Vintage Jag Tachometer Oscillation

Thread Starter

graybeard

Joined Apr 10, 2012
99
I am trying to repair a Gen 3 Smiths tachometer from an early 70s V12 E-type Jaguar. It has developed an oscillation when the IC has warmed up that is causing it to read incorrectly.

The tachometer meter is a 75 ohm D'Arsonval ammeter. The IC that drives it is a custom marked version of TI's SN76810 GENERAL PURPOSE MONOSTABLE/TACHO DRIVER. It generates a fixed pulse-width one-shot current pulse with each ignition event. As the frequency of the ignition pulses goes up, the duty cycle of the current through the meter goes up driving the needle to indicate the correct RPM.

This tachometer needle has started reading low and sometimes with some jitter once it has warmed up. It functions correctly when the IC is cold, but the one-shot pulse oscillates at about 500KHz when the IC is warm.

Here are the schematics of the tachometer:

1656608633795.png

I have a couple of scope shots that illustrate the problem. The top trace is the input signal from my signal generator and the bottom trace is from IC Pin 5 which drives the meter. The first shot shows the the correct signal when the IC is cold:

1656608930566.jpeg

This shot shows the oscillations in the one shot pulse:

1656608998353.jpeg

As the IC warms up, the ~500KHz oscillations start at the left and gradually grow to the right until it covers the pulse. Warming the circuit up with a hair dryer accelerates the oscillation. Hitting any of the components other than the IC with freeze spray has no effect. Hitting the IC with freeze spray instantly gets rid of the oscillation until the IC warms up again. The full oscillation causes as much as a 1000 RPM drop in the tachometer indication.

That led me to replace the TI IC with an SAK215, which is a pin compatible equivalent IC with a slightly different time constant for the one-shot timing. Strangely, that did not fix the oscillation problem, it had no effect at all. Just because I suspect all old caps, I then replaced the 0.047uF cap with a 0.05uF ceramic cap. That also had no effect on the problem.

All of the resistors are carbon resistors. I am able to measure R8 and R9, which are part of the Zener diode regulator circuit. They combine to make 120 ohms, about 50% higher than the marked 84 ohms. I clipped a 330 ohm resistor in parallel with those 2 resistors to make it 88 ohms. I was thinking that the regulator might not be providing enough current for the IC to work properly, but that had no effect on the problem. I suspect that all of the resistors have increased resistance due to age. Unfortunately, I didn't have the foresight to check the resistances that I could with the IC out of the circuit.

I don't see any solder joints that look cold. The traces on the PCB are very fragile, some of them were lifted a bit during the original soldering and it was very hard to replace the IC without lifting those traces. For that reason, I am hesitant to try to replace all of the resistors and caps to try to solve this problem. I would much prefer to understand what is causing the oscillation and simply fix that, if possible.

A friend of mine (Alex Miller) has written a document (attached) that dives deep into the SNSN76810 and the SAK215 ICs, including both data sheets. The context of the document is an analysis of Smiths' Gen 4 tachometers. The Smiths Gen 4 tachometers use the same TI ICs and roughly the same circuit as the Gen 3 tachometers. But it does add a cap between Pins 1 and 3. We speculated that that cap may be there to eliminate the possibility of oscillation, so I tried clipping different caps between those pins, including the values used in both versions of the Gen 4 tachometers, but that did not stop the oscillation either.

I am not very good at diagnosing oscillation problems, so I could really use some help. I know oscillation comes from feedback and I can see the feedback loops in the IC, but have no idea as to how to try to fix the problem. Any suggestions as to how to try to solve this oscillation problem will be greatly appreciated.
 

Attachments

Thread Starter

graybeard

Joined Apr 10, 2012
99
Thanks for posting that circuit diagram Dodgydave. I tried adding a filter cap across the meter and it knocked the oscillation down to a negligible level. 0.1uF seems to give me the smallest oscillation amplitude and 0.2uF doesn't improve that. I can still see a little on the scope, but it doesn't seem to affect the calibration.

I think I can make this work for the customer, but, I am not happy that I still can't understand what is causing the oscillation in the first place. The filter cap seems like a bandaid to me rather than a cure.

Any other input regarding the root cause of the oscillation?
 

Thread Starter

graybeard

Joined Apr 10, 2012
99
AlbertHall, Thanks for the suggestion. The power rails are clean as a whistle on the scope. I just replaced C3 and that had no effect. The oscillation is still there. Very perplexing.
 
Top