Tell me about your DL-90 Data Logger
 

I know a few of you must have one of these. How do you like it? It looks like a very comprehensive package. Did it help your driving? How do you like the software? Thanks in advance for the help.

Cheers, James

Look into AIM's systems also. For the $$ they are very sweet.

AIM!!??..Please don't tell me you're instant messaging while on the track :-) Seriously, What is AIM, Iss there a website?

Cheers, James

Hi James. Here's some more references off of the Production Racing Forum. I know that one of the guys here on Pelican is also working on integrating video since he posted a trial version here earlier.

James...

It does everything they say it does!

BUT.. (and this is a BIG BUT) ;) you must take the time to analyze the data to get value out of it, AND have the time to try different things on the track to see what works.

May favorite use is to find those 1 or 2 corners on the track where you are loosing the most time, and try several different lines through it in a session. Then go back and look to see a) which got you through in the least amount of time and b) what your time was in the next session to make sure your exit was the most optimal as well.

For instance in a "throw-away corner" you don't look at the corner so much as the time you were in the following straight. Then go back and find out what line was best..

James,

Chris gives (as ever) good advice, I'd add to it the following:

1. Initially the DL-90 is like trying to take a drink out of a fire hose. There is a ton of data and the interface takes a few hours to get comfortable with, so I find that I understand it better when I break the track down into discrete segments.

2. Once you have a) a consistent line in practice; b) good data on the car, e.g. you know the caster, camber, toe and other alignment settings and c) tire data = pressures and pyrometer readings, you can then begin to exploit the DL-90.

3. What Chris said is exactly right-- you only want to change ONE THING at a time. Car is understeering in slow corners. Reduce front tire pressure by one pound. Go out and run three laps with the DL-90 set up, then immediately come back, drop the tire pressure by another pound, go do three more laps. Once you get done with the session, you can then analyze things like the maximum speed through that corner, the segment time, the max lat g, the max long g, etc. It gives you quantitative feedback on the effect of changes to setup. But the only way to effectively correlate speed vs. changes is by only changing one variable at a time.

Which takes a long, long time. Sometimes I wish our sessions weren't 20 minutes long, but rather 7 minutes, so you could do a whole bunch of separate sessions with a tuning/analyzing period in between.

I love mine, every time I use it I say, "There is no performance, without measurement!"

No experience with it....but a local club member started a company this past fall to merge DL-90 data with your video camera to look at your driving style.

Check out http://www.apexvisualizations.com/

Men,
Thanks for the responses. I feel that right now I really need to analyse different sections of track to see where I need to improve. Right now all I have is a laptimer which is great but has it's limitations. I know the DL-90 has a ton of info on it so I guess it's pretty user friendly? Right now every track I go to I turn laps off within 1 sec of each other, they just need to be faster. My gut feeling is that on any given lap, I do a few corners well and a few not so well. It would be really helpful for me to be able to see what's working and what's not. Thanks again for the help. Also, how's the reliability etc Can I convert serial to USB signal? I have a laptop with no serial but USB?

Cheers, James

USB/Serial convertor is about $30, check www.pricewatch.com

That's be me! Thanks for the plug. Check out the web site for more details. I'll have a new beta release available in a couple days.

The DL90 is a really powerful tool... couple it with the apexviz software and its even more usefull...

usb/serial converters can be had for much less than $30

I've tried to connect my DL90s to a laptop via a serial-to-usb adapter, without any luck. Any of you make this work? I'm just a software geek ... :rolleyes:

How many of the external interfaces have you guys hooked up? I would be interested to know how you wired the data acquisition for the DL-90.

RPM
Wheel Speed Input
Digital input (brake on/off)
Analog Input (oil pressure?)
Analog Input (oil temperature?)

I'm gonna get some lightweight Tefzel wire and hook it up, would love to know how you guys did it. Chris?

John, have you tried the USB converter? It seems like there are conflicting reports in regard to this working.

Cheers, James

I have:

1. RPM (standard)
2. GPS (standard)
3. Ignition Advance from twin electromotive units (into analog)
4. Brake (into digital)

Where are you guys mounting the unit in the car?

I have had success duct-taping it to the top of the dash with a piece of folded-over duct tape under the aft edge to level it out. Also success taping to the passenger floor. Since I don't have a radio I'm going to fab an aluminum plate the same size as the radio grill and mount it on there, with some washers underneath to level it out.

Chris, how do you wire RPM pickup?

