AX22 versus Driftbox accuracy revisited
#1
Addict
Rennlist Member
Rennlist Member
Thread Starter
AX22 versus Driftbox accuracy revisited
I was sceptical about the Driftbox's ability to give accurate longitudinal Gs using GPS only relative to the AX22's use of accelerometers AND GPS.
The roads are cold and damp in the UK at the moment so the limit of G with my rear wheel drive 1480kg car wearing 315/30s on 12" rims was 0.745G
If you check out this graph you can see the green line is the lateral G, the red longitudinal G and the black speed, as the long G builds, at 54mph and 0.678G (long) the wheels spin and the car lurches to the side (green line showing lateral G lurch) about 0.3s later the peak long G is hit and I have backed off which brings the back end back and the green line shoots the other way recording a 0.25 lat G
Anyway having established that there was not going to be any high G heroics I did a couple of runs using 4th gear - neither of which managed to hit my world record run (fluke ) but I was doing them with both the AX22 and the Driftbox activated and measuring.
I have copied the runs below - this is the data from the AX22 (top) and Driftbox (bottom) from exactly the same run. BTW the Driftbox Gs had the "smoothing" setting at 5 that's why the line is nice and "smooth"
I am pleased to say that the times were the same to an accuracy of 99.7% and the long Gs were pretty much the same -you probably can't read the graphs so here is the Long G data taken from the software:
AX22 Driftbox mph
0.557 0.538 80
0.538 0.524 90
0.469 0.469 100
0.409 0.401 110
0.387 0.375 120
So there you have it, the Driftbox produces the goods - I am sending the files to Jean to make sure I've not goofed (entirely possible )
The roads are cold and damp in the UK at the moment so the limit of G with my rear wheel drive 1480kg car wearing 315/30s on 12" rims was 0.745G
If you check out this graph you can see the green line is the lateral G, the red longitudinal G and the black speed, as the long G builds, at 54mph and 0.678G (long) the wheels spin and the car lurches to the side (green line showing lateral G lurch) about 0.3s later the peak long G is hit and I have backed off which brings the back end back and the green line shoots the other way recording a 0.25 lat G
Anyway having established that there was not going to be any high G heroics I did a couple of runs using 4th gear - neither of which managed to hit my world record run (fluke ) but I was doing them with both the AX22 and the Driftbox activated and measuring.
I have copied the runs below - this is the data from the AX22 (top) and Driftbox (bottom) from exactly the same run. BTW the Driftbox Gs had the "smoothing" setting at 5 that's why the line is nice and "smooth"
I am pleased to say that the times were the same to an accuracy of 99.7% and the long Gs were pretty much the same -you probably can't read the graphs so here is the Long G data taken from the software:
AX22 Driftbox mph
0.557 0.538 80
0.538 0.524 90
0.469 0.469 100
0.409 0.401 110
0.387 0.375 120
So there you have it, the Driftbox produces the goods - I am sending the files to Jean to make sure I've not goofed (entirely possible )
#3
Thanks for the comparison. We've been seeing velocity drift errors in the AX-22 that have been causing very large errors 60 to 130 errors in the US. We're not sure why this happens, and it could possibly be due to the GPS satellite signals received in the US if no one outside the US is seeing this problem.
I hope to receive a Driftbox as a holiday gift and will post more 60 to 130 results.
I hope to receive a Driftbox as a holiday gift and will post more 60 to 130 results.
#4
For short duration measurements, the GPS input is less relevant I think.
Even the cheap G meters will produce accurate short-term measurements. AX22
et all take over when you start caring about where you are, such as mapping
courses, examining the effects of different lines through the same locations.
Even the cheap G meters will produce accurate short-term measurements. AX22
et all take over when you start caring about where you are, such as mapping
courses, examining the effects of different lines through the same locations.
#5
Addict
Rennlist Member
Rennlist Member
Thread Starter
Originally Posted by Bill S.
Thanks for the comparison. We've been seeing velocity drift errors in the AX-22 that have been causing very large errors 60 to 130 errors in the US. We're not sure why this happens, and it could possibly be due to the GPS satellite signals received in the US if no one outside the US is seeing this problem.
I hope to receive a Driftbox as a holiday gift and will post more 60 to 130 results.
I hope to receive a Driftbox as a holiday gift and will post more 60 to 130 results.
Because all the results I have got using the method above have been so close it would be VERY obvious to spot an erroneous run.
Surely the GPS satellites should work in the US - they are your satellites after all
Red
No external antennae on the DB, only for the AX
Joe
Accelerometer only units can give errors easily -check out the chat we had about them in the 60-130thread -near the begining.
#6
Addict
Lifetime Rennlist
Member
Lifetime Rennlist
Member
TB
This is great input, thanks for taking the hassle and time to perform this test. No energy right now to go through the data in detail, I have 40 laps of today's DE on Dubai's Autodrome (new to me) to analyze, including a 360 ..
The fact is that I have seen several DB files with errors in them after performing the test I use on the AX22 on the DB data, and since the long Gs and speed both seem to be accurate on both units...
.....My conclusions are the following..
