Suspension data acquisition for Traqmate?

[winefool]

I did a search and found some good information on different types/sources of potentiometers that could be used for measuring suspension travel, but it appears that you would need a masters in electrical engineering to fab up something...

The Traqmate web site lists a series of sensor that you can plug in but nothing for suspension.

Has anyone successfully installed suspension travel sensors and connected them to a Traqmate system? Are there any packaged solutions out there that I am missing.

TIA for any input or suggestions!

[TheOtherEric]

Traqmate probably doesn't suggest suspensions sensors since (a) it's very advanced stuff that most Traqmate users wouldn't be doing, and (b) the sampling rate of the Traqmate probably isn't the best for that. Many recommend like 1,000 hz sampling, but I personally suspect you can get meaningful data even at 100hz (but no, i haven't yet confirmed).

Can you point me to any discussions about sensors/pots? This is a project I want to get started this week. I use a DL1 that samples at 100 hz.

[winefool]

This was the best thread I found. It's great if you're an EE and have the know how to start with an industrial sensor and figure EVERYTHING else out on your own. I'm not there...

http://forums.bimmerforums.com/forum...d.php?t=701839

[POPS M3]

This may not help at all but RackPak sells position sensors for their systems and maybe -- just maybe it could be used for your system. I have not used these.

See Link: http://www.racepak.com/Sensors/Travel.php

Ken

[TheOtherEric]

This was the best thread I found. It's great if you're an EE and have the know how to start with an industrial sensor and figure EVERYTHING else out on your own. I'm not there...

http://forums.bimmerforums.com/forum...d.php?t=701839

I didn't see much in that thread that should scare you away. There was talk about using a low-pass filter, but if you're using a Traqmate then you're probably only sampling at 100 hz so you don't need to worry about that. Also some talk about FFT but here you just want your software to differentiate the displacement signal to get velocity. If Traqmate can't do that, then you can almost certainly do in excel.

As for sensors, I was inclined to try a cheap ride height sensor off of ebay for like $50. No they're not made for this but they'd probably hold up for a short time at least. Mounting would be more complicated than those nice string pots though.

[colinE30]

Just wanted to say, my Formula SAE team runs an ISAAC data act system that samples right around 1000 hz which is about optimal. You want it to be that high so that you can cut it down a bit or smooth it and still get really precise data.

I guess it all depends on how serious you are, if you are doing it for shits then 100 is probably fine, but for really good suspension tuning then you need more. Also if your suspension is not that adjustable then I suppose you don't need as precise of numbers.

If you have any questions about our set up I'd be glad to discuss it more, I didnt want to fill up your whole thread if you aren't interested in the ISAAC system.

Cheers

[endoforce5]

You may also want to consider placing a camera in the wheel well to visualize what the suspension is doing. There a lot of cheap cameras out there to accomplish this and you may already have a system or rent one. With multiple "lipstick type" cameras you can view your track position and suspension movement all at one time. I have found this to be surprisingly helpful at times, as you can see multiple things happening at once (camber change, suspension travel, steering input, movement of components that are and aren't supposed to move, etc.). Linear pots are great for spring/shock tuning but the information provided is pretty focused.

[neelvasavada]

I've been a data engineer in everything from IRL to offshore powerboats, including Grand AM and ALMS. I agree; the traqmate is just not appropriate for chassis data; once you get to that level you need to start thinking about something else.

These days, I deal with MOTEC, AIM, GET and Pi data systems; most of these are quite a bit more expensive than the traqmate, but there is a very good option in the AIM EVO4. Its less than $1300, has GPS built in, real logging memory and logging rates. You can run damper potentiometers no problem.

The Replay XD camera is AWESOME for doing video and data; I've hung them off the bottom of Superbike fenders to see suspension. Check them out here:

http://www.apexspeedtech.com/store/replay-xd

-Neel

[winefool]

Thanks guys - sounds like the Traqmate solution is not going to work. I'll just have to tune the suspension by the seat of my pants or try the camera approach...

[TheOtherEric]

I've been a data engineer in everything from IRL to offshore powerboats, including Grand AM and ALMS. I agree; the traqmate is just not appropriate for chassis data; once you get to that level you need to start thinking about something else.

-Neel

I tried working out the math to see if 100hz sample rate will get enough samples for these measurements, and it seemed to me that you could probably get decent enough data in the 0 to 3 in/sec range. Perhaps I did the math wrong, but it seemed ok to me. (IIRC, I assumed like a sinusoidal motion with amplitude perhaps 1 inch.)

And granted, you'd like to see beyond 3 in/sec, but the low-speed range will get you most of what you need to know. So 100hz isn't ideal, but wouldn't it get you like 80% of the data you want?

