THSCC Forums

So in the last blue classroom at VIR yesterday, they were talking about data acquisition. Made me start thinking...

Do these devices have yaw sensors?

If not, and you hook up two sensors on the car, one at the front & one at the rear, could you get it to give you yaw angle via the software? I'm thinking in terms of rallycross or rally, and speed comparisons for turning early vs. turning later, understeering through the corner, getting too much oversteer, etc.

Seems like you could overlay the two plots and tell way the car is pointing just by which line is where, but is it accurate enough to do it that way?

Hi Kevin,

To my knowledge, yaw is not a measurement offered in Traqmate.

http://traqmate.com/v-web/bulletin/bb/v ... d25152a336

I know they are measuring such variables in NASCAR these days. However you need a significantly more expensive and sophisticated system such as MoTeC.

_________________
“I feel safer on a racetrack than I do on Houston's freeways.” - A.J. Foyt

Kevin Butler
Mobetta Autosport Spec E30 #612
2003 C5 Z06 Corvette
AFR Miata, SM2 61, '93 w/200 SC'd RWHP - soon to be resurrected
Waaaay too many other projects....

You can put a yaw sensor on any data system that supports external sensors. I know you can do it on a DL-1, and I don't believe the sensors are terribly expensive. Any vehicle that has traction control and/or dynamic differentials likely already has a yaw sensor you can probably tap into, too.

You could use accelerometers at both ends of the car and do math on the result, but I don't think there's any reason to do that given that any system that would support that would also support a yaw sensor natively and the yaw sensor would work better.

And yes, MoTeC, AiM, Stack, and Pi systems will support yaw. The DL-1 is probably the least expensive data box that does as well.


--Donnie

A few years ago I was curious about this same question and dug around and in addition to ability to measure linear acceleration (x, y, z), you can find accelerameters that will measure changes in rotational acceleration.

I understand what you are looking for Kevin as far as yaw, but I was also curious to see if you would be able to measure pitch and roll as well. Overall I was curious if you could generate some type of suspension movement values. The "correct" way for what I wanted to do is to have sensors at each corner that measures the actual suspension movement instead of trying to infer it via pitch and roll. But I was thinking it could be cool if you could do this via a system that you didn't need to hook up to the car in an extensive way.

_________________
Richard Casto
1972 Porsche 914
2013 Honda Fit Sport
2015 Honda Fit EX
http://motorsport.zyyz.com
Money can't buy happiness, but somehow it's more comfortable to cry in a Porsche than a Kia.

The real answer there is that if you're trying to measure pitch and roll, you still need the sensors on the suspension corners. Just knowing you're rolling and pitching is of little use if you can't tell if it's due to compression on one side or rebound on the other, for example.

Putting sensors on each corner isn't that hard. The annoying part is those little buggers are sorta spendy (just over $100 each) and if you screw up they can get lunched easily. Plus you do have to run wire to each corner, which in a production based car can be a real pita. But it really is the only way to get that data.


--Donnie

Yeah, I am sure you are right on that. I was hoping that somehow via the magic computer analysis that it might be a cheap way to do it. But like you say at some point you can end up with something happening that could be caused by multiple things. Which at the end of the day is not going to tell you what you need to know.

It reminds me of another story about the problems with data acquisition. In that even with sensors at all four corners, you may not be able to see the entire picture. I can’t remember where I heard this from, but some racing team years ago was crunching the numbers from the four corners and it just wasn’t adding up somehow. That what they were measuring wasn’t fitting into their model as to what they should be seeing. They eventually put a camera into a wheel well to see what was going on. It was that the wheel was lifting a bit off the ground occasionally and apparently they were not able to see or infer that just by looking at the suspension position data. The point being that no matter how much you measure or how well you think you understand things, there is probably something you are missing.

It would really be cool to know just how much info something like a Formula 1 team has in this area. For example I have always thought it was interesting to watch slow motion footage of those cars crashing over curbs. Seeing the amount of sidewall distortion and oscillation of the wheel as it rings like a bell makes me wonder how complex their models of just the tires and their interaction with the suspension must be.

_________________
Richard Casto
1972 Porsche 914
2013 Honda Fit Sport
2015 Honda Fit EX
http://motorsport.zyyz.com
Money can't buy happiness, but somehow it's more comfortable to cry in a Porsche than a Kia.

That's why even the famed "shakers" aren't perfect since they still limit what the tire will really be doing by some amount. They really help, but there's nothing like data at the track, and yeah, even that can cause trouble. And yes, a wheel lifting like that would certainly cause the data to "not add up", or in this case the thing that they thought should have been zero was really just the amount of lift they had!


--Donnie

I think I now have the answer Richard! A 4 wiimote data acq. system!

(for reference, I think F1 teams use at LEAST 8 Wiimotes)

_________________
2012 MX-5 Sport SUV

Okay, the F1 bit was funny, but in all honesty you can do some useful stuff with a Wiimote on the suspension if you wanted to. Linear position sensors are going to be higher resolution, but based on what I know about the Wiimote, you could definitely get useful data from four of them. You're either going to go insane with spreadsheets or you'd still need to invest in some "real" data analysis software (there's stuff out there just for doing work like this, though).

The Wiimote could present a mounting challenge due to physical size in some places, too.


--Donnie

Bluetooth remote sensors is something else I have been thinking about, but had not thought about using a Wiimote

So far this is the coolest thing I have seen that someone has done with a Wiimote...

http://www.youtube.com/watch?v=Jd3-eiid-Uw

Bluetooth sensors would cut down on the wiring hassle. If someone would just come up with an open standard for sensor input using bluetooth you could have multiple vendors building compatable sensors and storage devices. The storage device would no longer needing analog inputs as everything would be internal (such as basic x,y accel and GPS) or use external via bluetooth. It might even be cheaper for the initial device as it can do without the A/D conversion and associated circuitry. Then you just buy additional sensors ala cart. However sensor cost would be a bit higher due to the overhead of having to support bluetooth and A/D conversion on the sensor

_________________
Richard Casto
1972 Porsche 914
2013 Honda Fit Sport
2015 Honda Fit EX
http://motorsport.zyyz.com
Money can't buy happiness, but somehow it's more comfortable to cry in a Porsche than a Kia.

I actually am trying out a wiimote data acq. system (just one though) in Greenville, it wasn't totally a joke.

Its definitely not anything I'll be able to look at in real time due to the spreadsheet calculations necessary (the setup just outputs .csv files), but since it cost me all of $12 (usb bluetooth dongle, I already own a Wii and the laptop) I figured I'd try it out.

_________________
2012 MX-5 Sport SUV

I just had to take this out of context...

_________________
Martyn Wheeler
AXing Kit's '05 Mazda 3, #29 HStock
(when The Gonzo Symphonic allows)