevlar), since the PTFE sleeve is relatively thin and can be punctured when the pack rattles on the bike. For now I'll probably add some reinforced tape pieces as protection between the cells and sleeved bus. I intentionally made a slight upward bend in each sleeved bus segment to add 1mm of separation from the cell.

The rest of this message are suggestions based on what I have learned in the process -

Building large packs resembles a production line - a lot of boring repetitive work. The key three things to survive it IMHO are (i) having proper tools, (ii) maintaining clean table with no metal bits and pieces to short the cells, (iii) being constantly alert (it is very easy to forget that you are working with charged cells and short something).

Cordless screwdriver, *temperature controlled* soldering iron are pretty much required for this work. Although I have a decent soldering station for precision work I got a Weller W100PG temperature controlled 100W iron specifically for working with massive wires. The other cheap department store non temperature controlled iron overheats and oxydises like mad.

Protection glasses are a must. Despite all the effort I did short the packs twice already :oops: Both times trying to pull the original wire from output tabs - it sticks there sometimes and it is too easy to slip the iron and short something. The spark instantaneously evaporates away offending metal pieces and that can fly i your eyes. I also plan on buying fire extinguisher - just in case. Honestly working with packs that big and powerful scares me somewhat :? .

Some Dewalt packs come with pieces of yellow insulation tape. DO NOT throw this stuff away. It is a very nice heat resistant Kapton tape and is very usfull in insulating something next to the solder point as it does not melt at all. The white insulation tape on other packs is junk - it melts just seeing the iron.

A lot has been said in this forum about checking cells and balancing them prior to assembly - just a reminder.

If using bus interconnect tin-coat every welded tab prior assembling the pack and soldering buses. Clean the blue marker dots with alcohol and use rosin flux pen. The welded tabs are fairly easy to tin coat if (i) they are properly cleaned and prepped (ii) soldering iron is not overheated. On the picture you can see some intentional offset in coating where two buses run closely in parallel.

When doing final soldering use rubber mat to cover the remaining cells (this has been mentioned in the blog on car pack build that Doc posted recently). Maintain utmost attenion not to short things.

The pictures illustrates prepared dewalt pack (balanced, hot glued with tabs tinned). Notice extra isolation where the output tabs are located - in a pack these will overlap th other cell terminal. I used two layers of Kapton tape during assembly - one inserted under the output terminal sticky side up and another covering the overlapped cell top sticky side down. Make sure the tape completely overhangs the cell top. Later I plan to add some fiberglass tape in that area to avoid puncture.

As you can see there is no output bus yet on the pack - I am still waiting for 12AWG silicone wire to arrive.

One thing I have not decided upon yet as what to use as a fuse for the pack. I need something in the 150A range that is not too bulky. I may just use a very short section of AWG 16 wire wrapped in several layers of fiberglass outside the pack as a fuse ;).
 
Ok, found solutions to two issues:
(i) Ordered 200A fuses from parts express (for *big* car audio systems) - they are large but pretty flat. I'll place it in a heat shrink and attach it to the outside of the pack.
(ii) A small drop of hot glue between teflon sleeve and the cell top prevents sleeve from rubbing against sharp edges of the cell. It is a quick fix and no need to insert pieces of tape underneath the bus wire.

Still waiting for AWG12 wire to finish the pack #1 and hunting for replacement dewalt packs (see my other thread on bad cells) on ebay to complete the pack #2.
I am considering running two AWG12 wires and two anderson connectors in parallel for each output lead.
 
I like your pack ... very clean ..

Why did you build such a big pack .. where will you put that thing on your bike?.


i like your 36v 66v switch idea ...

i will be doing a similar build like you ..l.

except i've choose to keep the packs in the original casings ..easier to mantence or check cells ... i can use different charging methods with differend harnes.. as well all balance each pack individually to ensure they all are working correctly..

-steve
 
Nice job. :) I've always found that running parallel groups like this, where in essence you have groups of six cells in parallel, and these groups are then in series, is better than keeping each series string separate. The packs/cells will take a lot more "abuse" this way. You just need to make sure you don't run the cells down too far, as that's the only real way you can do any damage to these. With six cells in parallel, this can deliver 420A continuously :shock: and a whopping 720A for 10 second "bursts". :shock: :shock: This is without any sort of damage, whatsoever, so you will start melting things way before damaging cells, in a dead short. Your big fuse will work fine, though.

