Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Pics/Info: Inside the battery pack

This site may earn commission on affiliate links.
Tony, this is the best evidence yet I've see of ~4.15 being Vmax for 100%. What's that a screenshot of?

It's from the Tesla powered Toyota Rav4 EV's BMS via CAN bus. We just decode it and put it in a display (again, it's all part of our "JdeMO" project to put CHAdeMO outlets directly on Tesla powered cars).

Like stated above (somewhere), 4.15v OCV is likely the cell voltage max, but SOC% is certainly variable with a "full" charge; about 96.X% to the highest observed of 99% in that photo.

All the other data I've posted (and which seems to be mostly discounted outright here as "doesn't apply") is from the Tesla BMS. It is live 24/7, whether the car is on or off.

4.202v for one update, then 4.200v continuously, is the absolute highest cell group voltage observed, during regen going down a big hill after a "full" charge (what Toyota calls "extended", like a Roadster "range" charge).

- - - Updated - - -

Model S is a lot more locked down by comparison, but it's just a matter of time before the collective intelligence of this forum figures it out.

Yes, maybe somebody will in fact post that stuff...
 
Last edited:
It's from the Tesla powered Toyota Rav4 EV's BMS via CAN bus. We just decode it and put it in a display (again, it's all part of our "JdeMO" project to put CHAdeMO outlets directly on Tesla powered cars).

Like stated above (somewhere), 4.15v OCV is likely the cell voltage max, but SOC% is certainly variable with a "full" charge; about 96.X% to the highest observed of 99% in that photo.

All the other data I've posted (and which seems to be mostly discounted outright here as "doesn't apply") is from the Tesla BMS. It is live 24/7, whether the car is on or off.

4.202v for one update, then 4.200v continuously, is the absolute highest cell group voltage observed, during regen going down a big hill after a "full" charge (what Toyota calls "extended", like a Roadster "range" charge).

- - - Updated - - -



Yes, maybe somebody will in fact post that stuff...
I would agree that the voltage will not go past 4.200 (With margin for error of a slight amount). This is why the regen power is reduced when close to a full pack. Ive even seen my car limit regen to a much lower level then the dotted line.
Now going higher then 4.2v speeds up degradation exponentially.
I'm sorry if I'm going off topic here, but thought this information would be helpful to some not as familiar with Lithium cells and longevity. I posted this link, How to Prolong Lithium-based Batteries - Battery University

about a year ago, and thought this will be a good place to show it again. Specifically would like to point out the voltage graph & Capacity VS Degradation chart. I've attached the picture below, but please visit the site (Dont want to plagiarize or break copyrights and want to give credit where it is due)

lithium2.jpg
Please note the increased degradation. at 4.3v, the cell will hold about 10% more power (10% higher capacity) however, service life is reduced by half compared to 4.2v, likewise 4.1v doubles the life of the cell. Good quote "Every 0.10V drop below 4.20V/cell doubles the cycle; the retained capacity drops accordingly. Raising the voltage above 4.20V/cell stresses the battery and compromises safety."

Forgive me my rambles, I'm on the Ny-Quil.
 
It's from the Tesla powered Toyota Rav4 EV's BMS via CAN bus. We just decode it and put it in a display (again, it's all part of our "JdeMO" project to put CHAdeMO outlets directly on Tesla powered cars).

Ah, ok thanks.

All the other data I've posted (and which seems to be mostly discounted outright here as "doesn't apply") is from the Tesla BMS.

Well if you are referring to my earlier post, my intent is not to necessarily discount anything, but simply to identify what's directly measured/observed behavior on the S, versus what's being extrapolated from other sources (or outright simply guessed at).

In this case, I'd suggest that a Rav4 of similar vintage to the S (as opposed to the older Roadster), is a pretty likely case for having similar behavior... so thanks for posting your data. Your experiences with seeing cells down in the 2.8V range is interesting as well.

4.202v for one update, then 4.200v continuously, is the absolute highest cell group voltage observed, during regen going down a big hill after a "full" charge (what Toyota calls "extended", like a Roadster "range" charge).

Am I understanding correctly that Vmax varied upt to 4.2V for one firmware update? Are your other pics demonstrating a Vmax of only 4.15V indicative of being a different firmware revision?
 
In this case, I'd suggest that a Rav4 of similar vintage to the S (as opposed to the older Roadster), is a pretty likely case for having similar behavior... so thanks for posting your data. Your experiences with seeing cells down in the 2.8V range is interesting as well... Am I understanding correctly that Vmax varied upt to 4.2V for one firmware update? Are your other pics demonstrating a Vmax of only 4.15V indicative of being a different firmware revision?

