Once the new cluster showed up, I immediately installed in into the car to make sure it worked, or at least take stock of the parts that didn’t. I turned the key on and I was immediately greeted with “bar: 0.3” in the center LCD, success!

I turned the lights on, and there was but a dim glow from the upper right corner. I started the car, everything just seemed to work, I went around the block, and counted 3/10 of a mile. Success!. I decided to run with this new cluster for a few days until I could get to replacing light bulbs, and decide what I wanted to do with regards to the odometer reading.

Of course, the next day, the odometer stopped counting at about 18 miles. Not unexpected, but an occurrence which, at this point, was very discouraging. At least I had the new gear set on hand.

On a day when I had ample time to mess with it, I engaged in major cluster surgery.

I swapped all the bulbs from the old cluster to the new, I swapped the rebuilt LCD control board in, I swapped the new capacitors into the speedo/odo control board. I disassembled the speedo/odo to set the mileage.

The car showed, 189k when I got it, I’d been keeping my gas receipts, so I knew how many gallons I’d burned through. Taking the number of gallons times the EPA avg fuel economy, gave me 2200 miles. So, I set the odometer to 191k — good enough for a 35 year old car. I, of course, also installed the new odometer gear set, I even spun things with a long small screwdriver to be sure nothing was hanging up.

With everything buttoned up, I reinstalled the cluster into the car and took it for a test drive. I got 2/10 of a mile before it stopped counting. In addition the speedometer wasn’t working right, it would barely register over 20mph. I was beside myself. I took everything apart again, using my long screwdriver to move the gears inside the odometer, I could tell something was hung up, I pushed a bit and things started moving again. In a burst of optimism, I put the cluster back in the car and went around the block. 1/10 of a mile before it stopped counting.

I took a deep breath and pulled the cluster out again. I disassembled the cluster AGAIN, and pulled the speedo/odo pod out AGAIN. I grabbed my bag of alligator clips and jumper wires along with a 9v battery in order to simulate the input the speed sensor would send to the cluster in attempt to observe what happens. (I was so frustrated at this point, I didn’t think to grab a picture of this process). I was able to un-jam the works again with my trusty screwdriver, hooked up the wires and started simulating input (this is accomplished by making and breaking contact to ground very fast in the hopes of generating a square wave signal. I observed the motor doing it’s step, step, step, twitch, twitch, twitch. I un-jammed it manually, and simulated input. It would step properly for a bit, then start twitching, I would unjam it. I continued this for a while before I took the pod apart again to check for debris, to make sure nothing was getting hung up…nothing.

Back together and simulate. At some point it seemed to be getting further between jam-ups. When I hit a full mile on the trip meter, I started having some hope. When I got it to five miles, I said, okay, good enough, lets put it back in the car. I also cleaned and added a little lubrication to the speedo needle shaft, where it passes through a nut which holds the hole works together. That clear the hangup there…thank goodness.

On the next day’s journey, it counted 1/10, 2/20, a mile and beyond. All the way to 22 miles….and stopped. I recalled, that when simulating input on the bench, resetting the trip meter sometimes had an effect. That action seems to jostle things just enough to sometimes shake things loose. Indeed, I reset and it and it started counting again. It got to some number of miles over the next few days. I stopped and topped up the tank and reset the trip meter. I was determined to finally be able to calculate my gas mileage. The Board computer, which displays through the center LCD was showing something silly like 60 mpg, which was clearly wrong (Will address this in another post). As I type this, the tank is near empty and the trip meter shows 257 miles. Honestly, I’m a little afraid to reset the trip meter, but I imagine whatever it was getting hung up on has clearanced itself.

So that’s it, it took 10 months of ownership, about $200, countless hours worth of poking and prodding, and countless remove/replace cycles of the cluster to get everything to work as it should. I must really like this car, because the time and heartache investment in just the instrument cluster, is objectively not worth it.

Note: This work was finished in at the end of December. It’s now beginning of April and everything has just been OK with the cluster! We’ve counted almost 2000 miles without missing a beat!

The procedure to pull the cluster couldn’t be more straight forward.
1) Remove the horn pad (just pull)
2) Remove the nut holding the wheel to the column, and pull the wheel off.
3) Loosen the clamp which hold the Multi-function switch to the column, and slide off.
4) remove two small screws from the cluster, pull it out and unplug 3 electrical connectors.

I knew I was in trouble when, as I completed Step 2 , I was showered with a dozen tiny ball bearings. You’ll note, that “pick up ball bearings and try to insert them into their mangled nylon carrier” is NOT in listed in the procedure above.

Having a suddenly acquired a new side-quest, I started searching the floor and by some miracle, was actually able to find all but 1 of them. Turning my attention to the mangled carrier, I was able to kind of bend it mostly back into shape, and pop the bearings into their spots. This bearing, sits just inside the column housing around the steering shaft, behind the steering wheel, offering support and smooth operation. Without it, there is a disconcerting amount of up/down left/right play in the wheel, along with unhappy metal scraping sounds when the wheel is turned. Thankfully, the bearing carrier seated well in the column. It seated so well, I was unable to remove it again, which make me wonder what kinds of abuse this poor car suffered before it came to me. Regardless, with this intermediate crisis out of the way, I proceeded down the list.

