-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
[updated with *] P85D 691HP should have an asterisk * next to it.. "Up to 691HP"
- Thread starter krisg81
- Start date
-
- Tags
- Driving Dynamics Model S
Andyw2100
Well-Known Member
As a very general question, how accurate are you finding the PerformanceBox data as compared to the data the track provided?
Time and speed wise they seem pretty close. I think the longitudinal acceleration data could use some work, though... the graph is all over the place, but the acceleration is smooth...
Here's a video I whipped up real quick of one run:
Tesla Model S P85D vs Nissan Skyline GTR - YouTube
Last edited by a moderator:
OK, so, here is the PBox graph of that run:
Although, I'm confused a little. (Maybe it's just late.)
The times and distances match up pretty much perfectly with my track slip... if I subtract the reaction time on the slip (0.516s) from the PBox numbers...
(Seems incorrect, see edit at bottom)
60' Time
PBox: 2.16s
Track: 1.626s
Difference: 0.534s (103.5% of reaction time)
330' Time
PBox: 5.32s
Track: 4.788s
Difference: 0.532s (103.1% of reaction time)
1/8th Mile Time
PBox: 7.96s (@ 94.82 MPH)
Track: 7.356s (@ 95.003 MPH)
Difference: 0.604s (117% of reaction time)
*shrugs*
I'm not sure what to make of it, honestly.
At 0.51 seconds the PBox data says I've moved 2.75' and am traveling at 7.85 MPH...
I'm not sure what to make of this data now. heh. Ideas? Is the track really subtracting my reaction time from the actual data? Would seem weird. I'm more inclined to blame the PBox.
Edit: Scratch the reaction time theory. Found a run that I had a reaction time of 0.224s and tested that theory and found the same ~0.5s discrepancy... now definitely confused.
Edit 2: Going to make another thread about this so I don't derail this one...
Although, I'm confused a little. (Maybe it's just late.)
The times and distances match up pretty much perfectly with my track slip... if I subtract the reaction time on the slip (0.516s) from the PBox numbers...
(Seems incorrect, see edit at bottom)
60' Time
PBox: 2.16s
Track: 1.626s
Difference: 0.534s (103.5% of reaction time)
330' Time
PBox: 5.32s
Track: 4.788s
Difference: 0.532s (103.1% of reaction time)
1/8th Mile Time
PBox: 7.96s (@ 94.82 MPH)
Track: 7.356s (@ 95.003 MPH)
Difference: 0.604s (117% of reaction time)
*shrugs*
I'm not sure what to make of it, honestly.
At 0.51 seconds the PBox data says I've moved 2.75' and am traveling at 7.85 MPH...
I'm not sure what to make of this data now. heh. Ideas? Is the track really subtracting my reaction time from the actual data? Would seem weird. I'm more inclined to blame the PBox.
Edit: Scratch the reaction time theory. Found a run that I had a reaction time of 0.224s and tested that theory and found the same ~0.5s discrepancy... now definitely confused.
Edit 2: Going to make another thread about this so I don't derail this one...
Last edited:
Does anyone know how PerformanceBox provides accurate speed information from GPS. GPS location accuracy is around 5 meters. So speed it not going to be accurate if computed from GPS location alone. Does it use doppler, i.e. frequency shift, of the GPS signal to determine speed? It must have an accelerometer to measure acceleration.
Danal
electricmotorglider.com
Absolutely yes, it has a 3-axis accelerometer. The better ones of these can take an accurate reading on all 3 axis 20 (or more) times a second... but they are NOT 'continuous'. Therefore, when the software integrates all those readings over time, errors creep in that make the absolute position of the car become less and less accurate. This is particularly important if the goal is to log the car's position as it circles an oval track or a road course.
Enter the 10 times a second GPS. To your point, GPS readings have an error radius, when you try to plot the absolutely against a map. However... those errors tend to change very slowly over time. In fact, the vast majority of literature on GPS errors talks about "pass-to-pass" accuracy, meaning the roughly 12 hours when a given satellite comes all the way around the earth. Two readings, taken 1/10 of a second apart, can be accurate to centimeters, relative to each other, and not a map.
So, yes, the secret of "track mappers" is lots of correlation between the errors inherent in "point" based accelerometer, and the GPS errors, and using each to compensate the other.
Cottonwood
Roadster#433, Model S#S37
GPS receivers use the doppler of the carrier frequency to compute speed. To state it simply, a GPS receiver needs at least 4 satellites to compute 3 axes of position and absolute time, and the GPS receiver also needs 4 satellites to compute 3 axes of velocity and frequency error in the local reference oscillator. Position and Velocity are related, but mostly independent, calculations. More satellites allow for more accurate calculations.
