1992 Toyota Truck Speedometer

[ranchhandbumper]

I am working on a 1992 Toyota pickup that has an electric speedometer. It is an analog needle speedometer and is driven by electricity.There is no cable. The transmission has a reed switch that sends a 12 volt pulse to the speedo 4 times per tire rotation.

According to GPS, the speedo is showing speed progressively slower below 50mph and progressively faster above 50mph. This also affects the odometer.

I've included a photo of the findings. The "control" is information I found on the net that correlates expected MPH based on a given hertz.

I removed the speedo and applied 12volt control voltage to the IGN and GND. And connected a function generator (square wave setting) to GND and SEN. This setup reproduced the error exactly as seen in the truck.

I replaced the 10uf and 100uf cap which were both looking pretty rough and cleaned all the circuit board and resoldered a bunch of components.
Also tested all of the resistors and they appeared to be well within specification.
What you see on the board at the top and right side is apparently pretty classic for these speedos. Every one I've seen on the net looks like this with the black discoloration. Must get a bit warm. I ohmed all those paths and found no broken connections and no high resistance.

Upon further testing, I discovered that if I run the control voltage down, the speedometer becomes increasingly accurate. It ceases to function below 9 volts.

I have read in a couple of places that the speedometer uses a stepper, but I'm not sure that mass of copper wire with the metal cup around it is a stepper. I finally saw a writeup somewhere that mentions a "moving coil movement" but gave no explanation.

I am at a bit of a loss and wondering if anyone here has any ideas.
Thanks!

[redwire]

The TB9213P drives the speedometer air-core sine/cosine meter and TD6905S drives the odometer stepper motor.
I would check the voltage regulator output, which is at ZD2/R6/gray power transistor and across the 100uF cap and at IC pin 16. The speedo takes in vehicle power and regulates it down to around 8.5VDC, so check the TB9213P IC pin 16 is around 8.5V, I think pin 9 should be around 5.2V; regardless of input voltage. It takes the input pulses and uses the on board quartz crystal to generate sine+cosine waveforms to drive the meter movement.
Otherwise I would check C5 is not cracked or damaged. It should be around 0.1uF or maybe larger for filtering noise.

[stj]

resolder the crystal,
when they are mounted like that, vibration slowly stresses the solder joints

[ranchhandbumper]

Thank you for the direction! It'll be few days before I get to it, but will post my findings.

[R_J]

C5 is not connect, the trace is damaged, I would also check ZD2 connection, That whole trace should be checked as I suspect it has taken on moisture at one time.

[CapLeaker]

… and I see rings too.

[redwire]

How does water get in the speedometer? Toyota has many speedometer boards with this damage on them. It might be a leaking front windshield.

I also see SMT R2, R4 not looking well soldered on one end.
With these old automotive boards, I will resolder all through-hole parts as the "rings" or fractures get worse with age. R7 pad has that happening.

I wonder if some lube light oil for the speedo motor is also needed.

[stj]

Originally Posted by redwire

How does water get in the speedometer? Toyota has many speedometer boards with this damage on them. It might be a leaking front windshield.

maybe condensation from the inside is getting into the housing.

[redwire]

It's along the top edge of the board - it must be some sulphur off-gassing attacking the copper, or high current is pulled through the traces during cranking causing heating.
I see the same PCB construction and black corrosion/soldermask in Landcruiser or Toyota Town Ace or Carina or Volvo 850 etc.
Weird.

[ranchhandbumper]

Thank you all for the input:

Pin 16 is showing 8.75v at 12v input and 11v at 14v input (which i would assume is the more likely input voltage at cruising speed with a good alternator.) Does that mean ZD2 is bad? The resistor checks out ok.

Pin 9 outputs 5.4v to 5.5v no matter the input voltage.

I resoldered ZD2, the crystal, and C5 for good measure all to no effect. I will resolder the other suggested components, but thought the 11v on pin 16 was probably significant.

[R_J]

Did you repair the corroded traces? From other pictures it seems Z2 is a 9.1 Volt zener. The purpose of a regulator circuit is to regulate the voltage so after about 11v input to the transistor the output should remain around 9 volts

[ranchhandbumper]

I have attached a picture. I cleaned the traces. I did not repair. Should I solder wire to each terminal on those traces? I haven't "repaired" because I have what I would call good continuity on those traces. I have very little experience with sensitive electronic components though.

[R_J]

It seems to be ok now, there should be no need for wires, I see C5 and ZD2 are connected now. I would check the voltage across ZD2 to see if it zeners at 9.1v like it should.

[ranchhandbumper]

heyyyyy, I think maybe its a bad zener?

So I get continuity in both directions on ZD2. (24k forward and 24k reverse)

Diode test on DMM shows 0.9V in reverse direction and 1.4V in forward direcion.

Also when voltage is applied to the circuit:
9V input = 6.3V output in both directions
11V input= 8.3V ouput in both directions
14V input= 11V ouput in both directions

Also, I have observed that the leads on the crystal (which I resoldered) are getting hot enought to be uncomfortable to leave my finger on. And could actually burn me if left long enough. This was the case before resoldering as well. Is this normal?

[stj]

crystals dont get hot

[ranchhandbumper]

Ok. I am thinking about the hot crystal backwards then. The whole trace is very hot and the heat happens to feel focused most on those two points maybe.

[stj]

you still didnt resolder the board, the pad near IGN and a lot of the lower half has "ghost rings" visible

[ranchhandbumper]

I am not understanding or don't know what to look for withese ghost rings. My untrained eye and multimeter says things are OK on the bottom. Of the board. I can resolder, but could you point out exactly on the picture one of these ghost rings. I appreciate all of your help!

[R_J]

Originally Posted by ranchhandbumper

Ok. I am thinking about the hot crystal backwards then. The whole trace is very hot and the heat happens to feel focused most on those two points maybe.

Check on the other side to see if a component is getting hot, It won't be the crystal but it might be the regulator transistor, or maybe the mov.

[stj]

the zener is near the crystal