James, I haven't used the USB/Serial as I have serial on my old track laptop. Presumably it would have to include a driver to spoof the software into thinking that the USB bitstream was coming from the serial port.

Mine is mounted to the cross-brace on the roll cage.

John, I just used the inductive pickup that they send along with it. I figured you could tap off the RPM wire, but it doesn't work for some reason.

Ah! Smart. That's what I'll shoot for.

I cut a aluminum plate about 8" long and tapped it to thread onto a tripod or camera mount (1/4" x 20) then plastered it, the DL90, GPS receiver and lipstick camera with velcro. The plate can be mounted to the rollbar with an I/O port or a tilt-and-pan tripod head (done both with good results). The camcoder sits on the floor.

http://forums.pelicanparts.com/uploa...1075934843.jpg

Greg;
If I were you I'd go for a positive mechanical attachment to plate, and then make sure that the plate has a safety strap holding it tight. Can your set-up withstand a solid 30-40G hit? A simple test would be to apply a force equal to 30-40x the pieces' weight against them and see if they push off. If they don't hold, you're liable to have all of that equipiment turn into projectiles. The worst case would be a hit on to left front of the car in which case these projectiles would be aimed at your head.

http://www.pelicanparts.com/support/smileys/spankA.gif

Good points to stress, all.

The plate also has a nut/washer keeping it attached to the mount, though you can't see it in the picture. On the track the tripod head is bolted to the harness truss (the picture is a test rig in another car) or to a rollbar via a I/O port mount.

Also, I'm not using "normal" velco, rather a plastic 3M material that works on the same principle. I've no doubt it'd stay put in an accident, given how hard it is to remove the DL90 from the plate afterwards!

The camcorder isn't bouncing around on the floor. It's strapped to a seat and/or sitting in a seat/door pocket.

The DL-90 only weighs a few ounces too. I've got L brackets in the side of mine, those are held to the roll bar with hefty hose clamps. Keeps the DL-90 secure and more in tune to car movement.

http://forums.pelicanparts.com/uploa...1075947160.gif

Anyone care to comment on this? What should it look like?

John,

Looks familiar but you may be giving up some on the left turns. Mine look the same in that respect.

One contructive criticism: The peak in the top middle means that you can also get on the gas a bit earlier... Mine generally have a flatter top, meaning that while I am in heavy corner loads, I am also accellerating hard as well. The peak in the top center shows less accelleration under corner loads.

Just checked back...
here is the web site:
http://www.aim-sportline.com/

I use the MyChron3 Gold XG & it is sweet....

Here's one of mine from a fast, but not so challenging track:

You can see that while I pull just over 1.2 G's on the let turns, I'm pulling over 1.3-1.4 on the rights... Pretty common since most tracks run clockwise, but notice the flat top.
http://forums.pelicanparts.com/uploa...1075955257.jpg

Whoa! Very sweet...don't mean to hijack the thread...but can you tell us more? A couple years back (5 years) I had a MyChron2 in my 944 Turbo and it was great for lap timing, oil and water temp...but that's about it.

Just saw the MXL which is akin to the Motec Dash, albiet cheaper.

Chris;
While trying to visuallize what the data means in the real world, I guess I can make a couple of observation which might be helpful -- and then again -- maybe not.

http://forums.pelicanparts.com/uploa...1075984579.jpg

* It seems like your data points are fairly tightly grouped which would I suspect suggest that you are being pretty consistent. That is a key skill and not always accomplished.

* If you look at the area under the arms of the "T", you can see that there are data points there which would suggest that the car can operate in that area. But there are fairly few data points in that area. Since those regions represent the transition from maximum braking to maximum turning (and vice versa), it would seem to be situations where you are not maximizing the use of the car's handling. Instead in most cases you have reduced your braking to .4 G's, but you are not using any of the turning traction that may be available. It might be an area to pick up some "free" time on the track.

Interestingly, John seems to have more points out there (relatively speaking) then you do.

Now I can also picture the data being skewed by the track geography, but still the fact that there are some points out there suggests that you went there occasionally, just not as much as you could. The image that comes to mind is the thread that you posted a couple of weeks ago about guys racing by coasting down a curvey road.

Caveat -- Having said that, I've never had my driving tracked by instrumentation before, so I can't say that I know what's involved in going into that area of the traction circle.

Good commentary here, thanks!

Chris, I see what you are saying, the "peak" at the center means that I am waiting until lateral gs come back close to zero before getting on the throttle-- pickup the throttle earlier and there should be a straight line all the way from the edge of the plot to the peak of acceleration, the inverse of what's on the other side.

