🌡️ Tested: Did Rivian Fix Charging Thermal Issues in 2023.2.3? With CAN Data!

[SASSquatch]

@branden just expressing my gratitude for both the content of your video and that you avoided making it 3 times longer than it needed to be. Hopefully you serve as an example to other YouTubers who don't seem to understand that concept.

Encouraging that Rivian is tweaking things, even if the results aren't impactful yet. I suspect as they have more data from charging sessions across the R1Ts and R1s that are on the road they will be able to make more impactful trade-offs.

Safety has to be a priority and I'm sure that they are hesitant to push things too far in either direction without more user data.

Sponsored

 

[Cosmacelf]

I will be very interested to see what happens to these tests when the weather turns warmer. My suspicion is that cold winter temps are having a big effect - my theory is that the lower layer of battery cells is much colder than the top layer, yet they have a single shared cooling plate in between. The lower/upper cell temperature difference won't be so pronounced in the summer.

Has anyone done similar tests located in Florida or some other warm weather state?

 

[BigE]

Branden, do you known how far into the charge that cliff drops off? Is it still ~ 30% SOC? Thanks for the data and all your work!

 

[SeaGeo]

I've been saying for months since I've been watching this that Rivian needs to kick in battery cooling much sooner. It had been doing nothing till the pack hits 50C, then immediately slowing to ~100kW which slowly ramping the compressor up to full to cool the pack.

The new update still does essentially the same, except it drops to 150 instead of 100kW. The full speed compressor still does not kick in until after the pack is too hot.
HOWEVER after two charges yesterday, and I think also shown in this video, I DO think rivian is running the compressor at a slower speed starting at a lower temperature. I swear this is new, I noticed the truck shimmying during both charges yesterday where previously it was dead still. So that may also help slow the heating slightly.

Branden has been logging just the data he is viewing in a format without timestamps. I am logging full raw candata and will be looking at the logs over the next few days. I still can't find coolant info but I continue to search. But it is awesome that branden could manage eliminating several values both with a reliable charger and charge rate (something I can rarely find) as well as near identical initial pack temps (forget that with the new england temperature swings lately!)

Just a shout out here, thanks for all the time you've dumped into this. @branden gets the clicks, but I know there was a hell of a lot of effort to get to that point.

 

[jwardell]

Branden, do you known how far into the charge that cliff drops off? Is it still ~ 30% SOC? Thanks for the data and all your work!

The cliff occurs when max pack temp rises above 50C. Nothing to do with SOC.

 

[seatosummit]

This is a pretty technical video, but essentially I’m comparing the charge taper curve between new and old software and how they’re managing the heat of charging. New software is marginally better.

Thanks for the deep dive. Love this stuff.

Look forward to consumer release of CANserver. I parse CAN logs regularly for work, so I should probably see if there is something I can contribute if the project is open source.

 

[jjwolf120]

The evidence does not support your conclusion. Once cooling starts, it appears to be very effective.

 

[zefram47]

The evidence does not support your conclusion. Once cooling starts, it appears to be very effective.

Yes and no. Yes it's capable of cooling part of the pack, but the temperature gradient within the pack is quite high and that's a problem that software won't fix.

 

[jjwolf120]

the temperature gradient within the pack is quite high and that's a problem that software won't fix.

Is it high? What is considered high? I'm no expert on the allowable temperature gradient within the pack, are you? No evidence was presented of what the recommended temperature gradient is.

 

[zefram47]

Is it high? What is considered high? I'm no expert on the allowable temperature gradient within the pack, are you? No evidence was presented of what the recommended temperature gradient is.

The data shows enough. Full cooling with the max temp dropping and they still don't return to max delivered power. Remember, the cells can get too cool for high powered charging as well as too hot. If some cells are hitting that 50C max temp while others are going below the min temp for max power, then you derate. Whether that's simply Rivian's decision for longevity or an actual cell manufacturer limit doesn't really matter. The gradient is out of spec, so they're derating and simply running the cooling earlier/more will not help.

 

[jjwolf120]

The gradient is out of spec, so they're derating and simply running the cooling earlier/more will not help.

What is spec?

I think you are arriving at a conclusion before all of the evidence has been gathered. The battery might not be returning to the previous charge rate due to the charging curve rather than thermals.

 

[Cosmacelf]

I don't think we have enough data to know what is really going on. I would love for someone to do similar measurements in a warm weather climate and maybe that would give us more data to form a conclusion.

 

[branden]

I don't think we have enough data to know what is really going on. I would love for someone to do similar measurements in a warm weather climate and maybe that would give us more data to form a conclusion.

Charlotte should be high 70s late next week - I’ll try and get more data then.

Sponsored