HELP! tach died and I need to know what can replace it!

[78targa]

Hi guys. I need a quick answer on this. I getting ready to start a race Sunday (the Targa Newfoundland) and my tach just died.

I run a HPV-1 electromotive ignition and the car had an APEXi tach in it. The output tach signal runs between 8.5 and 9.5 volts. Here's my question: what year tachs would work for this ignition?

I have heard about square wave etc, supposedly that is what the SC tachs work off of.

I have an SC tach with me that I have not yet put in (the car is not yet here) Do I need to hit the panic button or do I need to start overnighting parts? I have very little time to get this right.

Any SC guys out there that could check the voltage input on the purple wire to the tach?

I have gotten lots of advice, but nothing technical from some good sources. Unfortunately they all have conflicted. Hard data or BTDT would be of immense help. I need to know what I'm looking for!

Thanks!

[Rick Crider]

Casey ..

I'm just up the road from you, in Monroe.

Have '76 930 Turbo and '89 Carrera here. Mechanically and electronically savvy to a respectable degree.

Glad to check / test / etc ... on either of mine for you if it would be of help.

Could even pull the tach from the 89 for a few days if it'd get you out of a pinch.

Don't think there's any way to revive your existing tach?

Lemme know if I can help ....

Rick

[78targa]

Thanks Rick, but I am physically in Newfoundland right now so it would be a bit of a drive.

Tach died day before we loaded up and I ran out of time to get a replacement.

Looks like there are limited selections here which was a bit of a surprise. I believe an Autometer would work, but I can't find one so far. I would prefer not to blow out my SC one if I can avoid it.

Not sure if the existing tach is fixable. It appears to have just gone dead. All the power and grounds function....honestly at this point I can't afford to deal with a flakey tach.

Rick, do you know what voltage your tach used for its signal?

[Driver Ed]

I've got an '82 SC and an oscilloscope -- I check mine out this morning and post the results in an hour or so.

[Driver Ed]

Sorry that it took a little longer than expected. I forgot that I have to remove the steering wheel to get the tach out, but here's the 'scope image.

It's a little hard to see, but the 'scope is set to 10V/div vertical and 2ms/div horizontal, DC coupled. The long horizontal line is at about +12V. The signal then dips down to ground for about 2ms, shoots up to a peak of around +50V and then settles back down to +12. So basically, it's a 12V square wave with a huge overshoot, which is pretty typical in automotive circuits.

Also, if anybody's ever interested, my rough measurements were these:

RPM	period (ms)
1700	12
2500	6.4
3300	4.4

I've been thinking about building a data logger which would extract the engine RPM from noise on the +12V line, so I also took some pictures of the waveform on the +12V at the tach. The 'scope was set to AC coupled, 500mV/div vertical, 5ms/div horizontal and the engine RPM was about 1050.


Sorry about the crappy 'scope pictures, but the 'scope is the same vintage as the car! Good luck on your run, Casey, and I hope to see you at the DE in Kershaw in a few weeks.

Ed

[Early_S_Man]

Casey,

The SC tach should work fine with the Motronic signal!

Ed,

Try DC coupling setting ... 7A26 or 7A18, right?

Nothing wrong with that Tek 7000-series scope! You won't find anything 'current' on the market as durable & reliable, or as easy to repair & calibrate!

[Driver Ed]

Yeah, Warren, the scope is a Tek 7603 with a freq counter plug-in. I like it, but the pictures I posted don't compare to high-resolution digital scope pictures. On the other hand, my scope doesn't ever need rebooting!

I used DC-coupling on the first signal; it wouldn't have made any difference on the second since it's just a high speed transient anyway. Also, can you or somebody check my math? Here's my calculation to go from the period of the signal to engine RPM:

1. I measure 19.7 ms for the second waveform and noted that the tach said about 1050 RPM


2. since it's a four-stroke, 6-cyl engine, there are 6 blips for every 2 revs


3. so, 2 rev/6 blip * 1blip/19.7ms * 1000ms/s * 60s/min = 1015RPM

That's about 3% low, but that's probably because it was hard to read the tach with it hanging out over where the steering wheel used to be. Is the math right, though?

[Early_S_Man]

Ed,

Nothing wrong with your math ... and the 3% error is within the uncertainty range for the tach and the scope measurements, not all of which would be additive!

My tach calibrator/CDI tester puts out 60 Hz and 120 Hz signals, measured with Tek 222A at 16.6 ms and 8.3 ms ... computes to 1194 rpm and 2412 rpm with my trusty old HP-41. Close to the 1190 rpm and 2410 rpm results obtained on my old Pickett N-16ES, too!

[Driver Ed]

Thanks. Good to have verification. Sounds like we have similar tastes in equipment: HP-15c and Pickett N3-ES here!

[ChrisBennet]

I posted some scope shots of the tach signal to the 3.2 Motronic tach back in this thread
-Chris

[Driver Ed]

Thanks, Chris. Looks like 20 * f = r where f is in Hz and r = RPM. Interesting to see how much less overshoot there is on your signal.

[Early_S_Man]

Ed,

Those screen shots of a digitized signal aren't really looking at the signal, but rather a 'connect-the-dots' recreation, or representation of the signal ... and due to a sampling error, a narrow-width overshoot can be easily missed by most newer digital 'wonder' scopes. Plenty of people mistake a perfect, square risetime waveform as an actual measured risetime, rather than just a picture connecting two samples that had quite a bit of amplitude difference between samples.

This is an area where most 30-40 year-old, CRT analog scopes from Tek or HP have an advantage over most newer digital LCD storage scopes ... though not necessarily the high-end Tek or HP models in the $20K+ range! A good 10x probe and 100 MHz analog scope can accurately pickup and display noise transients missed by 'modern' digitals ... though they can be hard to photograph! I have used a LOT of Polaroid 3000 speed film in Tek C-50 series cameras trying to capture certain transients ...

[Lorenfb]

"I've got an '82 SC and an oscilloscope -- I check mine out this morning and post the results in an hour or so." - Driver Ed -

1. The SC tach signal does NOT exceed 12 volts as shown, so the
signal image posted is NOT the real TD signal from a SC CDI unit.

2. Chris' waveform is a PERFECTLY accurate picture of the 3.2 Carrera
& a basic SC tach signal. There's NO loss in image because of sampling,
total B.S.

3. The SC tach signal is a "clean" square wave with levels between 1.0-1.5 volts
to about 6.0-7.0 volts.

Bottom line: Check the hookup of your tach & re-test. Also, make sure
you have a stock Bosch CDI. The signal posted is the typicql Perma-Tune,
& etc, tach signal.

[manu72S]

I have change D8 diode with a 1N4148. No signal on the TD pin, even when tachometer is connected. T3 collector has got a signal described by Lorenfb. The engine is totaly functional too. The tachometer is functional with an another CDI box.

Any idea ?

Thanks