Yes, the Tesla part of the Toyota Rav4 EV project ended in about May 2012, with the Tesla Model S coming out the very next month. My original motor (replaced like on so many Model S cars, particularly those early ones) was Tesla serial #331. No, Tesla didn't reinvent the wheel for the Rav4 EV. It shares as many parts as humanly possible with Model S.

But, we do know the cells are the lower 2900ma cells, vice 3100-3400ma on Model S.

The cell voltages are the difference between fully charge OCV of 4.14v and maximum regen with a fully charged battery of 4.2v see the difference?

It's much like the data about charging. Sure, it likely charges at the equivalent of 4.2v (96 * 4.2v = 402v pack voltage), but the cells are more likely at 4.15v OCV, just like the Rav4 EV.
 
That's not how it works. The voltage rating of the fuse is the maximum isolation of the fuse, it has nothing to do with when it blows.

That is true but also consider the case of a short circuit, the fuse will blow faster on 690v than 400v as the current will be higher. For an equal overload current the voltage drop across the fuse is independent on system voltage, but in the case of a short circuit higher voltage means higher current. What you are saying is true because we on the same time vs current curve in either case. It is just in the case of a short circuit we are at different points on that curve. So is the blow time truly voltage independent in all cases? Yes and no.
 
But, we do know the cells are the lower 2900ma cells, vice 3100-3400ma on Model S.

Ok, so we know the cells at least are different. So your voltage measurements in a Rav4 setup are likely a good starting point for assumptions, but not direct observations of what the Model S BMS does with it's cells.

The cell voltages are the difference between fully charge OCV of 4.14v and maximum regen with a fully charged battery of 4.2v see the difference?

It's much like the data about charging. Sure, it likely charges at the equivalent of 4.2v (96 * 4.2v = 402v pack voltage), but the cells are more likely at 4.15v OCV, just like the Rav4 EV.

Ah ok, I wasn't sure what was meant by "update". Interesting that some regen is enabled even when the cells are at 100%.

Thanks.

It will be interesting to see how the overall system (pack + BMS) behaves when somebody is able to take the guts from a wrecked Tesla and charge a pack with it in some fashion where internal measurements are possible. Or makes a set of "extensions" for the HVDC and BMS connectors as well as coolant loops...
 
I've got a BMS from my Zero (logic buck switcher failure I suspect) where the part numbers are "readable". Let me know if you want to embark on a stand alone BMS solution to use the 85KW pack as is. If you do, I may be interested in pitching in with PCB layout and micro-controller development as I have all the tools (and a little free time :) ).

I'd also be interested in working with the group to solve the problems associated with using the pack in parallel with PV serial arrays of similar voltage. Specifically, how the MPPT stuff works when running off a low impedance source like the battery. It would also be nice to find alternatives that allow for using the MS pack as wired. I suspect this would also require an MPPT function when using PV to charge the battery directly.
 
I've got a BMS from my Zero (logic buck switcher failure I suspect) where the part numbers are "readable". Let me know if you want to embark on a stand alone BMS solution to use the 85KW pack as is. If you do, I may be interested in pitching in with PCB layout and micro-controller development as I have all the tools (and a little free time :) ).

I'd also be interested in working with the group to solve the problems associated with using the pack in parallel with PV serial arrays of similar voltage. Specifically, how the MPPT stuff works when running off a low impedance source like the battery. It would also be nice to find alternatives that allow for using the MS pack as wired. I suspect this would also require an MPPT function when using PV to charge the battery directly.
I have experience with running a Grid Tie directly from a lead acid battery pack. The MPPT function had no difference in performance with the battery pack. While it does make a difference with Solar, as the voltages and output fluctuate much more, with the steady current from a battery, no difference with the function on or off.
The problem I had was the cheap inverter I have to play with, didn't limit current. So while it was rated for 500 Watts, it would start low and ramp up, and keep going. I pulled the plug at 600 watts as I didn't want to fry the inverter. I need a way (resistors worked, but wasted WAYYY too much power as heat) to limit the current to keep it below 500 watts without such great loss, I know how to do this with AC, but not experienced in DC.
That may be a issue that one runs into with these setups.
 