Also, the geometry of satellite positions matters. The best accuracy comes from satellites spread over the sphere surrounding the receiver. The worst is all the satellites near each other. This is why an external antenna is better; it can see a better spread of satellites. GPS nerds love to talk about GDOP (Geometric Dilution Of Precision); Dilution of precision (GPS) - Wikipedia, the free encyclopedia.
Went over a bunch of the runs I did with the PerformanceBox at the strip last night.
Assuming I'm calculating correctly in my spreadsheet I see spikes as high as ~600 HP and a bunch more in the mid to upper 500s.
I used 5050 lbs as the weight for my calculations. This is with no smoothing on the data.
Assuming I'm calculating correctly in my spreadsheet I see spikes as high as ~600 HP and a bunch more in the mid to upper 500s.
I used 5050 lbs as the weight for my calculations. This is with no smoothing on the data.
If you take an average from say 45-65mph (power limited region with low drag) what would the whp be? I think the spikes in whp are just due to the accel spikes the graph had earlier, especially if you also see downward spikes.
Unless your power meter went above 480kW although that seems unlikely
*EDIT:
Using your posted graph, 50-60 took 0.8s resulting in avg accel of 0.57g (constant). So at 60mph P=2300kg*5.59m/s/s*26.833m/s =350kW or 460whp. FWIW.
Unless your power meter went above 480kW although that seems unlikely
*EDIT:
Using your posted graph, 50-60 took 0.8s resulting in avg accel of 0.57g (constant). So at 60mph P=2300kg*5.59m/s/s*26.833m/s =350kW or 460whp. FWIW.
Last edited:
Putting this here so I don't forget. Based on drive unit failure discussion, I submit the rear motor is being nerfed at some point to prevent rear drive unit failures.
We know the P85 puts around 430-440hp on a rear wheel only dyno. I submit the P85D rear motor number does not. I think the only graph of a P85D showed 400HP at the rear wheels. Is that true?
We know the P85 puts around 430-440hp on a rear wheel only dyno. I submit the P85D rear motor number does not. I think the only graph of a P85D showed 400HP at the rear wheels. Is that true?
Actually that graph really is combined both front and rear and total was only 400 hp at the wheels. The 4 wheel mustang dyno links the front and rear drums together with a belt that forces them to rotate at the same speed.
That said, mustang dynos also are notorious for under reporting compared to Dynojets which are far more widely accepted. Based on how much Mustang dynos typically under report, that particular P85D run would have been around 430 at the wheels on a dynojet....all 4 wheels combined.
That said, I'll bet the SOC on that particular run was much less than 100% and probably much less than 80%. I wouldn't be surprised if it was really more like 50%.
What's remarkable is that there isn't a single P85D Dynojet run anywhere that is public. As a result, this thread rages on yet nobody has collected any real data.
Danal
electricmotorglider.com
Not only that... Drag Times had to turn traction mode off, and launch with a gentle foot, or the Mustang Dyno drums and Tesla tires would slip. So that makes it even worse. I also am astonished that there is not one single credible measurement of a P85D in public.
I am trying to find a Dynapak. If we find one within 500 miles or so of Dallas TX, I'm up for being the guinea pig.
I am trying to find a Dynapak. If we find one within 500 miles or so of Dallas TX, I'm up for being the guinea pig.
Kalud
Active Member
Interesting peak power comparison from a Classic P85 (on 6.2 build 2.4.153) 85% SOC
345kW (463hp)
To a P85D (on 6.2 build 2.4.153) 93% SOC
414kW (555hp)
I can't wait to test the latest "More insane, insane mode" from build 4.2.168, hopefully it also increases the Classic P85 power.
345kW (463hp)
To a P85D (on 6.2 build 2.4.153) 93% SOC
414kW (555hp)
I can't wait to test the latest "More insane, insane mode" from build 4.2.168, hopefully it also increases the Classic P85 power.
I documented 363 kW on my P85's first drive unit long ago.
Peak KW on S60, S85, and P85
I don't have P85D data to contribute yet.
Kalud
Active Member
But that was on 4.5 or 5.0. Many things could have changed since. And being a signature its most likely a Rev. A pack which is even more impressive. I wonder what you get now on 6.2.
So far on my P85 I couldn't get more than 345kW but I haven't tried above 90% SOC. Same could be said to the P85D, I haven't personally seen above 414kW peak power from the API.
Similar threads