Many of the DB users (and AX22) are fooling the tool (in fact I have seen only a couple of good runs on DB, one was from KPG) which does not surprise me the least, in view of the ridiculous times posted by some guys, and the second is that both the AX22 and DB are very accurate tools when used as they were designed to, and misleading when not. As far as I am concerned, I can confirm whether a AX22 run was accurate or not, but I cannot do the same with the DBox it seems, although it should be no mistery.
Just to put an end (from my side) to this debate.. The Ax22 is extremely accurate, just as the DB it seems. I respectfully do not agree with Bill's theory about satellite issues. I have more than 40 runs that are spot-on from different users in the US, a handful were only wrong. I have had myself a couple of bad runs with my AX22 (out of well over 100), and in both cases it was my fault, such as calibration while driving or unsecured tool, but those errors were almost meaningless, and it happens of course, but certainly not consistently!!
Toby can you please extrapolate your data for every 0.01 s (speed, long G and distance) on excel and send it over to me.. KPG mught be able to help. I am interested to perform my test on both runs the DB and AX22.
Many Thanks for the great info.
This is great input, thanks for taking the hassle and time to perform this test. No energy right now to go through the data in detail, I have 40 laps of today's DE on Dubai's Autodrome (new to me) to analyze, including a 360 ..
The fact is that I have seen several DB files with errors in them after performing the test I use on the AX22 on the DB data, and since the long Gs and speed both seem to be accurate on both units...
.....My conclusions are the following..
Many of the DB users (and AX22) are fooling the tool (in fact I have seen only a couple of good runs on DB, one was from KPG) which does not surprise me the least, in view of the ridiculous times posted by some guys, and the second is that both the AX22 and DB are very accurate tools when used as they were designed to, and misleading when not. As far as I am concerned, I can confirm whether a AX22 run was accurate or not, but I cannot do the same with the DBox it seems, although it should be no mistery.
Just to put an end (from my side) to this debate.. The Ax22 is extremely accurate, just as the DB it seems. I respectfully do not agree with Bill's theory about satellite issues. I have more than 40 runs that are spot-on from different users in the US, a handful were only wrong. I have had myself a couple of bad runs with my AX22 (out of well over 100), and in both cases it was my fault, such as calibration while driving or unsecured tool, but those errors were almost meaningless, and it happens of course, but certainly not consistently!!
Toby can you please extrapolate your data for every 0.01 s (speed, long G and distance) on excel and send it over to me.. KPG mught be able to help. I am interested to perform my test on both runs the DB and AX22.
Many Thanks for the great info.
#7
Addict
Rennlist Member
Rennlist Member
Thread Starter
Originally Posted by Jean
Toby can you please extrapolate your data for every 0.01 s (speed, long G and distance) on excel and send it over to me.. KPG mught be able to help. I am interested to perform my test on both runs the DB and AX22.
KPG, if I send you the file can you do what Jean has requested ?
Trending Topics
#8
Addict
Rennlist Member
Rennlist Member
Join Date: Dec 2002
Location: San Luis Obispo, CA
Posts: 2,124
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by TB993tt
Surely the GPS satellites should work in the US - they are your satellites after all
#9
Originally Posted by TB993tt
Love to.......if I knew how
KPG, if I send you the file can you do what Jean has requested ?
KPG, if I send you the file can you do what Jean has requested ?
#10
Originally Posted by Jean
.
Many of the DB users (and AX22) are fooling the tool (in fact I have seen only a couple of good runs on DB, one was from KPG) which does not surprise me the least, in view of the ridiculous times posted by some guys
Many of the DB users (and AX22) are fooling the tool (in fact I have seen only a couple of good runs on DB, one was from KPG) which does not surprise me the least, in view of the ridiculous times posted by some guys
BTW..concerning the 360 at the AutoDrome, if you had a DriftBox...it would have given you a score on that spin
TB thanks for taking the time to do this...
#12
Burning Brakes
Join Date: Jul 2001
Location: Orange Park Acres, CA
Posts: 982
Likes: 0
Received 0 Likes
on
0 Posts
Just to make sure I got the right present for Xmas, I sent Santa my credit card with instructions to use it at racelogic and got the performancebox which is same as the drift box without one mode for drifitng. It shard to drift blackie with R tires and all wheel drive, althought TB and Jean should do a video with all that hp.
I chose the perfomance box based upon the technology, engineering and ease of operation. I was impressed by the built in engineering for their calibartion wihich is the only technical thing I understand. I will be running some tests in a secret location and will publish all good or bad in the near future.
I chose the perfomance box based upon the technology, engineering and ease of operation. I was impressed by the built in engineering for their calibartion wihich is the only technical thing I understand. I will be running some tests in a secret location and will publish all good or bad in the near future.
#13
Guru
Lifetime Rennlist
Member
Rennlist Small
Business Sponsor
Lifetime Rennlist
Member
Rennlist Small
Business Sponsor
Originally Posted by JJayB
I will be running some tests in a secret location and will publish all good or bad in the near future.
#15
Burning Brakes
Join Date: Jul 2001
Location: Orange Park Acres, CA
Posts: 982
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by chris walrod
Let me know where your racetrack is, and I'll come watch -- I live just over the hill from ya