Boxerworks.com

Heat related = electronics related = still could be a failure of filtering components, i.e. capacitors.

Barry's Cap comments can still apply. Lots of electronic components can develop failures when hot and then return to correct functioning when cooled off. Capacitors do this all the time on motherboards and in amplifiers.

Like I said before, the absolute simplest test is to see if the tachs behave the same way on a different bike. If you have such a vehicle available, that will immediately establish if you really have bad signal/interference, or bad receivers.

By the way, I found some speculative information on how it actually works:


Re: How does an auto tach work?

Postby LikwidCirkel » Mon Feb 27, 2012 6:57 pm UTC
As far as I know, each revolution will trigger a pulse of equal duty cycle using some kind of simple timer circuit. The output of this feeds to a very low pass filter - effectively a charging capacitor. This outputs the average voltage as a DC level, which will vary linearly with RPM, and can be used to deflect a needle.
http://forums.xkcd.com/viewtopic.php?f=36&t=80943

Later in that string someone posts the schematic for a frequency meter and LikwidCirkel says that the schematic basically reflects what he said in this post.

I can certainly see how such a circuit could begin to make the needle bounce around if the buffering capacitor stopped doing its buffering job correctly.

EDIT: Other more detailed info here: http://www.ehow.com/info_12132364_car-t ... -work.html and here: http://www.opelgt.com/forums/1e-other-e ... -work.html

A sketchy ground on my R65 first revealed itself in a bouncy tach at speed. It later revealed itself in a failing horn, flashing headlight, and in the end, a complete electrical shutdown. The bike ran well until the very near the end.

OK, back to grounds then. I'm trying to remember where they all are. Main one at the speedo bolt, seems like a few more on the top tube, what am I missing (besides my youth, energy, and marbles)?

Simplest way to check if it's a ground problem is to run a temporary ground wire from the tach right to the battery. If the problem persists, it's not a ground problem. If it cures it, THEN you can fight your way through figuring out exactly where the ground problem is.

Of course... thank god you guys have replaced my brains.

Update- always want this info in here in case someone searches later.

For tach issues on a 1992 R100GS- CHECK THE FUSE, DUMMY!!!! The tach and brake lights share a fuse- if your brake light isn't working due to blown fuse (maybe because you just pulled a lot of the wiring off during a rebuild???), the tach will be affected. In my case, weirdly, the fuse must have had a partial connection.

Now, I'm not sure this is the ultimate solution , because haven't test ridden the bike for more than 10 minutes, but it sure looks like it.......

Don

Still sounds like a ground problem to me. When systems that couldn't possibly affect an electrical component do, it's almost always a ground issue.

Which of your bikes are you doing this tach test on? I see in your profile you have a pot-full.

I agree with Barry too. I would change the capacitor in the circuit, they too are susceptible to heat issues. Usually come in 2 heat ranges 70c and 100c. The cap shapes the pulsating signal and should absorb irregularities, it's an inexpensive part, just change it and see if it makes a difference.