[Megatron451]

Lots of questions here. Has anybody dissected the 60.8V LG Chem battery (16S) (LG CHEM Lithium Ion Battery - 60.8V, 2.6kWh, EV West - Electric Vehicle Parts, Components, EVSE Charging Stations, Electric Car Conversion Kits I have the newer 44V model (12S) and I can't tell of the 60.8V one kept the integrated battery balancing circuitry or if it was bipassed. I'd like to hook up my orion BMS to each battery, so I'll need to figure out the best ways to do that, I was hoping for just plugging in somewhere.

Pictures seem to help with replys:

 

[mdrobnak]

My 60V modules are just cells with balance leads coming out to the connector. Looks like that PCB is a balance module. Probably needs a master of some sort, but you probably don't need the Orion, you need something like SimpBMS.

 

[OR-Carl]

Haha, yeah, everyone likes pictures! So it seems like the way production EVs are generally built is to use a pcb on each module that is connected over CAN to a control board. My module boards were connected with a molex wiring harness, yours appear to be soldered in place. I am not familiar with the Simp BMS, but it would not surprise me that people have been working on a drop-in replacement for the control board on common batteries. However, at 475 british pounds, it is not really a budget solution, although it would be a bit cheaper than an Orion, I think.

You can download the installation manual for the BMS you are thinking about using and look over how they are wired. Then take a multimeter and figure out how the cells within the module are wired, and where that voltage shows up on the board. It might route to one of those output connectors. If it does, your BMS can connect there, or else you will need to solder wires into those big connections to give the bms a spot to monitor each cell. The instruction manual should make it a lot more clear.

 

[mdrobnak]

If you want to just do balancing, you can run it on a Teensy 3.2, just need to get it stable power. The code is open source. So it depends on how much people want to do. Mostly PnP, buy the kit (I did) and it is pretty simple and works reasonably well...controls my charger, etc.

-Matt

 

[Megatron451]

Thank you for all the responses. Sounds like there are a couple directions that people would take with these. I need to do some more probing to hopefully find some points on pins that I can get all 12 cell voltages from, if not, I'll probably just have to run wires straight from the cells, which would suck.

I'm kinds stuck on the Orion because I already have the Orion 168 (without the cables though), so I'll probably just keep trying to use that.

 

[angusmf]

Ignoring the management stuff, how possible does it look to disassemble these and make something like a 60 or 72V module?

 

[Megatron451]

Ignoring the management stuff, how possible does it look to disassemble these and make something like a 60 or 72V module?

You could do it, but there isn't really an easy way to hookup BMS, probably better off finding the 60.8V versions of this if that is what you are looking for.

 

[angusmf]

You could do it, but there isn't really an easy way to hookup BMS, probably better off finding the 60.8V versions of this if that is what you are looking for.

Thanks, good to know. For what I'm doing, removing the OG BMS and soldering or welding on some leads would be reasonable. For the 60-72V scenario, I'm hoping it would only need one or two of these to deliver 500A for reasonably short bursts.

Hope you figure out something on the BMS tho. It would be nice to use a stack of these in a larger EV project also.

 

[OR-Carl]

I'll probably just have to run wires straight from the cells, which would suck.

I dont think it would be that bad. Seems to me you have 2 decent options:

Either solder a wire into the power connection for the PCB (red) or drill and tap the aluminum plates on the actual cells themselves (orange). The BMS connection does not have to be beefy, it can be made with 20 gauge wire. Its a little hard to tell just exactly how big those soldered connections are, you might need a decently sized iron to melt that much solder.

 

[Megatron451]

I really appreciate your response to this. The red circles actually didn't show voltages on my multimeter when testing them, I also assumed they would have the voltages of the cells. I might have to do your second option. But in the next couple of days I'll just go back through with my multimeter and set the black lead on the negative of the battery, and just go around every little connection I can find with the red lead of the multimeter and see if I can find anything.

I'd like to use the thermisters if possible, but I'm not sure where the connections are for those and how to tell what range they are. I "assume" if I can find the wire, it is a resistance wire compared to ground?

 

[brian_]

I dont think it would be that bad. Seems to me you have 2 decent options:

Either solder a wire into the power connection for the PCB (red) or drill and tap the aluminum plates on the actual cells themselves (orange).

Those don't look like aluminum plates - they look like copper bars with the cell tabs folded over them and welded together. Any terminal should probably be attached to the exposed end of the copper, not the tabs.