Hey, that looks like Road America! (Throttle, zzz, brake, turn, repeat). I agree with John, very very consistent!

John, thanks for pointing that out, this is the classic "friction circle" where you want to drive around the edges of the circle at the limit at all times-- max accel, then max braking, combination of brake/corner, max cornering, then combination of cornering/acceleration to the limit. I guess the ideal one would look like this:

http://forums.pelicanparts.com/uploa...1075990250.gif

John, Have you used the USB adapter with success?

Cheers, James

James, no. I use serial. Your question was whether it could be converted, I know it can, but I don't know whether the software would recognize it.

You might email the developers and ask if they have a USB version in beta. Seems to me that RS-232 is going the way of the dodo. . .

Interesting discussion about the G Circle data.
Have to admit I've never really looked at this data critically, but you guys make some good points. Pulled a recent run from VIR, which is a longer track with some fairly technical sections. Not nearly as technical as Mid Ohio or Summit Point. This is what I got.

http://forums.pelicanparts.com/uploa...1076005875.jpg

I guess it looks a little more like a triangle than a cross, although the cross distribution is still dominant. What do your experienced eyes see here? Kinda like a racer's Rohrschach [sp?] test.

Jleutjen,

THose areas that you have highlighted would be filled if I was doing a lot of trailbraking, which I don't. You would have to be experiencing sideways force while under braking. If you look at the graph in a FWD car, it almost looks upside down to the 911.

However I to John Cramers comment, you can also look at the total G's on a graph and find out if you are just "coasting" in certain places. This doesn't mean that it's particularly bad, in fact it occurs at every turn in when you transition off the brakes and onto the throttle, but haven't quite turned in yet. THe key is to minimze this.

John, it IS Road America.

I agree Tony. Really interesting thoughts.

At DE events they speak to "managing the tire patch", but now I think they should do more talking. It seems to me the G-circle chart is showing exactly how well the driver does managing the patch. If you have empty space to the left and right of the verticle you're squandering traction. Either trail brake deeper or get back on the throttle sooner.

Cool stuff.

Chris;
Actually I didn't want to use the Trailbraking word. The concept should be more of trailing off of the brakes as you feed in steering. Picture a string tied to the rim of the steering wheel and the brake pedal. As you start to turn the wheel, you let up on the brake proportionately.

This is where I would say the advice to the novice would be significantly different then the advice to someone who has spent quite a lot of time pounding around a track in anger.

Rule 1 to the novice: Break in a straight line.

Rule 2 to the experienced racer: Ignore rule 1.

The fact that there are points out there (Note I can see one at -.8 lateral and -.4 longitudinal) proves that the car has grip out there. The trick is being able to use it. If you can use it, I guarantee you that you'll make up a good 20 to 30 feet of track per lap in that corner alone on someone who isn't. It may not be comfortable at first, but it is the secret to being fast. It may also require some careful tuning of the chassis to make the car comfortable in that realm. I would start with the shocks if it were me.

You might also want to go back to the data and identify where on the track you were getting the points under the arms of the T. What is it about those corners (or laps) that let you take the car into that zone?

As Greg mentioned, you need to keep the car on the envelope in order to be fast. A T shaped graph would be perfect for a DE'er. But if you want to make it to the front of an SCCA National, you'll need to be on the envelope.

I've found karting to be a great place to practice this technique since (at least in the rental karts) the brakes only work on the rear axle. So they don't like to be agressively trail braked any more then a 911 does. But in order to go really fast, you need to be able to keep them on the edge of the envelope -- just like the 911 or any or race car.

JOhn...

All good points... I'll take a look at where I'm getting those values... Could be something special about certain corners where I am doing it... OTOH some corners you just DON'T trailbrake (off cambers for instance) because it's such a fine line between that and looping it, it's not worth the risk. :)

Agreed. It's a judgment call.

I still trying to figure this out, but trying to correlate these 'under the cross' data points with the track map suggests that there are at least three situations that seem to stand out:
- rolling onto the gas early in the apex to track-out transition
- trail braking
- left foot braking

The second two are limited to specific corners.
- The right hander into Oak Tree and the climbing left at the end of the back straight are good trail braking candidates.
- A few no-shift turns where the left foot brake helps keep the car balanced.

Still want to look at this some more, but my eyes start to cross after 30 minutes or so of this stuff.

Tony,

Why don't you look at the G-circle only for those sections of the track?