universal wheel speed based motorcycle traction control

[ogre]

Traction Control System for Motorcycles


somebody contact this guy and get the specs for the controller. then smithabusa can build it.

 

[Boosted Cycle Perf]

I already have traction control on mine. It's called my right wrist.

 

[ogre]

talking about for rain, not launching. for racing it's not traction control we need, just more traction.

 

[AJAY]

Might be nice. Bazzaz makes one for under $950. Cycle World tested it recently and, IIRC, they were about as impressed with it as they were with the unit on the Kawasucki.

 

[ogre]

the one described in the article would need to add a tilt sensor and new carriers for the brake rotors that have a tone ring, as well as a sensor for the tone rings. it'd be pretty simple really. just have to find the sucker who wrote the article to see if any software was actually written. it sounds startlingly similar to the system employed on the S1000RR

 

[ogre]

and the bazzazz unit is a piggyback, quick shifter, and traction control. my factory ecu does two of those things, so when they sell a tc module for 1/3 of that price, i'll think about it. plus it's only going to watch how fast you accelerate and slow you down. i don't know how well that would work in the wet, or in a turn.

 

[fallenarch]

Actually you don't need to sense wheel speed to do traction control. The early Kawasaki system looked at changes in engine rpm. Imagine an engine in gear and under full throttle. The engine will build rpm at a certain rate. If that rate changes a certain amount in a certain amount of time, then it's safe to assume the tire has broken traction and is spinning freely, so the TC retards the spark to reduce power. I think the bazzaz unit does something similar.

It's pretty simple to make something like this. You can pull the rpm off the plugs, then do some testing to figure out what typical rates of acceleration are. Then you could interrupt the spark. Add a lean sensor and you would have a pretty sophisticated TC setup.

All that being said, the bazzaz is already made and tested. Frankly at $950 it's a bargin.

 

[ogre]

for 1000 bucks? nah. that's not a deal. i wonder if petrik could add something to the ecu

 

[fallenarch]

for 1000 bucks? nah. that's not a deal. i wonder if petrik could add something to the ecu

The lean sensor and microcontroller cost about $300. Then you have to write the code, establish baseline values for wheel slip and finally determine where/how to retard the spark. When you get all that done you still need to deal with packaging and somehow test the system to make sure it's working and doesn't have any glitches. I figure you're talking $500 in parts alone to have something reliable. Remember this isn't a camera or something - having it work properly is a matter of life and death.

All that considered, the bazzaz is tested, packaged in a solid/reliable package and you get other goodies like mapping, quick shifter, etc. Frankly, if you count your time as worth anything, it would be hard to do this for $1000. Also, even if I got it to work I wouldn't sell it to anyone because of the potential liability. At best I would post everything with the source code and let people do it themselves with a big disclaimer saying I don't know if it works.

I have attached the lean sensor and microcontroller I have been looking at. They come from the world of hobby robotics. Technically doing this isn't that hard and the specific parts are available these days for a decent price. The first board I bought with gyros and accelorometers cost me $300 alone and it was useless on a bike because the drift on the gyros was so bad. I'm not sure how you would tell the ECU to retard the spark though. Maybe some of the ECU editor guys can answer that.

 

[AJAY]

I'm with Arch above. As an engineer, if you're building it yourself, you're a test pilot. For the $1000 you'd spend with Bazzaz, you've got their testing on your side.

I was surprised the Cycle World evaluation was so positive, with comments that it performed effectively at the ragged edge of traction. I would have thought you'd need at least the front and rear wheel speed sensors, with the lean sensor and accellerometers being nice to close the loop (more robust).

 

[ogre]

i always thought that wheel speed based tc would be simplest.

 

[GNBRETT]

for racing it's not traction control we need, just more traction.

sometimes but not at the drag strip. traction control for drag racing can be done by pulling timing instantly therefore cutting power to stop further wheel spin and then putting it back in as soon as the driveshaft sensor isnt reading tire spin any longer. one does the same with their wrist when they lose traction. one just cant do it nearly as fast as a computer can.