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

Differences Between Roadster and Model S Cells

This site may earn commission on affiliate links.

TEG

Teslafanatic
Moderator
Aug 20, 2006
22,101
9,490
Perhaps Tesla chose the cells they did for the 3.0 pack to minimize the cost and maximize reliability and longevity. But we likely will never know their reasons.

I suspect that they picked cells with somewhat similar electrical characteristics to the old Roadster cells. For instance, Model S cells might have a voltage range not totally compatible with the old Roadster electronics. I think they wanted a "solution" that didn't involve having to change other hardware.
 
I suspect that they picked cells with somewhat similar electrical characteristics to the old Roadster cells. For instance, Model S cells might have a voltage range not totally compatible with the old Roadster electronics. I think they wanted a "solution" that didn't involve having to change other hardware.

I think that's exactly it. The Roadster electronics couldn't work with the Model S cells due to differing electrical characteristics, i.e. voltage range. So they had to choose a different chemistry; otherwise they'd have to replace the rather expensive PEM.
 
I think that's exactly it. The Roadster electronics couldn't work with the Model S cells due to differing electrical characteristics, i.e. voltage range. So they had to choose a different chemistry; otherwise they'd have to replace the rather expensive PEM.

Frankly I would be shocked if that were the case. First, is there any evidence at all that the Model S cells operate in a different voltage window? Second, do we really think that Tesla built sophisticated electronics that couldn't handle a slight voltage change? I'd rather hope Tesla anticipated the likelihood of future cell chemistry improvements that might have different characteristics, since their whole business premise is based on cell chemistry improvements. A simple software/firmware update should take care of any issues, if there were any. If I change the capacity of the cells in my home built EV I just reprogram the gauge with the new size. If the voltage changes I can change that as well. I think some of you are making this much more complicated than it really is.
 
The voltage range of lithium cells used by both the Roadster and Model S should be the same as they are basically the same chemistry. There's no reason to believe that voltage differences would prevent the use of Model S cells in the Roadster and it's not like they are using lithium iron phosphate which has a significantly lower voltage range than the lithium chemistry used by the Roadster or Model S>
 
No, they are quite different chemistries. Roadster is Lithium Cobalt Oxide, whereas Model S uses Lithium Nickel Cobalt Aluminum. These have very different characteristics; for example, the Model S batteries are more tolerant of higher temperatures than the Roadster's batteries (Model S cools to 60C, Roadster to 40C)
 
Last edited:
No, they are quite different chemistries. Roadster is Lithium Cobalt Oxide, whereas Model S uses Lithium Nickel Cobalt Aluminum. These have very different characteristics; for example, the Model S batteries are more tolerant of higher temperatures than the Roadster's batteries (Model S cools to 60C, Roadster to 40C)

I fail to see how a more tolerant higher temperatures cells would be an issue, as if anything roadster will then keep the cells in a better temp environment than the MS.

Yes chemistries are different, but voltage top and bottom cutoff seem to be the same, to me the only thing at that point is having the proper SOC to Voltage mapping done in the firmware and updating the default CAC AH.

Roadster firmware seems pretty robust as I've connected Ultra capacitors, Lifepo4, and normal Lion as a replacement ESS Brick in the roadster without too much of a fuss (though I monitor the voltage on the brick manually vs the BMB/BMS).
 
My point was that the max/nominal/min voltage of both LCO AND NCA is the same. Charge both to about 4.1V max and don't discharge below 3.0V. A few parameter tweaks to the BMS should be a it takes to optimize for either chemistry.

Edit: spaceballs beat me to it.
 
At minimum they need a new SOC to voltage map.

My wild guess would be given the Model S cells are custom made (I recall them using a different cell cap than typical), you can't just drop in the cells into a Roadster. The Roadster used cells with a PTC (positive temperature coefficient) and CID (current limiting device) and the Model S cells don't have those. That's probably why they didn't use the Model S cells.

Just out of curiosity, I looked up the datasheets of the standard versions of the cells in the Model S (I know the NCR18650B is not the closest cell, but it is the same capacity and has a PTC unlike the other versions) and the Roadster. It seems the Roadster cells is less susceptible to lower temperature voltage sag. Given we know the Roadster doesn't have any anti-bricking protection (while Model S does), that may be a relevant difference too.
http://na.industrial.panasonic.com/sites/default/pidsa/files/ncr18650b.pdf
http://www.wickedlasers.com/image/wicked/pdf/SANYO-UR18650A-2.2Ah Specifications.pdf
 
Last edited:
... The Roadster used cells with a PTC (positive temperature coefficient) and CID (current limiting device) and the Model S cells don't have those. That's probably why they didn't use the Model S cells. ...

You sure about the Model S cells? I just took a look at a Model S cell on my desk, looks to me like the CID is present, also you can see the C shape disk vent.

... Given we know the Roadster doesn't have any anti-bricking protection (while Model S does) ...
I think that was a design choice at the time, and now they know better to not exclude this feature.
 
Last edited by a moderator:
Tesla used to boast that the Roadster cells had individual fuses at the anode and cathode of every cell.

