First, a bit of revision. The tacho signal (TAM) is generated by the ECU and sent to the speedo Printed Circuit Board (PCB) in the instrument cluster which then drives the input to the Tacho PCB. TAM runs at 2 pulses per rev.

AlienR2 in two sequential posts captured the tacho signal coming from the ECU then he deduced the circuit on the speedo board that processes it: https://www.rx7club.com/interior-ext.../#post12212569. (DO NOT follow his thought that it is OK to connect TAM direct to the tacho. Some have done it and claimed that it worked but it is possible that it will overload the tacho IC or ECU and lead to a shorter life. Without a data sheets and internal circuits for both it is impossible to be sure that it is safe.)

The line from the speedo to tacho is driven high to 8.75V by circuitry on the tacho board. When a TAM pulse is received, the speedo PCB circuitry applies a voltage to the base of an NPN Transistor on the speedo PCB which switches the tacho line to ground for the 1.5 milli seconds where the TAM signal goes high, yielding a waveform like the one below with a 1.5 msec drop to zero on every TAM pulse:




Fortunately, the tach circuit is not at all sensitive to the duty cycle of the waveform and reads the same over the full rev range receiving a 50-50 squarewave. At 9500 rpm the waveform is actually a 50-50 square wave.

Interestingly, the tacho drive circuit steps the motor currents up and down at 1 Hz intervals which equates to 30 rpm increments. In practice, a driver would not be able to detect the steps under driving conditions.

Enough revision. For this task, you will need:

• A smattering of electronics knowhow or a tame electronics Geek.
• A CRO and Signal Generator. There are numerous low-cost ones on Ebay etc. The signal generator should be capable of at least a 2V Peak to Peak square wave output over a frequency range of 25 to 300 Hz. Caution: check the specs, some only work at much higher frequencies.
• 12 V power supply
• A small NPN Switching Transistor. I used a BC546
• 3 x short M3 x 5 to 12mm long screws
• Breadboard (optional), hookup wire – thin and flexible is best, multimeter …..

Remove the cluster and disassemble. YouTube guides here: , and here:
You only need to remove the gauge face if you want to get to the underside of the board. If removing the gauge face follow the 2nd video carefully.

If you have got this far, I recommend replacing 2 low reliability components as a matter of course. See here: https://www.rx7club.com/3rd-generati.../#post12619950

Connect to the three threaded holes in the back of the tacho using short M3 screws in accordance with the following circuit:


Check the DC output impedance of your signal generator. Mine was 600 Ohms so it didn’t need another resistor between the signal generator and transistor base. If you are much under 400 Ohms, I recommend adding extra resistance in line to the base just to be safe

At this point it you should also hook your CRO up to the collector to confirm the signal going to the tach. My set-up below. Why do I always end up with a bench looking like this?:





Next, set-up the signal generator. You want a square wave signal with at least a positive 750mV peak. If you can, offset the waveform so as it drops to zero. If you can’t offset make sure that the negative does not go below the VEBO of your transistor; -6V for the BC546. You are now ready to set the input frequency. The tacho runs at 33.333 Hz per 1000 RPM.

There end details of a set-up for trouble shooting. More info over here: https://www.rx7club.com/3rd-generati.../#post12619950

Now for calibration. While the drive circuit steps the currents up and down at about 1 Hz intervals, keep at least 1 decimal place in the frequency as you can’t be sure where precisely the tacho will sense an increment point. The following table might help:



What worked for me with a series 8 tach with the zero at 6 o’clock was to set the gain, using the potentiometer on the tacho PCB, to show a needle a bit on the plus side of 6,000 rpm at 200 Hz. Then, drop back to 25 Hz and see what the error is at 750 RPM.

Adjust for the idle error with the tacho powered down. Swing the needle around to 9,000 rpm. Note the needle position at the stop just past 9,000 rpm. Remove the needle and replace it firmly, but not fully, back around 8500 rpm then rotate the needle against the stop to the point where it cancels the 750 rpm error. You can only position one way. If you go too far you must restart the process.

Recheck at 6,000 rpm and 750 rpm until satisfied, then check the other increments. As others have noted, these gauges are not real accurate. The gauge also has come lag so I recommend erring slightly on the high side in the higher RPM bands and as accurate as you can get at cruising speed.

My calibrated zero point on completion:



My final calibration:

 

This is some excellent info, thanks for this write up! Still glad my Youtube videos helped out in some way .

 

Great stuff!

I know there is also a pot on the tach board that can do some adjustment.

Dale

 

I will edit the post to specifically state that you adjust the gain with the potentiometer on the Tacho PCB.

Zero is set physically by adjusting the needle relative to the armature shaft, the readings around the dial are adjusted by the potentiomenter on the tacho PCB. It wont be precise everywhere, it is up to the person doing the adjusting to determine a best fit for their purposes.

 

Great, comprehensive write-up RGF, just what i was looking for to trouble-shoot my intermittent(more so dead Tacho now after pulling dash out a few times).

Before I pull apart the tacho itself(already done once to replace those 2 caps, although VR1 may be worth looking into), can I just confirm the the 1st signal in your thread(8.75v in amplitude) is the TAM signal at approximately idle or the Tach signal from IC4(out1) on the collector of TR3?

I want to put a CRO at the TAM connector(3F on cluster) to see if i'm getting that signal before delving further into the cluster circuit(IC4) & then the tacho cct.

Cheers

 

I owe the use of the term: “TAM” to AlienR2 in the post referenced in the 2nd paragraph of the first post. I can’t find any other useful reference to TAM in the service manuals or the net in general. Nevertheless, I have adopted “TAM” to refer to the signal sent from the ECU to the Speedo PCB.

The TAM signal can be seen in AlienR2’s post.

The signal in my post is not TAM, it is taken at the input to the tacho.

I have never actually looked at the TAM signal so can’t positively verify that it exists at 3F. If TAM exists at 3F at the instrument panel then it might be easier to get to at the diagnostic connector, IG-terminal, or the ECU plug – varies with year.

For me, the tacho input seemed to be easier to semi permanently wire up and provide more initial troubleshooting information, especially if you can view it in the car when the tacho is playing up:

• Is 0V floating away from ground?
• Is the 12V supply holding up?
• Does the signal between speedo and tacho remain clean?

 

Ok thanks for clarifying that, I'll checkout AlienR2's post.

I think you will find "TAM" is printed on the back of the cluster, which goes to the socket where wire 3F resides.

My plan is to first check continuity from the IG point to 3F, although it is meant to go through 2B on ecu, which mine has been swapped out with a Haltech so only assuming at this point it's connected.

12v & gnd seemed to be good when I checked at the Tacho screw points with engine running), but I was using a multimeter on TAM/3F & Tacho signal which is a bit useless in this case to notice any difference as it's just measuring Rms voltage.

 

Thank you so much for sharing it with us. I appreciate you for taking the time to write it in brief.