tach stripe problem

The stripes should be evenly spaced, I'd make the black and white the same width.  You might try the "poor man's" approach and see if just blacking out two of every three white stripes with a Sharpie works.  That should work a lot better than half the wheel being white.

Thanks, I'll try it.

Re-reading my post, I said it incorrectly. I made white strips of tape and covered up some of the black stripes around the wheel. It didn't end up even but I tried to spread out the covers as best as I could. It actually runs pretty good.

Im not contesting that "even" is the right choice. I'm just trying to "think train" and learn more.

  Does the board use both signals or signal size in its computations? If not then I'm not sure it would matter TOO MUCH as long as the one signal that was used is uniform and the unused signal could be noticed by the detector. I've divided tapes, rather unevenly, by adding thin stripes with a marker to change the speed of a non-train motor because the correct rpm motor was NLA and  I needed to shave 150 or 200 rpm off. That left me with fat white lines and both skinny and fat for black ones. No jerky movement; that CPU averaged the count well.

The same colors being even should make for better balancing and longer motor life (though a heavier flywheel would likely do more good). I.e. fat whites and skinny blacks or fat blacks and skinny whites should give balance if uniform and only one signal is used. But the action is push, coast, push, coast; and it seems to me the shorter the coast signal the more time the push has to operate and coasting happens during the push cycle on PWM too, essentially lengthening the short coast signal. Balance, cooling, or capacitor charging are the only other reasons I can think of to worry about the coast signal size unless it is carefully monitored and used by the board too.

I'd be very surprised if they don't just use either the positive or negative edge from the tach encoder.  You seem to be trying to design a system of way more complexity than is necessary for the task at hand.  I really doubt MTH did something that convoluted.

That's what I figured too actually. I was just trying to determine why all would need to be equal vs a mix of small but equal & large but equal, other than convenience or simplicity.