Pics/Info: Inside the battery pack

[jerry33]

Actually you can customize a fair amount if you go to "Settings" (top right of this page) and then "General Settings" (in the left menu), you'll see a long list of options including how threads are displayed and how many posts per page etc.

You're right in that it's not perfect, but this is a free-to-use forum which exists on donations, very little advertising income, and volunteer moderators. When Doug(Admin) logs in later he can hopefully take a look at the Wiki-post addition for this thread.

I'd be happy if it just sent every post as an email the way the Yahoo groups do.

As far as donations go, I donated some time ago now and I thought that Doug was doing something as a yearly reminder. The Mods here do great work (even if they're often not appreciated) and I'm aware of how much time, effort, and cash it takes to keep things going.

 

[NigelM]

And honestly, I had looked before and never saw the setting for posts-per-page. Set to 500 now and things feel much better on that front at least. Thanks for getting me to take a second look.

You're welcome! For folks that prefer the threaded type view (a little more like reddit) that option is also customizable from the "General Settings" page. There's also a help page describing individual options here--->faq_vb3_thread_display

 

[doug]

This useful side discussion will probably be moved to site feedback. Thanks for the input.

I'll work on a solution about post edits. In the meantime, for this type of content, the blog section might make more sense than the wiki. You could give that a try.

 

[Kiers]

... So, sanely applying a charge voltage to the pack should charge it fine.

Curious...how many kWh's of energy does it take to fully charge this 85kWh battery pack?

 

[wk057]

Curious...how many kWh's of energy does it take to fully charge this 85kWh battery pack?

Depends on the efficiency of the charger. The grid-based one I have is about 90% efficiency, so about 93kWh if it were completely dead... not going to let it get that low though. The solar charge controllers I have are more efficient.

 

[SeattleMatt]

I'm trying to understand the wirebonded fuse links on the 18650 cells. From this thread we learn that the pack is 96S74P, and if 3.1AH cells this means that 1C is 229A.

The Model S at 70 mph has an average C rate of about 0.25C (since it can cruise that that speed for nearly 4 hours). The P85D can pump out 691 HP for 2 seconds, which on a full pack means that's about 5.5C. So, we know the average current on the cells is about 0.25C, and the peak is about 5.5C. A rigorous hill climb would be about double the flat cruising (about 0.5C), but overall, we're left with a peak to average current consumption of roughly 10:1.

Using Onderdonk's equation for fusing and assuming copper, a 0.5 mm dia wire could withstand 5.5C for about 11.2 seconds, and the impedance of 10 mm of 0.5mm wire is about 84 mOhm per meter, so 1 cm would be under 1 mOhm. The cell impedance is probably around 50 mOhm, so the wire bonded wire contribution is minuscule.

Thus, it seems roughly that the wire bonding wires are probably on the order of 0.5 mm in diameter. And the pictures at the beginning of this thread hint that roughly 18 to 26 of these wires could fit across the 18mm cell diamter. But I wonder if anyone has published an actual measurement on the wire diameter? And I assume they aren't pure copper, but an aluminium alloy such is used for wire bonding power semi conductors.

TIA

 

[wk057]

I'm trying to understand the wirebonded fuse links on the 18650 cells. From this thread we learn that the pack is 96S74P, and if 3.1AH cells this means that 1C is 229A.

The Model S at 70 mph has an average C rate of about 0.25C (since it can cruise that that speed for nearly 4 hours). The P85D can pump out 691 HP for 2 seconds, which on a full pack means that's about 5.5C. So, we know the average current on the cells is about 0.25C, and the peak is about 5.5C. A rigorous hill climb would be about double the flat cruising (about 0.5C), but overall, we're left with a peak to average current consumption of roughly 10:1.

Using Onderdonk's equation for fusing and assuming copper, a 0.5 mm dia wire could withstand 5.5C for about 11.2 seconds, and the impedance of 10 mm of 0.5mm wire is about 84 mOhm per meter, so 1 cm would be under 1 mOhm. The cell impedance is probably around 50 mOhm, so the wire bonded wire contribution is minuscule.

Thus, it seems roughly that the wire bonding wires are probably on the order of 0.5 mm in diameter. And the pictures at the beginning of this thread hint that roughly 18 to 26 of these wires could fit across the 18mm cell diamter. But I wonder if anyone has published an actual measurement on the wire diameter? And I assume they aren't pure copper, but an aluminium alloy such is used for wire bonding power semi conductors.

TIA

I'm actually in the process of dissecting a damaged module (from a third pack.........) and I'll be doing a bit more tests on the cells and these fuses in particular when I get more time.

 

[llavalle]

I'm trying to wrap my head around the module configuration (basically how each cell is wired in a module).

Each module has 74x6 cells, right? I assume they are 6S74P, right? That means you should have 6 separate bus-bar type of backing connecting the negative terminal of the battery (one for each 74pack of cefor but from your picture I can only see 4.

Is there something I'm not seeing?

@SeattleMatt, I had to read the "96S74P" nomenclature a couple of times to realize what it meant...even though I'm very familiar with this (LiPO RC enthusiast here). I'm just not used to numbers that big! I run 4S2P packs in my drone

For those of you who don't know what it means : 96S = 96 cells in series. 74P = each series has 74cells in parallel. Since most modern lithium based cells have a max voltage of around 4.2 volts, you can calculate the max voltage of your battery pack by doing "number of S" x 4.2. In this case, 403.2 Volts.

