Data Chellenge

Commuta_Busa

Commuta_Busa

Donating Member
Registered
Ok, I figured this might be of interest to some. I'll provide the answers in detail tomorrow. You've got all night to figure it out.

The trace shows two speed traces, one from the OE front sprocket location and the other from a GPS signal. Can you explain why they're so different at the start then align during the acceleration part of the run. During deceleration the wheel speed is slower than the GPS signal but then spikes up much higher. What is the cause for the inconsistant speed traces?

tmp_speed_challenge.jpg
 
dunno but maybe some kind of "electrical noise" HID's, secondary ignition, charging system? GPS looks pretty smooth all the way down the graph and the inductive pickup is showing lots of "snow"...

guessing you already know what is really wrong?

the point of departure looks pretty consistent on both ends... maybe some kind of signal multiplier coming or going? TRE? etc?
 
Last edited:
Im going to guess that one of the teeth if you want to call them that, on the sprocket senser broke off. Cant wait to hear the answer.
 
You are going to have a time delay between signal sent and signal received plus if selective availability is in play you will have a variance between actual position and relative position which will create what appears to be and "anomaly" between actual and reported.
 
Thanks for the responses. As an analyst I really get into this "stuff". I knew there would be others here that also like looking at this "stuff". :welcome:

Ok, now for the answer.

Mr Bogus said:
maybe some kind of "electrical noise"? GPS looks pretty smooth all the way down the graph and the inductive pickup is showing lots of "snow"...

the point of departure looks pretty consistent on both ends... maybe some kind of signal multiplier coming or going? TRE? etc?
Click to expand...

good observation, the MPH departure points are very similar on both ends. :thumbsup:

tofas2c said:
Im going to guess that one of the teeth if you want to call them that, on the sprocket senser broke off.
Click to expand...

You're heading down the right path but if it had broken a tooth then the trace wouldn't realign later on. :thumbsup:

Dr E said:
You are going to have a time delay between signal sent and signal received plus if selective availability is in play you will have a variance between actual position and relative position which will create what appears to be and "anomaly" between actual and reported.
Click to expand...

I do understand what you’re saying and the thought did cross my mind. I record this sensor at a rate of 50 hz which is a pretty good rate. If it were related to latency the output data would incrementally decrease in proportion to the increase in speed. The change to the speed trace is more abrupt.

===================

So here's the answer: above a specific speed the signal is only being captured 50% of the time per each rotation. This doesn't explain why the rear speed trace drops below the gps trace during the initial deceleration. This additional decrease is from the change in the tires loaded circumference.

Prior to these wacky logs I had to swap my sprocket cover/clutch slave support. I swapped it with a different companies cover. I'm sure alot of people have noticed that alot of the billet stuff doesn't match exactly with the dimensions of the parts they're replacing. Of course the mounting holes/tabs align but many other measurements are different. This new cover placed the sensor pickup a few extra mm away from the teeth on the sprocket's washer. The result of this is that the faster I go the less chance of the pickup capturing the signal correctly.

I've gone through all of the raw numbers. These are the approximate numbers but the exact numbers do have a greater than 96% level of confidence to support this. The rear speed trace is 100% higher than the gps trace up until 137 mph. At this point the two traces align which means I've lost 50% of the rear speed signal. As soon as the bike begins its deceleration the rear speed sensor drops an additional 7% and the rear speed trace regains its full signal at 146 mph or 7% higher than 137 mph.
 
"I changed covers" hey that is with holding some pretty vital info.. :rofl:

and I guess the "signal multiplier" theory was kind of close.. :laugh: (and I do have one on the bench :)
 
I wouldn't have thought that replacing the billet cover with another billet cover would have made any difference. Same exact style/look only a different company. It's supposed to relocate the OE sensor to the same location. :banghead: what's a millimeter here or there ??? :poke: :laugh: I guess I'll have to shave things down a bit to get the sensor pickup a bit closer.


oh yea, I was curious to know by how much the loaded circumference changes from acceleration to deceleration at these high speeds. Seems as though my tire's loaded circumference went from about 71.2" to 79.2". This sounds like alot but when you look at the change in the radius it's a 1.3" increase.
 
commuta what gps sensor do you use? at work we use racelogic, vbox, etc, some are nearly $10K for the sensor.
 
hey Greg, The GPS is from AIMsports it has a 5hz sampling rate so it's affordable compared to other systems. What's the rate on the vbox? From what I learned is that once you're at the 20hz or higher not only does the price go through the roof but you start to run into latency from the NMEA stream which isn't the fastest.
 
we use them for measuring stopping distance, the vbox is 100Hz update rate, the race logic uses a 20Hz gps with a 200Hz accelerometer and interpolates between the points, both work very well for what we use them and very accurate.

we always used to use a 5th wheel mounted off the back of the car, these are so much more convenient.

What does the aimsport run? We use some 5Hz boxes as well but they run about $500, just use them on traffic type tests where you are steady state driving for the most part.
 
Ken since we are both data nerds you might appreciate this, 06 Ford GT speed trace from a month ago when I got to do some testing on it, car rips pretty hard I must say

oops messed up the screen shot, 1 mile accumulated distance, 172 mph

gt1mile2.jpg


IMG_3341.jpg
 
Last edited:
the aimsports unit is in that 500 range. Don't know the exact price. It's accurate enough for straight line "stuff". I'm a little confused on the 5hz gps with 100hz vbox. Wouldn't this leave you with so to say steps in your data? I'm starting to like the whole GPS thing. I wouldn't give up my wheel sensor s though since they're sampling at a much higher rate. :thumbsup:

I have to say it again, I'm completely jealous of the job you've got :bowdown::cheerleader::thumbsup::beerchug:

That GT40 pulls like gangbusters! I don't remeber if they were supercharged or not? Even in 5th gear that monser is pulling hard.
 
yep its supercharged, pulls pretty hard, especially for how its gear (top of second is like 95 mph). We use a cheap 5Hz gps for normal driving, 100Hz is for accelerating or decelerating fast. And the 200Hz one we like even better LOL
 
i think it would be way overkill for what you are doing, we are pulling over 1g deceleration for stopping distance stuff and need it to update that fast LOL

Whats your email wanna send you something.
 
You must log in or register to reply here.

Similar threads

WWJD
got cheap Radio Shack GPS...
Replies
19
Views
942
WWJD
WWJD
Rhythm
Cellphone GPS software
Replies
1
Views
568
Rhythm
Rhythm
el stumpo
New way to fight speeding tickets
Replies
7
Views
822
BlueHaya
BlueHaya
6
LM-2 Logs for Turbo Busa
Replies
9
Views
2,308
65Cobra
6
2
Since 1999 A Busa codes to c31 6Fixed
7 8 9
Replies
177
Views
9,653
2busa
2
Back
Top