Motec CDL3 speed input issue

Hi guys,

I have an issue with wheel speed logs.

The Motec CDL3 has front left, front right wheel speed (cherry type sensor) and gearbox output speed using the GT101 sensor (car has a locked diff) all are actual Speed inputs on the CDL3.

When I drive in a straight line without accelerating much the FL, FR, drive speed and GPS all read really close, within 1kph.

When racing on track, it appears to have considerable wheel spin when accelerating and steady rear wheel slowing compared to the front under brakes. This cannot be the case as the car is planted well under both full acceleration and braking.

There seems to be a speed input timing issue where the drive speed input is out of time compared with the front? But it isn’t a time delay or it would be always ahead or behind, but this is both depending on acceleration or braking if that makes sense.

The tyres are not rotating on the wheels, they are definitely not spinning or locking up always the same amount..

See log screen shot, the curser is at a point where there are no longitudinal G forces so the speeds are very close.

Any ideas?

Cheers,

Brent

Can you put your configuration in a zip file and upload it?

There are several settings that can make a difference, such as the number of triggers (pulses per revolution) as well as the sample teeth. Increasing sample teeth will probably smooth out your drive speed noise.

What sample rate are you logging each of these channels?

Your front speeds are always delayed relative to the rear speeds -- I suspect a higher sample teeth, or fewer number of teeth. I think you are looking at the amplitude (value) not the "when" the peaks / valleys occur.

Hi David,

Thanks for taking a look, see attached.

As you will see, front wheel speeds are 24T, gearbox output is 10T and the diff ratio is 3.23:1, the SW only allows whole numbers for tooth count, so I've adjusted the tyre circumference a bit to get it to match the fronts. Let me know if you want me to upload some logged data also.

Cheers,

Brent

I did have the drive speed on a CDL-3 Dig input as I was using the 3rd speed input for fuel temp via Ethanol content sensor, I changed it to a speed input but same result.

Your time plot pic is using "Ground Speed L/R" channels which are only being logged at 1Hz, whereas your drive speed is logged at 50Hz, this will generate the transient delay.

Your "Wheel Speed FL/FR" channels are logged at 50Hz so I expect if you plot those they will show a trend more similar to the rear.

Sorry, old log image… see attached.

Ah, thanks Adam, I’ve had a mix of drive speeds, wheels speeds and ground speeds depending on the worksheet in i2. I looked at a few different worksheets and some showed almost matched front and rear speeds, some didn’t… that’s an oversight and a half!

OK, I believe the delay in the original screen shot was because you were plotting the ground speed channels calculated and transmitted from the M800.

I think the noise in your Drive Speed channel is likely do to un-even spacing of the trigger teeth, and you would get a much better result with sample teeth of either 3 or 5.

It is not unexpected to see the drive speed 1-3 % faster when accelerating.

The best way to calibrate these speed channels is to simply coast (gearbox in neutral or clutch in) for a few seconds on the longest straight. You can then compare GPS Speed to the other speed channels and scale the tire circumference for each to get the same values. You can just scale the channels in I2 first, then transfer that to your CDL3 config for the future. Now any measured difference is what's really happening.