DIY Electric Car Forums banner

Model 3 Battery Pack for Tesla Cobra EV

10643 Views 14 Replies 7 Participants Last post by  The_Sasquatch
6
Product Electronics Toolroom Machine Furniture


We decided to build a 3rd generation battery pack for our Tesla Cobra EV road race car. Our 1st pack was from a 2015 Kia Soul EV. It had very high power output (we were able to pull 375kW), but the capacity (27kwh) was a little short for our 20 minute races. We replaced it with 2/3 of a Chevrolet Bolt pack. Despite being water cooled, we found this pack ran hot. To keep it cool, we had to resort to building custom cooling plates between every cell pair. We did manage to keep the back cool, but were never able to get the capacity we wanted due to high voltage sag under load.

Despite the weight penalty, we decided that the Tesla Model 3 pack would give us the capacity we wanted. From our single cell test data, we measured the cell capacity at 4.5 Ah at 20 amps constant draw. This equates to 74 kwh for normal use. Under racing conditions, we estimate that we will have at least 48 kwh usable.

There is a fuse wire on each 21700 cell that will limit the maximum current that the pack can pull. We measured these to be .035” aluminum. We believe these are good up to about 30-35 amps peak. We have run the completed pack up to 1050 amps (25 amps per cell) for 15 second bursts with no problems.

One of the things that attracted us to the M3 modules was their water cooling arrangement. It is much improved over the model S. There is a parallel water header on each end of the module. We built a custom manifold for each end for connection to our cooling system. We are utilizing a Mezeire 70 lpm pump and a full size aluminum radiator. If you are buying individual modules, make sure that you get the water connections as they are not readily available!


Machine


The modules will be difficult to fit in many cars. The only way the pack would fit in our car was to stack all four modules on top of each other. Even at that, we had to modify our chassis to fit.
Short modules. 23 cell groups. 185 lbs. 70.5 x 11.5x 3.5” tall (4.25” including BMS). Length is measured from the outside-outside of the water outlets.
Long Modules. 25 cell groups. 220 lbs. 76”x11.5”x3.5”

Because there were no known solutions for communicating with the BMS boards when we started on the project, we elected to use an Orion 2 BMS system.
The cell tap wiring for the factory BMS is thin film aluminum tape bonded to the top case. The BMS boards attach to the aluminum tape via thin ultrasonic bonded wires. It is very difficult to solder to the aluminum pads, so even if you don’t use the BMS, leave it in place so you can connect your BMS wires to the conventional solder pads on the board.

Technology Electronics Machine Vehicle


The factory connection points for the bus bars/cable that connect the modules can be cut down and new holes drilled to shorten the modules - however beware that the material is laminated aluminum so you must be careful when cutting/drilling so you don’t separate the laminations.
The only structural part of the module casing are the sides where the factory hold downs are. The modules must only be supported from the sides and cannot be allowed to rest on the bottoms. The bottom is exposed foam insulating material that covers up the fusible links.

Architecture Metal Glass Steel


The only structural part of the module casing are the sides where the factory hold downs are. The modules must only be supported from the sides and cannot be allowed to rest on the bottoms. The bottom is exposed foam insulating material that covers up the fusible links.
The hold down points on each of the modules are in different positions and except for the small modules are difficult to attach to. To “simplify” matters, we elected to cut off the Tesla mounting points and bond aluminum angle down the sides. We do not recommend this as the plastic is difficult to bond to! We also learned that at elevated temperatures, the strength of most epoxies is shockingly low. We ended up having to use an expensive 3M structural epoxy, DP820.

We used Misumi 20 series extrusion to build a framework around the modules. The framework was sheeted in aluminum. The bottom is steel tube with a Carbon-Kevlar panel from penetration resistance.
Total weight of the finished pack and enclosure is 850 pounds. Exterior dimensions are 77” long x 13.5” wide x 18” high.

Land vehicle Vehicle Car Formula libre Auto part


Vehicle Automotive exterior Auto part Technology Computer cooling
See less See more
  • Like
Reactions: 1
1 - 15 of 15 Posts
Wow, it's beautiful. Thank you for gifting the community with this information!

From the car shot, it looks like part of the pack is actually in the passenger compartment.
Look at that battery box! Thanks for all the specs. I'm bummed that Tesla is moving away from the smaller modules.

If 27kwh was almost good enough, couldn't you have gotten by with two Tesla modules?
Great step in a great build! In a previous post there was mention of introducing AC based cooling to the battery - did that pan out?
Wow, it's beautiful. Thank you for gifting the community with this information!

From the car shot, it looks like part of the pack is actually in the passenger compartment.
Yes, all three battery configurations for this car place the rear end of the battery pack beside the driver, and angle the front of it toward the middle of the car. This is feasible with a competition-only vehicle like this.
Each module is a nominal 90V, so 2 modules would have only given us 180V.

The chiller is dedicated to motor cooling. It works fine, but preliminary data suggests that no matter what the coolant temperature, the stator temp continues to build under hard use. Unfortunately, we can only get in about 5-6 hard laps before we are power limited. We will be moving to a Model 3 motor as soon as there is a good control solution.
Each module is a nominal 90V, so 2 modules would have only given us 180V.
That's a rough approximation, right? The short modules are 23S and the long ones are 25S (as you described earlier), so they're nominally 86 V and 94 V.

The enormous size of these modules - both their length (because each is the entire length of the under-floor pack which fills as much as possible of the Model 3 wheelbase) and their total mass and volume (because there are only four in this high-capacity pack) make them very hard to package in anything but the intended location (under the floor of a big sedan)... and all four are needed to reach typical 360 V operating voltage (without internal modification of the modules). There might be some vehicle which could make good use of just one in a project with a low-voltage motor. :unsure:
Really impressive project. You note that when you started the project there were no know means to communicate with the BMS. Has that changed?
No open source or published info, but EVTV and Simon Andre have solutions for sale.
Did you guys ever find who makes that snap fit coolant connector Tesla uses to connect to the modules, or did you manage to get all yours off without damaging any?
No, we are using the OEM connectors that we got with the pack. I could not find a source.
No, we are using the OEM connectors that we got with the pack. I could not find a source.
I have been searching this weekend, I think I just identified the manufacturer (NORMAQUICK V2 series) but I haven't found a source yet. I take it then you were able to get them all off without breaking the tabs, or did you come up with another means of securing them into the modules?
It's very easy to remove the fittings. Just gently open up the two tabs with screwdrivers, or even your fingers.. The connectors have an o-ring seal and just slide right off. To re-use the connectors, we cut off the corrugated tubing and directly attached silicone hose via hose clamps.
It's very easy to remove the fittings. Just gently open up the two tabs with screwdrivers, or even your fingers.. The connectors have an o-ring seal and just slide right off. To re-use the connectors, we cut off the corrugated tubing and directly attached silicone hose via hose clamps.
Good to know you had success. I tried to use something close to the technique you described, and the tabs were very brittle... I was working in February inside an unheated garage, which may have been my problem.
1 - 15 of 15 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top