The Model S is the same.

- - - Updated - - -

You sure about the Model S cells? I just took a look at a Model S cell on my desk, looks to me like the CID is present, also you can see the C shape disk vent.

I think that was a design choice at the time, and now they know better to not exclude this feature.

I've confirmed by blowing up several Model S cells since my initial mishap when shorting one that they seem to have little or no protection. All of the cells so far that have gone up in flames have burst from the negative end of the cell where there is a small circular score in the metal casing. This seems to be the weak point in all cases and it pops at this spot like opening a can. You can see this in the pic I posted in my old thread from when I first accidentally shorted one and it went off like a rocket...

This is under the stress of my high-load and high-charge rate testing that I've been doing where my goal is essentially to test the limits, so, not something I'd expect to happen in normal use. Many cells will still need to be sacrificed in the name of science before I'm done. lol. (Testing lately has been outside, inside a steal enclosure...)

Eventually I'll get around to compiling all of the video and data for this. :p
 
Take a model S cell apart and compare to a standard laptop or Roadster cell with both "fuses." You will find one has been replaced with a solid metal disc. The equivalent of a "penny in the fuse box."

The model S cells are not drop in replacements for a Roadster pack.

GSP
 
Last edited:
The Roadster is older technology. They've learned a lot since they built it. There's no reason to expect they'd have foresight to design it for batteries that hadn't been developed yet.

I'd say there is every reason to expect that, since it's so easy to do, especially when the new technology has the same operating voltage as the old, so essentially it looks like the exact same cell to the electronics. I agree with those who think it's the internal physical construction that prevents them from simply dropping in Model S cells, and why when they announced the Roadster upgrade they said something to the effect of "We've found a new cell that allows a higher energy density Roadster pack to be built". We know they aren't Model S cells, not even "B" grade Model S cells, since those don't have the same internal protection as the Roadster cells. They are probably NCA cells with the same internal construction, PTC, CID, as the original Roadster cells.
 
Yes ma'am.

- - - Updated - - -

My point was that the max/nominal/min voltage of both LCO AND NCA is the same. Charge both to about 4.1V max and don't discharge below 3.0V. A few parameter tweaks to the BMS should be a it takes to optimize for either chemistry.

Yes those appear to be similar. That said, Tesla has never confirmed exactly what chemistry is in the Roadster pack. Some claim it is actually Lithium Manganese Oxide (see https://en.wikipedia.org/wiki/Tesla_Roadster).
 
You sure about the Model S cells? I just took a look at a Model S cell on my desk, looks to me like the CID is present, also you can see the C shape disk vent.
I stand corrected then (of course it is still possible that the underside is missing the insulation and second layer, but at least the outside looks like a CID). I was recalling from memory about the patented cell design (I will have to look up the actual patent).
 
OK I carried on in the other thread...

I get that they can't put a Model S cell in the Roadster pack as they have been cost-optimised with the removal of some safety features, but why could they not put in the off-the-shelf NCR18650B?
 
OK I carried on in the other thread...

I get that they can't put a Model S cell in the Roadster pack as they have been cost-optimised with the removal of some safety features, but why could they not put in the off-the-shelf NCR18650B?
NCR18650B (which at 12Wh per cell would mean 82kWh) is not power optimized. The advertised peak discharge rate is 2C for that cell, which means 164 kW, while the Roadster peaks at 225kW. Something like the NCR18650PF (10.44Wh per cell so 71kWh) might be closer given it is advertised at 3.5C (10A), which means 249kW, enough to cover the Roadster's peak power demands with a little bit of margin.

That's my best guess (other than low temp voltage sag I mentioned in other post).
 
Something like the NCR18650PF (10.44Wh per cell so 71kWh) might be closer given it is advertised at 3.5C (10A), which means 249kW, enough to cover the Roadster's peak power demands with a little bit of margin.

I think you might have found it. The pack size of roughly 70kWh and 30% more energy is exactly in line with the Tesla blog. Matching the Roadster C-rate makes a lot of sense.
 
NCR18650B (which at 12Wh per cell would mean 82kWh) is not power optimized. The advertised peak discharge rate is 2C for that cell, which means 164 kW, while the Roadster peaks at 225kW. Something like the NCR18650PF (10.44Wh per cell so 71kWh) might be closer given it is advertised at 3.5C (10A), which means 249kW, enough to cover the Roadster's peak power demands with a little bit of margin.

That's my best guess (other than low temp voltage sag I mentioned in other post).

However if you look at the spec sheet for the original Roadster cell (Sanyo UR18650F), that also states a peak discharge rate of 2C. Tesla were obviously taking their chances that drawing >100kW wasn't going to happen for more than a few seconds of hard acceleration and they could get away with it :wink:

So taking the same philosophy, if they used the NCR18650B then they would be going above 2C for less of the time.

If you take the P90D in ludicrous mode, that would be drawing 6+ C. Yes those are "automotive optimised" cells, but I doubt that includes a step change in power density from the NCA chemistry.