Btw guys, do you know what's the max discharge and recharge rate of these cells? I was thinking of building a small battery back for my hexacopter using 18650A cells but I'm concerned with my amps draw. I alreay have IMR cells that can take way more but they are obviously a lot less dense (weight wise) and since this is a flying machine - every grams counts.

 

[peter_east]

Hey wk057,

I've designed large lithium ion packs and systems so might be able to help if you have questions. Great photos, btw. One thing I noticed is that you refer to the board on each module as a BMS but it's really just a monitoring and balancing board. How do I know? Because everyone does it that way and I can tell by looking at it. The board has a micro, a precision reference, and a set of resistors to drain and high VSOC cell groups. There is also an isolated CANbus transceiver, you can see the trace separation areas on the PCB. There's an inside pack CAN and the external CAN. The actual BMS is the big PCB near the contactors and I'd love to see a high res photo of it as well. Unless you want to do some time-consuming reverse engineering you're not likely to use their BMS. Among other things it wants to talk to other vehicle components and unless you have an internal spy in Telsa who is willing to risk their job you're not going to get the mapping of its registers on the bus. You're probably not going to figure out even the small balancers either, although that's at least a doable project in a reasonable amount of time if you're good at code and they haven't locked the chip.

As I said, everyone is basically doing the same thing, a small balancer and measurement board on each battery module, powered by the module, and an isolated CAN transceiver talking to the BMS which makes all the decisions and sends balancing commands (because it has to have knowledge of all series cell groups).

Cool project you've undertaken! Like I said, if you have any questions, please ask, and I'd love to see a nice photo of their BMS (just the component side) to compare to what I had done in my designs.

 

[wk057]

Each module has six sets of 74 cells. These sets are wired in series using five bus plates that run on the top and bottom of the modules. Half of the cells are inverted. So the positive terminal connects to the positive of end of the first set of cells, then the negative of those cells on the opposite side is connected to a bus plate that connects those 74 negatives to the next set's 74 positives. Then the next set's negatives on the opposite side are connected to a bus plate that connects to those 74 and the next 74's positives. It goes up one side of the module and back down the other side of the module so that the ends of the series are on the same side of the module on the top. Pretty nice design.

Then the modules are all connected in series inside the full pack.

 

[llavalle]

Wow ok! Pretty neat design indeed. Very compact and lightweight.

One of the module would make an excellent battery source for my lipo charger while I'm on the go.

I usually use my car battery but the 12v batt in the S is scaring me a bit...unless the dc to dc converter is charging it...

 

[wk057]

So, I have a module from a 3rd pack that is damaged and I decided to dismantle it to harvest the good cells. Honestly, I don't know what I've gotten myself into. My original assessment of the pack being built like a tank definitely applies.

Each cell is individually glued to the hard plastic framework that holds everything in place. This framework is glued to the aluminum rails on the sides. I butchered the rails off of the thing in hopes to be able to maybe torque the whole thing apart. No dice. I literally jumped up and down on it while having one end propped up and it was solid as concrete.

I managed to remove about 6 cells without damaging them through brute force efforts on the plastic on the edge row of cells, and 3 others I punctured in the process. I think the only way to safely remove them would be to somehow dissolve the glue, unless I'm missing something.

Oh well.

 

[Johan]

So, I have a module from a 3rd pack that is damaged and I decided to dismantle it to harvest the good cells. Honestly, I don't know what I've gotten myself into. My original assessment of the pack being built like a tank definitely applies.

Each cell is individually glued to the hard plastic framework that holds everything in place. This framework is glued to the aluminum rails on the sides. I butchered the rails off of the thing in hopes to be able to maybe torque the whole thing apart. No dice. I literally jumped up and down on it while having one end propped up and it was solid as concrete.

I managed to remove about 6 cells without damaging them through brute force efforts on the plastic on the edge row of cells, and 3 others I punctured in the process. I think the only way to safely remove them would be to somehow dissolve the glue, unless I'm missing something.

Oh well.

Pictures? Pretty please, with sugar on top...

 

[wk057]

Pictures? Pretty please, with sugar on top...

Yeah, was getting there.

 

[Johan]

Very cool. So it's pretty clear then that the Tesla engineers are treating the modules as the smallest functional piece of each pack.

 

[kennybobby]

Great photos! Maybe a low temperature heat gun could blow 200-300F air and help loosen the adhesives for you. My curiousity is awake and now i have questions.

Do you have a micrometer or calipers to measure the diameter of the fuse wire? Can you determine the material of the fuse wires--copper, silver, aluminum, steel...?

It looks like the copper cooling tube is covered with ?foam tape, and there is also some sort of cloth tape showing. Is the cloth tape only at the end turns or touching every cell?

The sheet of metal that connects the 74 cells in parallel appears like some sort of foil that rolled up--i always thought it would be solid copper or aluminum "plates". Can you determine the thickness and material of those?

Those modules are an amazing assembly of engineering excellence--and that battery pack is a game changer to make BEV transportation an economically-feasible reality.

 

[JRP3]

Great photos! Maybe a low temperature heat gun could blow 200-300F air and help loosen the adhesives for you.

Or maybe a hot wire knife to slice through the glue without heating the cells.

 

[doug]

It looks like the copper cooling tube is covered with ?foam tape, and there is also some sort of cloth tape showing. Is the cloth tape only at the end turns or touching every cell?

Looks more like 3M thermal conductive tape.

 

[apacheguy]

Looks more like 3M thermal conductive tape.

Doesn't it have to be hollow on the inside to allow the coolant to flow?

 

[JRP3]

The tape holds the hollow copper tubing to the cells, the coolant flows inside that tubing.