Speedometer Repair (Electronic)

[ClickClickBoom]

Like the title says, my speedo is the electronic version, mfg 12/83. It has always worked pretty good. I have a Yellow Box correction unit installed, my speed has always been within 1-2 mph of the GPS and was dead on, with a little wiggle/nervousness.
I have been noting that the speedo would drop off line and mysteriously return randomly, hot cold, no rhyme or reason, but so infrequently I just lived with it.
I have ordered a Classic Retrofit alternator(whole nuther post to follow), because my voltage had dropped from 14.1 with boring regularity to 13.5 to 13.8 randomly. The PO had installed a "rebuilt" alt prior to my purchase. With no reciept showing the rebuilder, I was expecting it to go off line almost at any time.
Well it didn't fail, but the regulator spiked the voltage to 17.1 volts for about 90 seconds. Speedometer joined the club, and went on strike, no amount of CPR or trickery could wake it from its Snow White slumber.
Fast forward to the interwebz and the consensus is the electrolytic caps being 35 years old tend to quit quite frequently.
Yanked the speedo, got my small screwdrivers and pried the bezel off carefully. There in all their glory were 2 of the dreaded FRAKO caps in the 47mfd and 100mfd x 16V flavor, there were no obvious signs of distress. Mild panic set in, eBay was consulted, prices are Gold Standard Porsche tax grade for speedos.
Hopped on the motorcycle, scooted to the local electronic supply and purchased 2 each of NTE 47/26 and 100/26 electrolytic caps. Rushed back home, and broke out the HAKKO soldering station, and associated tooling and removed the offending slacker caps, and replaced them with the new high temp grade NTE replacements. Buttoned everything back up, plugged the requisite wires back to their rightful places in the electron circus, and took it for a spin.
Whew, everything worked even better than before, needle is smoother and rock solid.
All in all if you possess an decent soldering iron and some magnifying equipment, the job is a solid 2 banana job. Requiring little more than minimal soldering skills and the willingness to dive in and get er done. One might note, I replaced the caps with a higher voltage rating capacitor but nailed the mfds right on.
I would suggest that if you have a electronic speedo, and havn't had any work done, to replace the caps. There were multiple posts with burnt resistors with the cap failures and even a couple IC bugs failing from the voltage spiking. Not being an electron head, I don't know if they were collateral damage or the cause of the failures.
Get out and drive!
eric
P.S. for the members with short term memory issues, take pictures and make notes.
I didn't and it took me an extra hour to jigsaw puzzle everything back together.

[ClickClickBoom]

[ClickClickBoom]

Final notes, new caps worked flawlessly, whilst driving my speeds would sharply drop by 20-30 mph, go completely to 0, then back to correct speed. I thought the voltage spike wounded my yellow box. Disconnected the yellow box completely, speedo was dead as a doornail. Crap, Occams Razor says the last significant thing I did was the speedo.
Checking the speedo is almost impossible without a signal generator. I did have a sensor on the 915/67 in the corner of the shop, I tried to get a signal by sticking a magnet on the lathe and spinning it by the sensor and hooking up the ohm meter, nothing from either my OEM 128,000 mile unit or the one from the /67. I put the 915/67 unit into the car thinking, "gots nothing to lose" While running errands after 20 miles or so the sender came to life, speeds normal, working pretty good. So the speedo wasn't the issue after all.
Ordered a new Porsche OEM sensor, put it on the bench and hooked the multimeter up to it, waved a magnet 6" from the unit resistance went from infinite to almost 0 just by waving the magnet at it.
All these shennagians started after the stupid voltage spike. The sensor appears to basically open and close the ground for the speedo signal, so maybe the sender was close to its final days and the overvoltage pushed it over the edge.
I guess 128,000 miles wasn't bad for the original unit. From the FSM schematic it appears to be a reed switch, which means its a mechanical device, kinda amazing.

[ClickClickBoom]

Well,
The voltage spike claimed it's last victim, the beloved yellow box. As noted above, I first replaced the capacitors in the speedo, then the trans speed sensor, and at last the yellow box. All the symptoms above were displayed. The final symptom was, the speedo worked perfectly for about 20 minutes then failed dead. I disconnected the yellow box, and reconnected the speedo, works perfect, 11% fast, but works dead smooth. It should be noted this wasn't merely throwing parts at a problem, it was in response to an initial condition(overvoltage), then a series of cascading failures. Stupid alternator, good thing I replaced it with the Classic Retrofit alternator, and new shroud to boot.

[Solamar]

Originally Posted by ClickClickBoom

Checking the speedo is almost impossible without a signal generator. I did have a sensor on the 915/67 in the corner of the shop, I tried to get a signal by sticking a magnet on the lathe and spinning it by the sensor and hooking up the ohm meter, nothing from either my OEM 128,000 mile unit or the one from the /67.

I have a couple Speedos I want to function test and have been thinking of an acceptable test. I have a Hall effect sensor that I think will work.

While you were spinning the magnet to generate the pulse, did you have 12v and ground hooked to the speedo?

[ClickClickBoom]

Originally Posted by Solamar

I have a couple Speedos I want to function test and have been thinking of an acceptable test. I have a Hall effect sensor that I think will work.

While you were spinning the magnet to generate the pulse, did you have 12v and ground hooked to the speedo?

The sensor on the trans is a magnetically switched reed switch. The sensor shows either open(infinity) or closed(2-3 ohms) when there is a magnet within 6”. Pretty sure it isn’t Hall effect. To test my speedo, I jacked up one side of the car and ran through the gears(no LSD). Easier than driving and quicker.

[Walt Fricke]

Solamar
Testing for base function is easy. Hook up the speedo ground and positive to a 12-13VDC source. Clip a jumper on the signal terminal, and rapidly touch it to the ground terminal. This simulates the reed switch alternately grounding the signal wire and making it an open connection. The speedo needle will rise, how far depending on your technique and speed. A ham radio guy could hook up his Morse key and probably get a very steady speed.

[Solamar]

Originally Posted by Walt Fricke

Solamar
Testing for base function is easy. Hook up the speedo ground and positive to a 12-13VDC source. Clip a jumper on the signal terminal, and rapidly touch it to the ground terminal. This simulates the reed switch alternately grounding the signal wire and making it an open connection. The speedo needle will rise, how far depending on your technique and speed. A ham radio guy could hook up his Morse key and probably get a very steady speed.

Appreciate that Walt, great info!

Funny enough, I am a Ham radio (cough) hack (cough). Don't think I still have a key around but still have this old Heathkit frequency generator. Probably the wrong wave type to excite the speedo though...

I'll try the tapping to ground, have a couple to test.

[Bigtoe32067]

Hold up. I'm confused! Where do the bananas come into play?

[ClickClickBoom]

Originally Posted by Bigtoe32067

Hold up. I'm confused! Where do the bananas come into play?

Check your tailpipe....