Hi guys!

Another day, another issue.
So I drove into the city, did some stuff, came back to the car and noticed my speedometer needle wasn't moving.
After a short while (or after some travel distance up the road) it seemed to remember it had a job to do.
What I have been able to observe since then:
- Car sits for a while -> no speedo activity at first
- Speedo starts working again, sometimes after hitting 10 mph, sometimes at 20 or 30, so maybe a time-based thing
- May or may not cut out for a moment after first starting to work again
- Usually keeps working till the next cooldown (parking the car). Longest drive so far was roughly an hour
- no jaggy needle movements, no speed is being skipped

First thoughts:
- Vehicle Speed Sensor (VSS) gear
- Speed Buffer
- Instrument panel cluster or gage

After first research:
- My car has the electronic speedo, so no drive gear or cable
- The VSS is therefore also electronic: Transmission shaft turns the VSS gear, thus creating a magnetic field. If the gear was damaged, there's no way that magnet would turn until those teeth met again
- There is no speed buffer in this setup, the VSS signal goes straight into the ECM, which then converts the pulsed signal from the VSS into pulses the speedo understands and then sends into the needle coils
- No cruise control dropouts, no TCC issues, no SES light, so for now I'm ruling out the VSS/ECM combo as the cause (though the ECM will be the next stop on this road)
- A LOT of threads talk about leaking capacitors and faulty resistors

Aaand off I went into the tedious process of getting the cluster out to check for resistors and capacitors:
- Knee bolster and attached plastic thingy (and those finnicky plastic screw cover plugs that are never ever allowed to break!)
- Radio bezel (to avoid scratching)
- Accessory button bezel
- Instrument cluster bezel
- Steering wheel tilt lever (as it always gets in the way)
- Instrument cluster clear cover
- Instrument cluster inlay
- Instrument cluster (I could NOT get it out of there all in one. The steering column just sits slightly too high.)

Well then. We want to get at this PCB here:


There are 3 screws in the back for the PCB and 3 more for the speedo unit.

This is it. I have numbered all the resistors and will write down the values suggested by their color coding as well as the factual measured resistance values.




You can see this is nowhere close to those Camaro boards and there is nothing burnt out and not a single capacitor is bulding or leaking, at least not in an obvious way that I would have recognized.

Let's get to the promised values then:
All of these resistors have a golden ring, meaning a tolerance of ±5%#1: purple, green, black, gold, expected 75 Ω, measured 25 Ω -> totally out of whack, so what's right here: The expectation or the measured value?#2: orange, blue, black, gold, expected 35 Ω, measured 25 Ω -> also slightly out of balance
#3: white, brown, orange, gold, expected 91 kΩ, measured 90.5 kΩ -> within tolerance
#4: red, purple, black, gold, expected 27 Ω, measured 27 Ω -> spot on!
#5: brown, blue, orange, gold, expected 16 kΩ, measured 15.8 kΩ -> within tolerance
#6: orange, black, orange, gold, expected 30 kΩ, measured 30 kΩ -> spot on!
#7: brown, black, orange, gold, expected 10 kΩ, measured 9.9 kΩ -> within tolerance
#8: red, red, red, gold, expected 2.2 kΩ, measured 2.15 kΩ -> within tolerance
#9: brown, red, orange, gold, expected 12 kΩ, measured 11.8 kΩ -> within tolerance
#10: red, red, red, gold, expected 2.2 kΩ, measured 2.14 kΩ -> within tolerance
#11: red, red, red, gold, expected 2.2 kΩ, measured 2.22 kΩ -> within tolerance

I also noticed 3 diodes like these, which only work in one direction and they also checked out fine:


I'm not sure about these next things here. A friend mentioned they could be powerbased A/C resistors, which I would think very strange for a D/C circuit, but he pointed me to some on the net that looked similar.

Anyway, I measured them too, and one was exactly 20 Ω, the other 24 Ω, just as it is printed onto them.

Soooooooooo...
Any ideas?

I reflowed all the solder joints on the back side of the PCB, just to be sure, but that didn't help at all.
As long as this is intermittent, I wouldn't know what to do about it. I can't drive around with my laptop hooked up all the time, just hoping to look at it at the right moment to see what the ECM has to say about this.

 

Remove the electrolytics and check them with an LCR meter to make sure they're in spec and don't have some ridiculous ESR. Or pony up for a capacitor wizard and you can check them in circuit. Caps don't always degrade visually, they can dry out but look totally fine. Prolonged exposure to high temperature will accelerate this form of failure.