Also, as the text moulded into the case end frame suggests, this is the rear of the module; the front (opposite end) will have seven more similar bars, with the end ones connecting to the module's terminals. The BMS will need to tap into all 13 locations. The front is shown in the photos with the orange covers in place and the terminal posts protruding. The disconnected ribbon cables visible at each side presumably go the printed circuit board at the front.

 

[OR-Carl]

Those don't look like aluminum plates - they look like copper bars with the cell tabs folded over them and welded together.

I am not an expert, but I thought those welds look like friction stir welds, which made me think they would have exposed the underlying copper if those tabs were made of plated material. But that was just a guess.

Looking at this picture again, and I feel like we are missing something obvious. I also assumed we were looking at the back of the module, but now that I think about it, I do not see that the cells are actually wired in series... This would explain why there was no voltage - you might be measuring cells that are not connected anymore! I am starting to suspect that there must be interconnections on the reverse side of the orange cover plates? Can you post a picture of the front and back of the exposed modules, plus the inside of the cover plates? That would be a clever design, because as soon as you take off the cover, you break the circuit and reduce the voltages. Remove both covers, and you have 6 isolated cells, which would only read voltage from opposite ends. Does that make sense?

As for the BMS taps, Brian, wouldnt you need 7 connections for a 6 cell module? It is true that you will need one tap from the far side, but it has the terminal stud that can serve as the connection point, right?

 

[brian_]

I am not an expert, but I thought those welds look like friction stir welds, which made me think they would have exposed the underlying copper if those tabs were made of plated material. But that was just a guess.

No, the friction welds used in battery connections typically just join the materials where they touch each other, like a traditional electric spot weld; they're not friction stir welds that mix the materials to a substantial depth. Look at the tiny wires that attach each end of each 18650 or 2170 cell in Tesla module to the aluminum bus plate - it looks like the wire has been smashed onto the bus plate and stuck there. Anyway, this module is similar to any other LG Chem module of pouch cells, so the people who have worked with them at the cell connection level can confirm how the welds work.

Looking at this picture again, and I feel like we are missing something obvious. I also assumed we were looking at the back of the module, but now that I think about it, I do not see that the cells are actually wired in series... This would explain why there was no voltage - you might be measuring cells that are not connected anymore! I am starting to suspect that there must be interconnections on the reverse side of the orange cover plates? Can you post a picture of the front and back of the exposed modules, plus the inside of the cover plates?

Yes, they're wired in series. Each of those visible connections is presumably the positive tab of one cell (one one side of the copper bar) connected to the negative tab of the next cell, supported and located (and connected to the BMS) by the copper bar. There are 12 cells in series, so there are module positive, module negative, and 11 inter-cell connections... there appear to be six here on the back and so the the other seven (module positive, module negative, and 5 inter-cell) must be on the front.

The feature which makes this module look unusual - which threw me at first - is that the cells have one tab (terminal) on each end. Pouch cells more commonly have both tabs on the same end, each less than half the cell width, so connections in those modules alternate side to side as the current flow path works its way from one end of the stack to the other.

I can only guess that the soldered points have no voltage relative to anywhere because they are some sort of isolated intermediate point in the monitoring circuit for each cell.

But yes, photos of the front and back, both with covers removed, would make everything more clear.

As for the BMS taps, Brian, wouldnt you need 7 connections for a 6 cell module? It is true that you will need one tap from the far side, but it has the terminal stud that can serve as the connection point, right?

These are 12S (not 6S) modules, so you do need 13 BMS connections, 7 of which (including the two terminal studs) are on the other (front) side.

Here's a text drawing attempt to represent what appears to be the module configuration:

< Front Back >
(+)--[+ cell 12 -]__
/----[- cell 11 +]-/
\----[+ cell 10 -]__
/----[- cell 9 +]-/
\----[+ cell 8 -]__
/----[- cell 7 +]-/
\----[+ cell 6 -]__
/----[- cell 5 +]-/
\----[+ cell 4 -]__
/----[- cell 3 +]-/
\----[+ cell 2 -]__
(-)--[- cell 1 +]-/

 

[OR-Carl]

Sure enough, it says 12s on the first post in this thread , I had forgotten that this is a 44 volt module... Disregard my previous speculation...