-- Gary
 
Thanks for kind remarks.
Does anyone sell silicone insulated wire thicker than 12 AWG in small quantities ? I'd prefer using a single wire for output leads. 150A (the peak current of the Kelly BLDC controller) is too much for a single 12 AWG and this is the biggest to my knowledge that RC industry uses.

steveo said:
Why did you build such a big pack .. where will you put that thing on your bike?.
I am a heavy guy and need some reasonable runtime in hilly terrain. The pack is not that big actually, just a bit over 1' long. Two of these will fit on a rear rack just fine. I am modifying a hardtail so rack strength is not much of an issue. I'll put a pair in a bag so I can take them off easily for charging.

GGoodrum said:
You just need to make sure you don't run the cells down too far, as that's the only real way you can do any damage to these.
LVC will be there no question, the pack is too expensive to be ruined by overdischarge of a single 6p section. I decided to make it external to the packs so I can finish the packs before selecting what type of LVC to use. It is going to be either yours or something of my own design (I have few ideas that I want to check before posting it here).
 
curious said:
Thanks for kind remarks.
Does anyone sell silicone insulated wire thicker than 12 AWG in small quantities ? I'd prefer using a single wire for output leads. 150A (the peak current of the Kelly BLDC controller) is too much for a single 12 AWG and this is the biggest to my knowledge that RC industry uses.

steveo said:
Why did you build such a big pack .. where will you put that thing on your bike?.
I am a heavy guy and need some reasonable runtime in hilly terrain. The pack is not that big actually, just a bit over 1' long. Two of these will fit on a rear rack just fine. I am modifying a hardtail so rack strength is not much of an issue. I'll put a pair in a bag so I can take them off easily for charging.

GGoodrum said:
You just need to make sure you don't run the cells down too far, as that's the only real way you can do any damage to these.
LVC will be there no question, the pack is too expensive to be ruined by overdischarge of a single 6p section. I decided to make it external to the packs so I can finish the packs before selecting what type of LVC to use. It is going to be either yours or something of my own design (I have few ideas that I want to check before posting it here).


Go here for your silicone wire: http://www.trossenrobotics.com/store/c/3051-Wiring.aspx

I found that because i was searching cable for my 12s 30A charger!! 10 ft long x 24 wire of gauge 10!! :shock: :lol:

i'll need 240ft of cable!! for my 10 ft (x24 wire)cable between the charger an the bike


They have the best price.. BELIEVE ME !! I searched a complete afternoon for that and i used RCgroup forum as great source of info for that.

Doc
 
Doctorbass said:
10 ft long x 24 wire of gauge 10!! :shock: :lol:

I think you may have bought all of their 10AWG stock. They do not list 10AWG any more on that page :D
CBP has some 10AWG listed in stock, I may get some there. Since I need about 10ft total the price is not that critical.

BTW another useful resource I found on rcgroups for anderson connectors and fuse assemblies is this
http://www.powerwerx.com/
 
You might check places that sell high powered car stereo amplifiers. It might not be silicone, but they usually have some nice heavy wire that has lots of tiny strands, making it very flexible.
 
fechter-
I so much prefer to work with high temp insulation when soldering is involved that I'd rather pay some extra for silicone wire. Just ordered some 10AWG wire from CBP.
 
I see your point. I remember seeing some nice military grade 8ga. braided jumpers that were teflon insulated on some surplus site. I can't seem to find it now, perhaps they sold out.
 
I stared at the half-build pack for few days and decided to trim overhanging output tabs. (BTW using dremel cutter on the tabs leaves a lot of sharp shards that tend to stick to the tab so I used sandpaper underneath the tab to clean the cut.)

I almost dropped pliers on the top of the pack once so it is critically important to have a rubber mat covering the pack when working on it. For the second pack I am going to cut a hole a bit larger than battery diameter in a rubber mat and use it as a mask when working on a pack. Shorting the individual pack tabs with soldering iron was bad enough but dropping a metal instrument on a full 6p pack is a guaranteed disaster.

Hot glue was used to fill the gaps between teflon sleeves and battery terminals. If I do the build again I'd get a heavy duty braided sleeve and skip that step. BTW I found that where the two bus wires are located nearby (there are 2 pairs like that on the back side and one on the front) it is better to solder the wire closer to the positive battery terminal and have a sleeved span closer to the negative terminal. This way there is less chance of sleeve abrasion damage. On a first (pictured) pack I did the back side correctly but front side the wrong way.

I decided to use the same bare bus wire (0.064" or AWG14) for connecting output terminals in parallel and then run a pair of AWG12 silicone wires for each battery terminals. Output wire connection points to the bus bar for 6p battery are chosen as xXxxXx pattern such that there is no more than one span of AWG14 bus wire separating a cell from the AWG12 output wire. Each output wire is terminated by it's own Powerpole plug so I have 2x current capacity of the connector.