 

I have ordered an LCR last night and will see what I can find out next weekend. Thanks.

 

Update

I removed and measured these five bad boys today:


The results are in:
#1: Rating: 33 μF, 35 V, measured: 36.58 μF, Vloss: 2.2 %, ESR: 0.3952 Ω -> seems OK
#2: Rating: 33 μF, 35 V, measured: 37.59 μF, Vloss: 2.3 %, ESR: 0.32 Ω -> seems OK
#3: Rating: 33 μF, 16 V, measured: 38.48 μF, Vloss: 4.7 %, ESR: 2.3 Ω -> this looks kinda bad to me
#4: Rating: 4.7 μF, 25 V, measured: 4.786 μF, Vloss: 0.7 %, ESR: 0.5 Ω -> this one looks like new
#5: Rating: 47 μF, 25 V, measured: 46.18 μF, Vloss: 2.4 %, ESR: 0.71 Ω -> seems OK

What's your opinion on these values?

I haven't checked those goldish ceramic capacitors or the blue droplet style ones yet, as my desoldering syringe seems to have gone missing at some point and I already had a hard time with these guys.
I'm going to restock my equipment, get some flux and a bunch of new capacitors (all of those) if I can figure out all the ratings, and then I'll swap those 5 out and, if so required, the rest of them.

 

I'd can #3 for sure and probably #4 and #5. Ceramic disc caps don't go bad I believe so I wouldn't worry about them. Those are all pretty normal value caps so if your doing this I can imagine it wouldn't hurt to place an order with Jameco or Digikey and add them to your stash. If you really want to get into it, pick up a cheap function generator from AliExpress and pump a 5v(?) square wave into the input from the buffer. Poke around with a scope and you might be able to pinpoint your problem. Bench testing is a hell of a lot safer and more repeatable than doing it on the road.

 

Thanks for the advice and confirmation!
I have read that about the ceramic capacitors, too, and that you could even de-age them by exposing them to 150 °C for a moment, because they then recrystalize or something like that.
Anyway, since those don't provide any voltage info (same goes for the blue tantalum pearl ones), I'm going to leave them alone for now.
I did order some of the regular caps and will replace them just to be sure.
And thanks for the tip about the function generator. I was wondering how I could pulse that board to get it to do something. I'll check it out!

 

I have swapped out capacitors 3, 4 and 5 today. Still no speed. Instead now a wonky tach that can't decide what my RPM is.
That may be because I don't know any of the tolerances of those old caps and the replacements tended to go upward of their rated capacity by a much larger margin than the old ones did.
And the capacity is the one value you're supposed to keep close to the original one. Rated voltage can go higher, which makes finding replacements easier, yet did not enable me to find caps 1 and 2 in their original size.
New as I am to this stuff, I'm not sure whether it would be OK to install caps of a smaller diameter instead and only found conflicting information on that.

Another lesson I learned is that caps are available in several dimensions, even though the main ratings are the same, so here's the dimensions (diameter x height) of those caps, for anyone who might need them.
#1: 8 mm x 12 mm (or 11.5 mm, somewhere around that area)
#2: See #1
#3: 5 mm x 11 mm
#4: See #3
#5: See #3

Additionally, there are the brown ceramic capacitors, rated simply 103 and B 471.
103 is 5 mm wide, B 471 somewhere between 3 and 3.3 mm. Apparently the diameter has to do with their voltage rating.
The ones I ordered were about 1.5 mm and rated for 50 V, so they wouldn't have helped anyway.

There's a red capacitor labelled .U: 104... I guess that one's from panasonic. It's 5.94 mm wide, if that matters.

And the blue droplet ones are tantalum caps, ~3.2 mm wide.

So what did I do then?
I put the old caps back in and went for a drive with my laptop on the ol' ALDL connector.
Yep, no speed from the ECM for sure.
Man, this is so typical. I read about the caps and resistors issue and wanted to beef up my electronics know-how and skills anyway, so I chose to go the long and hard way.
At least I now have some capacitors at hand, an LCR box to test components with, a new solder syringe, flux, conformal coating paint, a field generator (yet no BNC cable, because that's still on the road) and more knowledge about caps.

Well, so this ends my investigation into the T/A speedo for now and I will have to focus on whatever's between the VSS and the ECM. Hopefully just a loose, oily or corroded connector somewhere, because anything else would stump me, considering that this issue disappears after being on the road for a moment.

Thanks for your valuable input, folks.