I can only guess that the soldered points have no voltage relative to anywhere because they are some sort of isolated intermediate point in the monitoring circuit for each cell.

If they are monitoring voltage, though, then they should show voltage relative to somewhere, right? If those solder joints show no voltage relative to each other, then they must not be connected to the copper bars that hold all the cell tabs. I am not sure what purpose they would serve in that case. I assume the OP verified that there was voltage between the sets of tabs/copper bars? At any rate, hopefully more pictures and an update will make things clearer.

 

[Megatron451]

so the two connectors that connect to the battery ballencer board have all the voltages needed for my BMS. Does anybody know how to source connectors? I was a digikey for an hour and couldn't really get anywhere.

 

[OR-Carl]

That last photo you posted is the back of what I am going to call Board B, right? It mounts on top of Board A with 3x bolts above, 2x in the slots next to those weird connectors (6), and also holds the ribbons marked 4 and 5?

That appears to be more of a 2 part socket than a wiring harness. The two bolts on each side are likely what is holding it together. Also, if Brian is correct (and he usually is) then some of the cell taps are being routed through those ribbons, and so those signals would be lost if you remove board B?

I see a lot of clear goop on those circuit boards, but did they by any chance not cover the back sides of those sockets? (i.e. the solder points visible at 6 on board B?) You could solder your wires to there if they show the voltages. It seems like you are keeping an awful lot of unknown circuitry, but I dont see how you would get all the voltages otherwise on account of those ribbons.

If it was me, I would be tempted to think about ditching the PCBs entirely and trying to make a mechanical connection to those copper bars or aluminum/silverish tabs. Your BMS would have to route wires to the front and the back of each module. I suspect those BMS boards would not be a problem, per se; they probably are designed to not draw meaningful amounts of power from the pack, but they will be serving no function for you if you keep them.

 

[Megatron451]

I appreciate the conversation about this. So I am going to call board b the battery balancing board, I assume it is only to use resistors to make sure all the batteries are about the same voltages. I pinned out #1 to see if there were any voltages, but there were not, it is only an 8 pin connector, so I wouldn't have the 12 voltages needed plus ground.

I think #4 and #5 are either some type of pressure sensors or the thermisters, either way I don't know enough about those technologies to use those.

#6 is what I am most interested. Those two black connectors have all the voltages needed for my BMS. They are two 8 pin connectors that connect to the back of board B. If I can't get anywhere with sourcing the connectors, I might just de-solder the connectors from the back of board B and run wires to the orion BMS that way.

Pictures below are removing the first plastic on the back side of the battery to get to the board B. It is held in by 5 screws.

 

[OR-Carl]

I am going to call board b the battery balancing board, I assume it is only to use resistors to make sure all the batteries are about the same voltages. I pinned out #1 to see if there were any voltages, but there were not, it is only an 8 pin connector, so I wouldn't have the 12 voltages needed plus ground.

Yeah, I think you are on the right track. The outgoing harness (1) is probably connecting all the modules to a CAN bus, so it would not need to pass the voltages; it would just send data.

I see now that there are 3 ribbons, so the center one must carry the voltages from the front (I am at this point clearly not the Shelock Holmes of batteries, but it is fun trying to figure out how it works ). Desoldering the connectors - or just soldering wires to the tabs - seems like it would be a simple and efffective solution.

or the thermisters, either way I don't know enough about those technologies to use those

Thermistors are pretty easy to identify - they will be a pair of wires that read somewhere around 20k ohms across them. The exact value will depend on ambient temperature, and they do come in different ranges, I believe, but I think 20k is a pretty common value.

Do these battery modules have liquid cooling? Is that what the plug is on the left side?

 

[Megatron451]

Thermistors are pretty easy to identify - they will be a pair of wires that read somewhere around 20k ohms across them. The exact value will depend on ambient temperature, and they do come in different ranges, I believe, but I think 20k is a pretty common value.

I'll check voltages against ground first, if no voltages, mabe I'll get brave and try some resistance measurements across the side ribbon cables and see if that is what they are.

Do these battery modules have liquid cooling? Is that what the plug is on the left side?

Hmm, which plug might you be referring to? There isn't internal cooling, like it's 60.8V LG Chem big brother, I'll need cooling plates on the bottom.

 

[OR-Carl]

I was looking at this:

but going back and looking at it again, it seems to be the protruding end of a bolt, not a plug.