My assumption is that you would be using battery for off grid operation. In this instance, you would have something like an 8 KW AC rated inverter supplying your house whose consumption would, presumably, stay below the constant and peak capabilities of the inverter. The inverter would also have to be of the type that does not require the presence of voltage to sync to (ala grid tie inverter) which means it would likely be more flexible on input type (in anticipation of this type of question).

All guessing and speculation on my part. Once I found that FPL's net metering was a no transaction fee or free "battery storage" solution I lost interest in the battery option.
 
All guessing and speculation on my part. Once I found that FPL's net metering was a no transaction fee or free "battery storage" solution I lost interest in the battery option.

This was my thinking as well, until I really thought about it and realized nothing prevented this from changing soon after I made the investment.
 
So, haven't had a chance to do much with this project the past few days.

I checked the pack voltage earlier and it has self-discharged only 0.3V since I charged it 5 days ago. Can't complain too much there.

(The main BMS board is still disconnected)
 
so I'm firing up my new PV array and some inverters are not cooperating.....
A quick internet search and I find some neat stuff to play with-
http://www.ti.com/lit/ml/sprt615c/sprt615c.pdf A full point/string inverter development kit from TI. This is where answers can be generated when it comes to seamlessly integrating a HV battery like the MS with a PV system.
http://www.ti.com/lit/an/snosb76c/snosb76c.pdf PV based battery charging development board with MPPT. This might be scalable to work with much higher voltages???
 
Well, finally gave up on keeping the pack in tact. One reason being I want the garage space back... hehe. But mainly because nothing off-the-shelf and even remotely reasonably priced (I found a solution that would work for roughly $150k...) works with this high voltage DC in a useful fashion. So, breaking it down into modules is the only plan that makes sense since it will essentially allow me to configure the pack for whatever voltage I need.

Anyway, started the full tear down tonight. Just going to reiterate: This thing is super heavy duty. Dismantling it is not simple.

I had to first drain the coolant loop. I couldn't find anything that would mate with the quick disconnect easily, so, I rigged up some PC liquid cooling tubing to each outlet, which fit somewhat snug inside them. I then fed air into one side and drained from the other until I didn't get any more coolant. About a gallon of coolant was liberated.

In order to remove the modules I had to first remove the entire spine where everything ties together... which consisted of a heavy gauge HV cable, dozens of bus bars, the BMS cabling, and all of it inside injected rubber insulation...

The modules would not lift out with the bus connections in place because you have to lift that side out first, slide it up and over, then disconnect the coolant loop connections, then lift the module out. It appears the modules sit on rails and are elevated about 1/4" above the the pack's inside floor... I have no way to easily do this outside of the pack, currently, so, I'm just carefully setting them down and out of the way sitting on their thin plastic covers. The plastic covers have ridged sections over top of the rows of cell fuses, so, they're reasonably well protected from damage even sitting like this. I plan to build a frame for them later, though, or at least some legs to keep them off the ground.

I stopped after removing all of the spine connections and six modules. Ten to go... another day.

I did learn that all of the modules are in fact identical. The driver and passenger side modules are just rotated 180 degrees. The top module in the front (the oddball) is just upside down. But they are in fact all the same, just their orientation is changed to suite its position.

I was kind of surprised to find a heavy gauge wire inside the rubber in the middle spine. The only things visible prior were heavy bus bars. So, I had assumed there was one which ran the entire length of the pack returning to the main contactors, but I was fooled! The exposed bus bar portion is welded to the heavy copper wire (2/0 gauge I believe) at both sides inside the rubber of the spine. The wire is 150 degree rated (C) and is *also* sheathed in a thick fire resistance material (the same material used all over the inside of this pack).

I'll sort through some pictures and post some later on.
 
Bummer. I was hoping against hope that this would progress as a HV integration into a full PV system. That would have been a lot of work and I do not blame you for a second making the decision to deal with bite sized chunks at reasonable voltages. My only fear for that path was the lack of cell level safety making having the modules exposed a serious hazard.




Anybody in the S. Florida (WPB) area interested in doing some experimentation/development of a high voltage MS pack integration into PV? My curiosity is killing me here.
 
Last edited:
One thing I didn't see in any of the pics -- where are the coolant loop connections made between pack and car? You showed the two electrical connectors (HV and LV) on the back edge of the pack (w.r.t how it is mounted under the car). Are the press-fit coolant connections on the front part/edge?

Th coolant loop is interesting to me b/c to support quick battery changes, there is no air-purge step between disconnecting battery #1 and reconnecting battery #2. Loop has to be designed to capture and deal with what little air is introduced during the swap...