For charging/balancing wires I used AWG16 silicone wires. I think this is a bit of an overkill and next time I'll probably use AWG18.

Edge rubber guards and rubber spacers were added before applying heat shrink so that the battery will rest on the guards / spacers rather than bus wire (at least on a flat surface). The pictures below show both sides of the pack before applying heat shrink.

Honestly I did a rather lousy job with heat shrink and will likely redo it once I finish the second battery (so no picture of the covered battery yet). I have to yet figure out how to trim the front and back sides of the shrink wrap cleanly and seal the pack.

This pretty much concludes the build. The second pack should go a bit faster (pending the arrival of two replacement dewalt packs). Four good dewalt packs are already prepped for the build. If I discover some major deficiency of pack construction while using it on the bike I will certainly post here.
 

Attachments

  • upper_2.JPG
    upper_2.JPG
    63.1 KB · Views: 3,240
  • lower_2.JPG
    lower_2.JPG
    64.2 KB · Views: 3,232
Hello thanks to curious this the format that i am going to follow in constructing my 10 s 4ack construction to ultimately come to a 2 pack of 40 cells.

efreak
 
Hey great work. Would like to see a description of how you plan to charge it.
 
Yes the plan is using tyco dc-dc converter modules capable of being adjusted to work at 3.6 volt around 18 amps to charge a 4p m1 cells times 10.

10 modules feeding off of a 48v or so power supply ( still in the works)

the packs will have Gary Goodrums lvc board connected for low voltage protection. and the modules will charge each cell like a single cell charger so all around should be foo; proof at least I hope i don't mess anything up. the are a couple of threads by doctorbass that i will following to help me along.

http://endless-sphere.com/forums/viewtopic.php?f=14&t=2824

efreak
 
Nice and clean looking pack.

When I build mine I used clear Duct tape to cover the exposed leads as a precaution. Shoe-Goo works really well for sticking the cells together. It's common practice in RC.

Where did you get the edge rubber guards? That's really interesting. I was planning of using those anti-shift kitchen pads.
 
Looks very nice!

How are you going to charge and balance the packs?

I haven't quite got my head around the layout - it's 6 cells in parallel, then these sub packs are in series? It's 5 wide, that's all...


Oh, and I'd forget the 33v/66v idea. 33v will suck so much you'll never use it once you have experienced 66v!!
 
my pack is only 4 x 10 cells then two of them series

10s4p%20Sub-Packs.gif



charged similar to this without he voltage detection circuit
10-Cell%20Boost%20Charger%20Circuit.png


using dc dc modules explained here somewhat

http://endless-sphere.com/forums/viewtopic.php?f=14&t=2824&start=90&st=0&sk=t&sd=a


http://endless-sphere.com/forums/viewtopic.php?f=14&t=4196

efreak
 
Jonathan said:
Where did you get the edge rubber guards? That's really interesting. I was planning of using those anti-shift kitchen pads.
mcmaster# 93755K43 for rubber guards, I am sure there are cheaper alternatives though...
mcmaster# 8871K71 for bare tinned copper wire for connecting packs in parallel
get a high temperature *braided* sleave for insulating bus wire (many options in mcmaster) it will be much more abrasion resistant against tabs

Mark_A_W said:
Oh, and I'd forget the 33v/66v idea. 33v will suck so much you'll never use it once you have experienced 66v!!
I have a lot of long steep hills where I live and the motor controller efficiency is not good at high current/low rpm combination. I want a long range setup.

swade said:
Would like to see a description of how you plan to charge it.
Working on it. My first attempt at a DIY switch-mode balancing charger did not work well for multiple reasons. However I believe that now I finally understand how to build a high efficiency and compact integrated balancing charger 8). I am contemplating the idea of making a commercial product based on things I've learned. It is not a big market but may compensate my hobby expenses.
 
One more note - in the final assembly (not shown on the pictures) I've added few layers of FG tape to the cell tops where the balancing wires rub the contacts. Again a braided sleeve around balancing wires would be a cleaner option.
 
Hi guys, How long does the shoo goo take to get tacky/hold reasonably well? I used black vinyl floor runner to protect my last packs. It is easy and cheap to work with. Comes from Home Depot. Has ribbed texture. Then I use poly tape over the whole thing. These packs are 6p4s Lithium manganese @ 15AH
Jeff K. "Deep Cycle" project
 

Attachments

  • DSC02922.JPG
    DSC02922.JPG
    59.3 KB · Views: 2,029
Back
Top