If there's not a cheap logger on the market by the time I finish my BMS and one other project (month or so) then I'll make a Tesla specific one to sell at near cost for folks here, probably < $100, assuming we can find the new connector part number and pinout and all too.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Chassis CAN Logging To ASCII Text Plus Graphing
- Thread starter lolachampcar
- Start date
-
- Tags
- Driving Dynamics Model S
WK,
If my approach is acceptable, I can set you up with some PCBs (found a bunch in my closet) along with a lot of the BoM on reels. I think I even found a rack of fifty processors......
If my approach is acceptable, I can set you up with some PCBs (found a bunch in my closet) along with a lot of the BoM on reels. I think I even found a rack of fifty processors......
whitex
Well-Known Member
NICE! Thanks MikeBur for good data.
Quick question, at the end of your runs, I see the current starting to taper off faster while the speed keeps on climbing (so I assumed your accelerator foot was still firmly on the floor), were you seeing any limits on your instrument cluster? Is this due to battery or other components heating up and needing to cool down to provide the >1500A?
The only negative I see now is that my upgrade is not for another 3 weeks. *sigh* But I guess there are people here who still don't have a date, so it's not so bad relatively.
Quick question, at the end of your runs, I see the current starting to taper off faster while the speed keeps on climbing (so I assumed your accelerator foot was still firmly on the floor), were you seeing any limits on your instrument cluster? Is this due to battery or other components heating up and needing to cool down to provide the >1500A?
The only negative I see now is that my upgrade is not for another 3 weeks. *sigh* But I guess there are people here who still don't have a date, so it's not so bad relatively.
MikeBur
ManualPilot
The reason I plot TPS (that thin purple line is to show when (and at what %age) the accelerator is at. The compare chart is all 100% throttle pressed - just check the thin purple line.
I did manage to get max battery status to go from "Ready!" to "Cooling..." at one point. I believe unrelated, though does show how much you can heat these things with spririted testing
- - - Updated - - -
Last edited:
We've been puzzling over this from several pages back, where the current tails down from it's peak around 6 to 6.5 seconds after first motion. Several theories and ideas have been proposed, but after seeing the L update have proven not to be: voltage-limit triggered, 6 seconds time triggered, 80mph-speed-limit trigger, motor back emf, etc.
i think you are on to something with a thermal component reaching temperature that is triggering the limit to pull back the current, e.g. the igbt's in the inverter, the stator windings in the motor, ???.
p.s. Mike, did you ever weigh yourself and the car.
pss Interesting how slowly it seems the current falls when the TPS goes to 0--like .25 seconds its still full amps whle the foot is off the pedal.
Last edited:
MikeBur
ManualPilot
Yes, looks like this is a brain-fart, about to delete these out. Just noticed as I was creating a merged one. As my only defense, the auto-generation of the min/max takes the whole run into account - not the filtered graph. Sigh, sorry - will redo
MarcG
Active Member
Great stuff Mike, thanks again! Yeah getting the SoC and battery temp as close as possible would be ideal, but this is already a great start.
Would it be possible to overlay the relevant parts of graphs 1 and 3?
Would it be possible to overlay the relevant parts of graphs 1 and 3?
MikeBur
ManualPilot
Ok, this is better - still not perfect with varying temps / SoC, though try this:
Slightly differing start and end speeds, though full data for transparency - SoC, bat Temp, etc for Insane on left, Ludi on right, dotted for Insane, translucent for Ludi again. mph on left vertical axis, rest on right axis
Ok, now to cheat a little - aligning on same starting speed (gives advantage to Insane a little) and aligning the time, then zooming in a little gives this:
Still searching for better aligned data at these speeds Marc, though hopefully gives a feel for the difference. There's obviously some squinting required... though I'm going cross-eyed already looking at all this data, so that likely helps me, lol.
- - - Updated - - -
No. Truck scales always seem to be closed when I pass one, and haven't prioritized finding one. Everything for me is relative - ie. I still have the same "stuff" in the back of my car for the last week+. I'm hopeful I haven't gained (or lost, yeah right) much weight personally in same time.
On second point, yea, makes me wonder if there's need for energy use ramp down going from full power to full regen. Though I would be speculating.
Cool!
Although, for production runs like that I go to a friend who works for a company that has a turn-key setup. I give them the PCB design and BOM, they source it all (either from me or whoever), bill me at cost for the BOM items, and build however many I want, from 1 to 1,000,000.
Generally only cost effective for 100 or more though.
MikeBur
ManualPilot
MarcG
Active Member
That's a great one, thank you so much Mike!
So if I interpret the graph correctly, it looks like insane did 44-75 mph in 2.8 seconds, and ludicrous did 43-75 in 2.4, or to interpolate: 44-75 in 2.3, then which correlates nicely to my 50-80 runs where ludicrous had a 0.5 second advantage.
Glad to see real data behind it all! :biggrin:
So if I interpret the graph correctly, it looks like insane did 44-75 mph in 2.8 seconds, and ludicrous did 43-75 in 2.4, or to interpolate: 44-75 in 2.3, then which correlates nicely to my 50-80 runs where ludicrous had a 0.5 second advantage.
Glad to see real data behind it all! :biggrin:
MikeBur
ManualPilot
That's the way I read this too. Of course, I'll try to reproduce the insane one better tomorrow, i.e. Cold battery and set out with 90% SoC and close to 49mph I can accelerate from. I wouldn't expect much difference though
MarcG
Active Member
D'oh I just realized I wrote 44 when it should have been 49mph (and 43 instead of 48).. my bad!
One thing I found to help was to set TACC to a certain speed (in my video, it was 50mph for one test then 60 for the second) then floor the accelerator pedal when ready. That way you're sure to have a consistently repeatable starting speed.
Nice graphs. Don't suppose you have the torque plots for the front and rear motors? (This is where I'd expect the biggest difference between the 85D and the P85D)
(I have the C# source for those messages if that's what you are using)
llavalle
Member
Haven't had the time to code that part of the parser, yet. I am using C#, send me your code, I'l plot the torque.
Very nice graph, I would love to see the same graph with a 85D vs Insane, although I know that can not be done with the same car
I still think it would be interesting to see what difference there is between a 85D and P85D accelerating from 50 - 70 mph.Doh! Just realised I only have implemented the messages for rear torque, front torque should be similar, but I don't have access to a gvret dump of an AWD car to test.
Code:
public class RearDrive : BaseMessage
{
public double ThrottlePos;
public double RearTorque;
public override void Init(byte[] data)
{
ThrottlePos = data[2] * 0.4;
var tqA = data[5] + ((data[6] & 0x1F) << 8);
var tqB = 512 * (data[6] & 0x10);
RearTorque = (tqA - tqB) / 4;
}
}