P90D Supercharging Data

[apacheguy]

So something odd was happening between 15-35%. The power reading seemed very noisy from large fluctuations in the amperage. Can't say I've ever seen this before. At any rate, I'm still confused as to how the 90 kWh charges much slower than an 85 kWh for any given SOC.

 

[Pete90D]

So something odd was happening between 15-35%. The power reading seemed very noisy from large fluctuations in the amperage. Can't say I've ever seen this before. At any rate, I'm still confused as to how the 90 kWh charges much slower than an 85 kWh for any given SOC.

I started a post asking about the amperage fluctuation that I saw at the end of a charge the other night. The dash showed it fluctuating a lot faster than on the touchscreen.

Power Fluctuation

 

[hiroshiy]

As apacheguy said, the taper curve seems to be steeper than 85. I have E pack 85, and at 72-75% SOC the kW crosses 50kW. Pete90D's data seems at 70% already lower than 50kW. Different chemistry?

 

[Pete90D]

As apacheguy said, the taper curve seems to be steeper than 85. I have E pack 85, and at 72-75% SOC the kW crosses 50kW. Pete90D's data seems at 70% already lower than 50kW. Different chemistry?

Would heat be an issue? If they put more into the same space then maybe they are tapering differently for heat reasons.

 

[Khatsalano]

Well done. Keep investigating. This is pretty in-line with the knowledge base before the 90 packs came out.

- K

 

[Troy]

I compared this data (in message #15) to data from other supercharge sessions. I created a Google document here. Anyone can edit it. This chart is from that document. I think Tesla uses 3 different pre-programmed tapering curves. This time Pete's car selected the lower curve. Links to the data sources for the other sessions are in the Google document. I added 3.94 minutes to the beginning of Pete's session to compensate for the missing first 8%.

 

[apacheguy]

IMO, each battery pack rev has it's own taper curve (only one). However, there exist preprogrammed offsets in the event cooling issues arise and the pack is not able to maintain the desired set point. I would find it enormously interesting if Tesla released the basics of their SpC logic.

 

[Pete90D]

I talked to a technician today about the fluctuation in amperage and he didn't have an answer so he's going to see if an engineer will reach out to me. I'll update when/if I have more information

 

[apacheguy]

So apparently the latest SW updates have been targeting the BMS among other things. Might be worthwhile to check if Tesla's gotten around to fine tuning the SpC taper for the 90 kWh.

 

[Pete90D]

So apparently the latest SW updates have been targeting the BMS among other things. Might be worthwhile to check if Tesla's gotten around to fine tuning the SpC taper for the 90 kWh.

Thanks for the heads up. I may not have a chance to record it for a couple weeks, but if I do I'll post the data.

 

[apacheguy]

So apparently the latest SW updates have been targeting the BMS among other things. Might be worthwhile to check if Tesla's gotten around to fine tuning the SpC taper for the 90 kWh.

There are plenty of 90 kWh (90D and P90D, doesn't matter) out in the wild and yet we only have one SpC taper curve from Pete. Pete's done enough already for our community. How about we get some others pitching in here?

 

[jeffro01]

When I get my S90D in a couple of weeks I'll be happy to help, you'll just have to tell me how... LOL

Jeff

 

[apacheguy]

Instructions:

1. Arrive at a SpC with sufficiently low state of charge (35% or less).

2. Connect to SpC and log the session in your favorite app that interfaces with the Tesla API such that you track volts, amps, and kW every 2 minutes or less for the duration of charging.

If I lost you on 2, simply take screenshots of the "Charging" screen on the mobile app for every % change in SOC.

3. Post data here on TMC.

 

[jeffro01]

Roger that, will be happy to do.

On number 2 since yes, I was lost... (although I shouldn't be), do you have any recommendations for an app that'll keep me from having to take screen shots?

Thanks,
Jeff

 

[apacheguy]

I'm the author of Tesla Mobile, which you can find here:

REST API Tools and Apps - Tesla Motors Club - Enthusiasts & Owners Forum

It's the only app I've ever used to interface with my car so that's all I can speak to. Alternatively, you may simply grab a video of the center console as shown below for an 85 kWh:

Supercharge Pack Rev E - 120kW supercharger - No Audio - YouTube

 

[liuping]

I'm going to try do a test tonight. I should be down around 20% (I've skipped charging for a few days now), and I'll charge to 90%, with visible Tesla graphics the data.

It will be probably around 10pm, so there should be few other people there. The outside temp should be around 58 degrees, so cooling should not be a factor. I'll leave the cabin heating off.

Any thing else I should consider?

 

[apacheguy]

Sounds good. Just as long as you make sure VT is logging volts, amps, and power with sufficient frequency. IIRC, for some reason VT polls charge state infrequently.

 

[liuping]

Sounds good. Just as long as you make sure VT is logging volts, amps, and power with sufficient frequency. IIRC, for some reason VT polls charge state infrequently.

I'll make sure VT is in do not sleep mode.

I'll also do the low tech backup of logging the time of each 1% gain on screen.

 

[liuping]

I ran the test last night. I arrived with 16% and charged until 90%. It took 1 hour and 4 minutes. Outside temp was 61 degrees. There was on one other car at any given time at the sight (and they were not using my paired charger)

Visual Tesla did indeed do the sampling inconsistently. It did most about 4 minutes apart, then some just one second apart. However, I also manually logged the Time, Volts and Amps each time the % went up by one, so I think I have all the data we need.

I'm not sure the best way to add it to the existing data, since I have sample data at each % increase, and the graph data is per minute. Should I have excel plot a graphic and then resample the curve at the existing time stamps used in the google doc?

I can try to do that tonight (I'm not great at Excel, but me wife is an expert), or if anyone wants to give it a try, I put the CSV data here: https://www.dropbox.com/s/a2c3a4rmbwzkabi/supercharge.csv?dl=0

 

[apacheguy]

A couple things:

1. I assume the data point at 23 % was a typo since that = 150 kW. We've never seen rates above 120 kW.

2. You're 90 kWh seems to have followed the less aggressive taper that was observed by Pete90. I'll need to take a closer look for sure, but just eyeballing it, your data indicate slower charge rates for a given percentage compared to my 85 kWh.