Once the cluster was removed, It was immediately obvious that I’d acquired another side quest. The contacts between the various bulbs and needles and the electrical connectors, consist of very very thin strips of copper, embedded in a fragile plastic sheet. The plastic contact mat was torn in several places, preventing signals from getting to where they needed to be.

Repairing a broken circuit trace is generally a straightforward operation. You can either bridge the trace with solder if it’s a small crack, or run a bodge wire to give a new electrical path. This is a not-uncommon occurrence when working with old computers in general. Unfortunately, given the light plastic nature of the “board”, it would simply melt and shrink when the soldering iron would come near it, so it was VERY challenging to try and bridge the broken the traces, but I persevered and managed to make some repairs which showed that I had restored electrical continuity.

So, with repaired traces and all new light bulbs, I plugged it in, turned the ignition switch to “on” and learned that while the fuel gauge came back to life, the center LCD did not. Bummer. I clicked the lights on to be sure all the new bulbs worked and several bulbs in the cluster EXPLODED.

Sigh. So, I took it apart, cleaned out all the broken glass, and replaced the exploded bulbs and the various newly burnt out ones, with old working ones. After getting everything back together I tried again. Success, illumination and no explosions!

All of the above happened while I was working on the breaks and the power steering pump. Once I started driving the car, a few things became apparent. First I really wanted a boost gauge in order to know what the turbo was doing. The boost gauge is a digital readout, provided by the non-functional center LCD. Second, the check engine light worked, I know because it would light up every time I accelerated. Third, by some miracle, the odometer still worked (this is a known weak spot for gauges made by VDO from this era) — at least for 36 miles, when it stopped counting miles.

That the odometer broke wasn’t too surprising or upsetting. There is a set of nylon gears which interfaces between a stepper motor and the gears behind the numbers. The gears simply crack and fall apart due to age. This is such a big problem for every 80’s and 90’s European car with a VDO speedometer, that new gear sets are readily available for a few bucks. I removed the cluster (again) and fully disassembled it. I found no cracks in the gears, which perplexed me. As I opened things up, I was hit the face with a vaguely fishy smell, which is indicative of leaking electrolytic capacitors. Dead/leaking caps can absolutely cause circuitry to just not work, so I took inventory and placed a Digi-Key order. I also ordered a new gear set. I put everything back together and reinstalled it, so I could continue to drive the car.

Once all the parts came, I made quick work of the capacitors and the gear replacement. I did find a missing tooth on the old gear, which helped explain the stoppage.

I put it all back together and turned the key on, but I still had no center LCD – at this point I figured it was just a bad LCD panel, and I’d have to source a new one. I started the car and drive down the street. It through clicked 2/10 of a mile and stopped counting again.

So, I took everything apart again and found a small bit of debris in the gear track. With that cleaned out, I figured I’d solved it and put it all back together and reinstall it into the car. 1/10 of mile and it stopped. At this point, I was beyond frustrated and was questioning if an oily rag and a match might be a better solution to this issue. I took everything apart AGAIN and decided to test the circuitry itself on the control board for the speedo/odo. I knew it was getting signal from the speed sensor, as the speedemeter worked without fail.

So, on the bench, I simulated input to the board. I measured voltages coming out of the IC chip and everything checked out, the motor was simply not turning when it received voltage. It’s at this point I noted, that the motor, wasn’t sitting flush to the board, a very small amount of wiggling and it popped off the board all together — except the pins which are used to solder the motor to the board, which were still soldered in place on the board.

It turns out, this is possibly one of the worst designs I’ve ever seen. The windings of the motor, which are about as thin as a hair, are soldered to a metal pin, which is held in place with nothing but friction to the plastic housing. The other end of the pin is soldered to the circuit board. That’s it, there is no other mechanical attachment of the motor to the board, so with any movement or manipulation of the board, you risk damaging the connection from the windings to the pins, irreparably damaging it.

I mentioned above that tons of cars with VDO gauges have an issue with the nylon gears. Unfortunately, almost no other car uses this particular motor and control board. It seems only 86-88 Audi 5000 Quattros and similar year Volvo 740s used it. I’m guessing that VDO realized they had a turkey of a design on their hands as all other cars from this era that use VDO gauges seem to have a far more sane motor/board interface.

This also means that it’s incredibly hard to find a replacement. There are no markings on the motor itself, so I needed a donor cluster from one of the above cars. The only thing I was finding on E-bay was a cluster for $450 out of Latvia. Considering the dead center LCD, I needed a donor cluster anyway. But $450 seemed a little too steep, so I kept periodically checking E-bay, hoping for a cheaper option.