[neelvasavada]

I tried working out the math to see if 100hz sample rate will get enough samples for these measurements, and it seemed to me that you could probably get decent enough data in the 0 to 3 in/sec range. Perhaps I did the math wrong, but it seemed ok to me. (IIRC, I assumed like a sinusoidal motion with amplitude perhaps 1 inch.)

And granted, you'd like to see beyond 3 in/sec, but the low-speed range will get you most of what you need to know. So 100hz isn't ideal, but wouldn't it get you like 80% of the data you want?

Yeah, those assumptions won't work... the problem is also aliasing right? Its not like you're getting an average of the suspension movement - those outliers (which happen all the time in - this is a mass-spring-damper system with highly variable inputs) make you think you're looking at something you're not. I've seen this many times - even simple positional calculations - how much travel were you using? What is pitch? Roll? All of those can be fairly invalid with 100Hz suspension data. Of course, that's not to stop people who don't have better data from saying its fine.

Do a little test for yourself - get record a corner into 2 inputs - 100Hz and 500Hz. Then look at the data and ask yourself what you know about the car's performance. Sometimes I fear that the low-end data systems are doing nothing but giving people false information about their cars.

Here's another test - log at 100hz put put a zip-tie firmly on the shaft. Drive on a fairly smooth track and you're data will show you only using some of the travel when the zip-tie will often be crammed into the bump rubber.

Oddly, its easier to tune with video (30 hz) than 100 hz data, because at least then you SEE what is happening; your mind connects the dots and you don't filter the data in your head.

-Neel

[miata33]

The Traqmate was never really intended to measure suspension movement, however as it has been said in this thread, video is actually a great way see what is happening.

With that component in mind the Traqmate's ability to sync with camera systems is very helpful. In your case, the chase cam would be the best with it's bullet cam setup.

The Traqmate would control the camera system so all your data and video are pre synced as you go around the track. Main benefits is the ease of use and simplicity that the Traqmate offers as well as the great software.

If you need more specific 'data' about what each shock is doing, follow some of the other post for a system that measures at a higher rate. More data is helpful in some cases... but keeping it simple has it's benefits as well.

Good luck either way.

Andre'

[TheOtherEric]

Yeah, those assumptions won't work... the problem is also aliasing right? ...

Oddly, its easier to tune with video (30 hz) than 100 hz data, because at least then you SEE what is happening; your mind connects the dots and you don't filter the data in your head.

-Neel

Ok, I see your point. I guess the problem is that high frequency motions won't be seen by such low sampling rate, and will just show up as low-frequency (or random) motion. Ok, so 100hz is out.

But how do you use video to tune shocks? A few people here have mentioned that, but for the life of me I can't understand how you'd use that. You just see the wheel going up & down... what does that tell you?

Now that I've found $11 micro cams that shoot 30fps, I'm interested!!

[neelvasavada]

But how do you use video to tune shocks? A few people here have mentioned that, but for the life of me I can't understand how you'd use that. You just see the wheel going up & down... what does that tell you?

Video can tell you a lot by watching the tire contact patch. Its not a substitute for data but its very helpful. We used to tune Baja 1000 Trophy Trucks by watching the suspension from a chase helicopter!

Tell us more about these $11 micro cameras!

[TheOtherEric]

Video can tell you a lot by watching the tire contact patch. Its not a substitute for data but its very helpful. We used to tune Baja 1000 Trophy Trucks by watching the suspension from a chase helicopter!

Tell us more about these $11 micro cameras!

Oh, I get it. With video you're not looking at suspension motion but rather tire contact patch. Interesting. I guess the idea is to just keep the tire on the ground?

I haven't used the $11 ebay cameras. I just read about them in the Putnam Park thread:
http://forums.bimmerforums.com/forum...&postcount=130

[fizzx]

Oh, I get it. With video you're not looking at suspension motion but rather tire contact patch. Interesting. I guess the idea is to just keep the tire on the ground?

I haven't used the $11 ebay cameras. I just read about them in the Putnam Park thread:
http://forums.bimmerforums.com/forum...&postcount=130

Hey that's my post :-)
Here's a shot of the micro-cam velcro taped to the underside right behind front right tire. Pretty fun to watch the tire deflection and liftting off the ground.
[ame="http://www.youtube.com/watch?v=lyWbtN1bHFs"]http://www.youtube.com/watch?v=lyWbtN1bHFs[/ame]

and here I just taped it to the door pointing at front left wheel
http://www.youtube.com/user/gccjason#p/u/6/kNZkIYC3xnI

I need to spend a little time figuring out placement/securing for track speeds.