winders

With the 893 mV resolution I have for 360 degrees of rotation, I have .40 degrees of rotation per mV. I would think that should be plenty of resolution for this application.

ProCoach

Thanks, Matt.

Sorry, Scott. I learned something new, thanks to your issue.

I have always used and installed sensors on the shaft immediately behind the dash before angular change OR used a linear pot on the rack, on the axis of the rack.

winders

In an air-cooled 911, there is virtually no real option to to access the shaft the steering wheel is attached to. In fact, there is no easy way to access any shaft other than the one I am using. There is no simple way to attach a sensor to the air-cooled 911 steering rack either. The space is tight and access is limited.

After examining my options, it would appear the only practical thing to do is live with what I have now.

Matt Romanowski

I think with the custom sensor as you did, you will be fine. You'll get good results. You'll be able to do simple xy plots for under/over steer, and you'll be able to do more calculations to see the same plus figure out your optimum level of the same.

Gary R.

I'm still trying to Google the clogged belt....

ProCoach

Hah!

winders

No worries, Peter. I knew cardan u-joints were not constant velocity but never encountered a situation in real life where that fact was evident. I should have put 2 and 2 together right away but did not. This shows the difference between knowledge and experience!

kgorman

Nice Scott. thx for sharing.

ProCoach

Scott, I went back and checked on this and while it appears that measuring on the other side of one carden joint results in non linearity, if you measure on the other side of the second carden joint, assuming there is one, it is linear with the wheel. Interesting!

winders

Yep. See post #15!

T&T Racing

That is why there are two universal joints in every misaligned steering column. The input to the first is linear, the steering wheel, the output from the first is non linear, the input to the second is non linear, the output from the second is linear. One cancels the other if connected in series, a basic problem in college dynamics class.

winders

Yes, I know. But there is more. They need to be installed with the yoke ears aligned and the shaft angles equal for it work out as you describe.

As I said, I knew all of this at an academic level. I just hadn't encountered a situation where this characteristic of the cardan joint was an issue. I suspect not many people have. Again, the difference between knowledge and experience.