sbelles

Now I know I'm getting old when 385 hp is the low hp option.

dgm8138

Thank you again! I'll see if either of these guys plan to be at grattan anytime.

Gofishracing

Night track driving- easy to out drive your street headlights. I knew I was in for a difficult evening when my student said at dusk that he has problems seeing at night. Haven't done another. There is an upcoming one at Pocono.
http://www.motorsportreg.com/events/...0#.Vz8YiEZZ2ZY

kespiritu10

Yeah, night driving was definitely different. I have to really hand it to the instructors riding in students cars. I've heard some stories that weren't so pleasant.

Anyhow, Coach VR, did this guy not correct fast enoughwhen he was sliding (scrub to 4:52)? Btw, I've been reading the thread since page 1, it has helped ALOT.

Veloce Raptor

I dunno
Looked to me he had too much steering angle and was determined to hit that rumble strip (possibly driving point to point?) whereas the 993 displaced it properly and had a much more relaxed posture...

TXE36

For the first time, I got data analysis to "work". Not work in the sense of collecting data, but work in the sense of analyzing the squiggly lines, formulating a plan, and actually achieving some measurable improvement. Recall that back here and here I discovered a rather ugly coasting problem going into T1 going CCW at TWS.

The original plan for dealing with my taking of T1 was to use speed as reference. This failed spectacularly because the car just doesn't have enough power to really pull this off. Rather than holding the speed constant and working up, what worked was holding the throttle constant, starting at say 75% or so, and then gradually working up. Finally, during the last session Sunday, the magic happened and I got close enough to pulling it off early in the session that I was able to work up the courage to do it. I was realized I was 90% of the way there and just needed to cowboy up to finish it.

Attached are long G vs distance and velocity vs distance plots. The dark blue oval is T1. The blue plot is from last February and the pre T1 coast is clearly visible as well as the drop off in speed before the brakes are applied. The red plots are the last four laps of the session. No coast!!! (well, mostly)

Down below is the video of the session overlayed with data. I posted the whole session because you can see the courage building up and the first really good pass through T1 is just after 15:17. The next three are good as well just after 17:25 (note the whoohoo!), 19:33, and 21:40. At the time I did not care about the rest of the track - it's only purpose was to get me back to T1. Lap times are well of my PBs, however, this T1 performance is better than I've ever done.


Still needs work - still some small hesitations and I could have gone deeper. Concentrating on more speed out of T15 will results in speeds going up a bit as well.

When I finally did it, holy ****, what a rush!! Made this older guy feel like a boy again . Reminded me of why I've dumped buckets of money into this hobby. The intensity of the session is quite obvious at 22:44 during the cool down lap.

Thanks for the help guys.

-Mike

Veloce Raptor

Knew you could do it...

kgorman

Nice work TXE36!

Awesome write up and display of how data matters.

Throttle time matters!

Thundermoose

Good job. Do you have true throttle position or is it inferred by Longitudinal G's? It looks to be actual TP. If so, compare your TP plots to VR's. Easy way to find quick time especially since it is the same car. I am able to do this with other E46 M3's with AIM SOLO DL's. It was amazing to see how quickly my friends are able to get to throttle.

ProCoach

In my opinion and experience, I can infer TPS with Traqmate's long g sensor.

Thundermoose

I don't disagree, but if you have true TPS, it makes the comparison a little more straightforward.

Veloce Raptor

I didn't have a true TPS in the car so I used long g. Not sure whether Mike has added one

TXE36

I upgraded the data collection in the car to include an analog tap of the throttle body TPS and a digital input is connected to the brake light. Thus, the throttle seen on my videos is true gas pedal position and the brake indicator comes one whenever the brake light is activated. Indeed, you can see some of the chicken wavering to hold the gas down. It took a lot of effort to hold my foot down. When I finally did, it was nearly a religious experience .

The discussion has got me thinking there may be a way to express the long-G data in the video overlay. Also keep in mind my TM is an early unit that does not have accelerometers, thus it's long G is computed from GPS.

One other important point I left off my post is the reason I knew that holding speed into T1 was not working was because I actually looked at my long-G data on Saturday and realized nothing had changed even though my butt dyno said I was faster. If I hadn't looked until Sunday night or Monday morning, Sunday afternoon would not have happened and the results disappointing.

I am at a point where I'm fast enough that mere lap times and video isn't optimal information to make progress. With this long-G business, I've got plenty to work with for a while.

-Mike

ProCoach

This is NOT TRUE. The early TM have a two-axis electromechanical accelerometer on the board, the later TM's have three-axis.

Traqmate has NEVER calculated from GPS long or lat g...

AiM MXL's have NO long g accelerometer and VVBOX has only calculated from GPS speed and heading delta. Everyone else uses real live accelerometer data.

Veloce Raptor

+1. Spot on.