[NA] Fixing an old glitchy tach

[drummer4468]

Hey all,

Shot in the dark here, I've been trying to run down a problem with my old '95 Nissan Skyline's tachometer(electronic, not cable driven). Sometimes it's dead and resting on the stop, sometimes it's jumpy and random, and most often it works kinda "normally" but chooses an arbitrary zero point around 5k rpm or so. This leads me to believe it's getting a good signal from the ECU but fails to translate it to the needle accurately. Again, it's a '95 so there really isn't much computerized wizardry to contend with. The cluster circuitry itself is astonishingly simple

I've already re-soldered all the points between the cluster board and motor, and there's no change in behavior. I know this is a common problem with these clusters, and pretty much all I've read just says to re-solder any bad connections. Anyone more deeply familiar with them have any advice? These clusters are only getting harder and more expensive to come by, and I'm trying to keep everything as "stock-ish" as possible.

Any direction is appreciated!

 

[Brad W]

I'm not at all familiar...so not the answer you're looking for....but my 1st shot in the dark thought is bad ground...and maybe even at the windings that move the needle.

 

[Zeldman]

Have you tried rapping on the front of the tach with your knuckles.??

(that's all I got...)

 

[guest user]

Just thinking (typing?) out loud here.

The flow would be Tach sensor -> wires -> ECM -> wires -> cluster -> motor -> needle (which yes, is mechanically connected to the motor)

Each of these could be corrupt:
* Note that there is nothing saying it cannot be a combination of the below either

Tach Sensor
- bad sensor

Wires
- if the wire is exposed there could be rust, or it could be intermittently grounding somewhere
- if the wire is damaged internally, it could be flexing and causing an intermittent connection that way

ECM
- as the originator of the signal, could give bad data
- doubtful that the code is being inconsistent, so probably based on input rather than physical damage due to inconsistency
- improper seated connector (although this should mess up a lot of other things)
- improper seated pin in the connector, not making full contact

Wires
- Same as above

Cluster
- if the cluster is not transferring the signal to the motor due to bad solder
- if the connector is not fully seated
- if the tach pin is not fully seated in it's socket

Motor
- as stated, the windings
- an internal short

-------------------------------------

Troubleshooting suggestions:
* motor here is the small motor driving the tach needle
** signal from ECM means the voltage/pulse width/etc of how the ECM communicates with the tach

(The below is not specific to your car, nor are there any special knowledge hidden, but here is how I'd approach T/S this issue)

- find what signal you're expecting from the ECM
- Test for that signal at the motor.
- If you find that signal at the motor, then rev the car and see if the signal changes as expected and smoothly.
- If signal going to motor smooth and correct, but actual needle movement is erratic, I'd replace the motor.

- If signal to motor is bad, I'd back up and check the circuit board
- if the signal at the board is fine, but the motor signal is bad, look at the connectors from board to motor

- If signal to board is bad, back up and look at signal to the wire harness on the circuit board end
- If the signal to the wire harness end is fine, I'd verify the connector on the board for the tach input

- If the signal to circuit board side connector is bad, look at the signal from the ECM side connector
- If the signal on the ECM side connector is good, then you get to inspect the wiring to find bad spot
- - If you can't find a questionable spot on the harness, you can always replace the wire itself

- If the signal on the ECM side connector is bad, then I'd jump to the other side of the ECM and test from tach sensor and harness back to the ECM
- If signal is good to the ECM but bad from the ECM, then there's your problem
- If the signal is bad to the ECM, I'd check at the sensor
- If signal is good from sensor, check/replace the wire
- If signal is bad from sensor, replace sensor


Hope this helps.

 

[SoonerAviator]

You might search some of the Z-car forums (ZCar.com, TwinTurbo.net) to see if they have some solutions for it. Probably the same components as the 300ZX ('90-'96) and a much larger audience in those forums.

 

[drummer4468]

Have you tried rapping on the front of the tach with your knuckles.??

(that's all I got...)

lol there’s been plenty of knuckle-rapping and wire-jiggling. Was hoping it was just a bad ground or something, but after cleaning and tightening them there was no change.

Just thinking (typing?) out loud here.

The flow would be …

thanks for the write-up. Luckily all the parts and wiring should be fairly easy to access as this engine uses a cam angle sensor right up top for all those signals. My only problem is I believe it uses a pulsed signal as opposed to varying voltage. Meaning my multimeter won’t cut it, so beyond ohming out the wires I’ll need to get my hands on an oscilloscope to chase the signal itself.

I’ll pull the cluster apart again and check deeper for any bad/dry solder joints. I didn’t notice any signs of anything burnt out or blown the first time around, but I was mostly focused on the motor connections.

the pulsed signal is what confuses me. The gauge is smooth most of the time, just reads the wrong rpm. I’d think bad wiring or dropped pulses would make it very erratic and/or read low, but for the most part it’s not, just idles at 5k-ish rpm and goes up from there. So maybe it is a bad capacitor or resistor somewhere.