944S2 - Tachometer Issues

[weir501]

Hi everyone, I recently attempted to fix the tachometer in my 944S2 which was non responsive since I bought the car.

I have some hobby electronics experience and found that one of the capacitors on the circuit board was dead - after replacing this with another of the same rating, the tach does respond and seems to bring the needle to the correct revs, however the needle seems very loose and will wobble around before sitting still. Unfortunately since I have never seen this tach working properly, I have no reference for how it's supposed to be.

Could this be due to the clock springs losing tension over the years or something else maybe?

Any help on the topic would be appreciated

[blueline]

Hi everyone, I recently attempted to fix the tachometer in my 944S2 which was non responsive since I bought the car.

I have some hobby electronics experience and found that one of the capacitors on the circuit board was dead - after replacing this with another of the same rating, the tach does respond and seems to bring the needle to the correct revs, however the needle seems very loose and will wobble around before sitting still. Unfortunately since I have never seen this tach working properly, I have no reference for how it's supposed to be.

Could this be due to the clock springs losing tension over the years or something else maybe?

Any help on the topic would be appreciated

Giving your request a bump to get it back in front of a few more eyeballs. I'm sure several here will have ideas that might help to resolve your question. Will also tag @Tom as this sort of thing tends to be in his wheelhouse.

By the way, welcome to Carpokes!

[Tom]

Hi everyone, I recently attempted to fix the tachometer in my 944S2 which was non responsive since I bought the car.

I have some hobby electronics experience and found that one of the capacitors on the circuit board was dead - after replacing this with another of the same rating, the tach does respond and seems to bring the needle to the correct revs, however the needle seems very loose and will wobble around before sitting still. Unfortunately since I have never seen this tach working properly, I have no reference for how it's supposed to be.

Could this be due to the clock springs losing tension over the years or something else maybe?

Any help on the topic would be appreciated

Giving your request a bump to get it back in front of a few more eyeballs. I'm sure several here will have ideas that might help to resolve your question. Will also tag @Tom as this sort of thing tends to be in his wheelhouse.

By the way, welcome to Carpokes!

The tach should rise and fall smoothly. When you say the needle seems loose and wobble, I'm not sure what you mean? Is it physically loose -- i.e., does the pivot point in the middle bounce around? Or do you mean the needle is twitchy and increases/decreases erratically until it settles in? Have a video you can share?

[weir501]

It's acting more like the needle just has no tension to it, it moves a lot more freely than the other gauge needles
This video should explain a little more


I'm testing the gauge by putting a square wave generator to the signal pin on the back, thanks to a rennlist forum for the diagram on that


Also, excuse the speedo being missing, I'm replacing that pesky odometer gear while the cluster is out

[Tom]

I suspect you are looking at a RL thread I was on -- I was Tom M'Guinn over there. My hope is that your tach is fine and the signal you are feeding it is responsible for the wild mood swings you're seeing. Three things might contribute -- the voltage, the on/off duty cycle, and the rate of change from zero to 4k rpms. The scope shot I posted on RL back then is below if helpful. Each little pulse twitches the tach upward a bit, and I suspect but have never confirmed that the duty cycle can create a stronger twitch. So that 'might' be part of it. But the other thing I have seen it going from nothing to a 4k pulse instantaneously like that -- the tach will definitely overshoot from all those twitches coming all at once. Try doing a little program with a 90% duty cycle (90% high, 10% low) AND ramps the rpms (pulse rate) up at a more realistic pace -- say start at 800rpm for a few seconds and then and ramp it up to 4k over a second or two -- and see if the tach still bounces around like that. It shouldn't, but if it does then I'm guessing the cap wasn't the only bad component.

Image 15.jpeg (206.99 KiB) Viewed 1642 times

[weir501]

So I have changed the pulse width to 90% and tried again - this did not seem to change the stength of the pulse twitch
However by incrementing by +10Hz up the gauge from around 800RPM, it does appear a lot smoother, there is just a little bit of wobble still if the needle moves quickly

I have noticed that when the gauge is not powered, moving the needle by hand it will snap back to zero fine but it does so quite quickly, whereas doing the same with the speedometer gauge, the needle returns to zero slowly as if the magnet in the back is slowing it down some - not sure if this is just a red herring though

[Tom]

No guarantees, but I'd be inclined to try it on the car. Reducing the instantaneous changes to smaller increments clearly helped. I'm guessing you are using a function generator and not a microprocessor, and that you are still stepping the frequency up by a fixed amount rather than increasing ratably over time? My guess/hope would be that all of the over-shoot needle swinging would go away if the rpm's increased smoothly up and down without any steps -- like you'd see on an actual running motor. Like I said, no guarantees, but the change between those videos would be encouraging enough to give it a try in my mind.