Racepak.com

Hey guys - I installed the USM I got from you a couple months ago and set it up with steering and throttle potentiometers..

Did my first weekend with the new sensors a week ago and while the throttle position worked just as expected, the steering trace looked way off. I calibrated the steering signal to go from -360 to +360 (rotation of the steering wheel). If I remember right, the resulting trace was between 10 and +25, and had basically no relation to where I imagine I was pointing the steering wheel.

I'm in the garage right now with the car and trying to figure it out:

- measuring the signal wire voltage directly at the potentiometer gives me the readings I expected, i.e. the readings I had calibrated the USM in datalink with. When I set it up, everything matched and I used the calibrate window to set the range, etc. The readings at the potentiometer match the ones in datalink - 1.4v at full left, 4.78v at full right.

- measuring the voltage at the 'signal' terminal also gives me the same result, as expected. So I'm not suspecting the wiring..

- the 'calibrate' window in datalink tells me that the G2X is getting random values around 3v - 3.020 to 3.1 or so. I'm rotating the wheel from lock to lock, my multimeter tells me the values are going from 1.4 to 4.7v. The G2X is getting something else.. The number does change when I move the wheel, but it randomly goes up or down and always within 0.1v..

Any ideas?

Lets try something first, in terms of calibration...just to verify operation. When calibrating steering, there are multiple ways to do so. Trying to calibrate to a 360 degree layout can become cumbersome, if calling anything to the left - and to the right +. This ends up being a 720 degree calibration.

First, recalibrate with the steering straight ahead as 0 (Set A point). Then turn the wheel 90 degrees to the left and call that 90 (Set B point). This will then graph anything turning left as positive and anything right of zero as negative. This will let us know the values coming back are ok..as typically you would not be putting much more than 7-10 degrees (at the wheel) into turning the car through a corner.

Calibrating by steering wheel angle is more familiar to the drive and helps in relating in terms of viewing the data...What is considered the "better" method is to calibrate at the front wheel..actually calibrating the sensor to the true angle of the front wheel..not the steering wheel. This can be accomplished by using setup "turn plates"..If you dont have those, you can do a quickie method of projecting a straight light from the sidewall on the floor (that straight ahead 0 point) ..then project a 15 degree line from that line..turn the wheel to match the line and set that as 15 in Set B point. It is not perfect but would get you close in terms of wheel angle at 15 degrees/

Thanks
Tim

On my car, I installed a steering sensor but it only reads one revolution. I believe I did what Tim said and calibrated straight ahead as zero, and then I did 180 at one half turn. (more like 170 because the sensor can only do one revolution, but as Tim said, you hardly ever turn that much direction)

I just made my first Dashware video with the steering angle sensor displayed as a steering wheel. If you would like the link, let me know. I also did a writeup on my steering sensor on another forum and can provide the link as well.

Hey Tim,

I had calibrated it at steering wheel angle initially for simplicity sake, but it was my intention to get actual wheel angles (which I did this weekend).

Anyway, I got it working again last night by starting from scratch - changing the sensor type to a stock 0-5 volts, saving the VNET config to the G2X, shutting it off and back on, then calibrating the sensor again.

Just to be clear - the problem I was having was, in the calibrate window, the sensor was reading ~3 volts no matter what I did to the steering wheel. This was despite measuring the voltage both at the sensor and at the USM terminals and it being clearly NOT 3 volts.

I had calibrated it 3 weeks ago before my last race and it worked fine in the garage, but none of my sessions recorded the correct data, I guess somehow the config file got corrupted between me setting it up and my first run at the track.. Any idea what could have caused this?

Again, configuring the USM channel from scratch completely seems to have fixed it.. As soon as I burned the stock 0-5v setup to the logger, the voltage reading matched the multimeter and everything was fine.

For reference - I did end up calibrating it to the actual wheel angles. No turn plates, but I stringed the car up, measured the wheel angle with the wheel pointed straight ahead (to account for toe in).. Then turned the steering wheel 180 degrees, measured the distances to the string and used basic trigonometry to calculate the angles. Repeated for the other side and everything matched up to about 0.2 degrees, so I'm confident it'll be close enough! Turns out on my little formula vee, 180 degrees of the steering wheel is about 15.6 degrees at the front wheels.

Yeah..I was going to go into the whole trig thing...but thought maybe I better just make is a simple explanation for the forums. What probably happened it is somehow become "tangled" up in the cal values between the data logger and the config file. What you did was the correct solution..just set it back to 0-5 degree sensor, send that to the data logger..power the logger off and on and then recalibrate. That typically solves the issues.

Now that you have calibrated to the actual front wheel angle, you can start looking more in depth, in terms of data/tire analysis.

Thanks

Tim

Just a follow up to let you know the steering position channel worked as expected this weekend Found some interesting stuff with it too, at least until the engine blew up!