What's the Frequency, Kenneth?

    The good news is that it's not some weird silicon breakdown inside that mystery IC or any other bit of unobtanium. It appears to to be yet another example of a pretty good piece of engineering let down by sloppy construction. It's really easy to fix. And after you've fixed it, it's easy to calibrate too. If you put a V8 or V6 in your TR7, it's even easy to make the original 4 cylinder tach work correctly with the new engine.

   Heck, it's even easy to remove the tach from the dash. Just pop up the fake speaker cover on top of the dash and remove the handful of screws that secure the big cover over the instrument cluster. Lift it off and slide up the clear faceplate. Remove the black gauge fascia, then the three screws securing the tach. If you're lucky, the three screws will be Phillips. If not, I've heard they can be dealt with using a precisely sized straight slot screwdriver and a little finesse. Once you've done that, just pull the tach straight out.

    The tachs in my Wedges have always worked perfectly, but in the course of examining some other tachs to verify the schematics on this page, I found out exactly what makes a 'sticky' tach sticky. I just fixed two of them, and found that the culprit's not the electronics, the design, excessive wear or poor quality materials. It looks like it's just the application of a little too much of a good thing.

    With the tach pointer is at zero RPM, the coil is at the 5000 RPM position and the top of its bobbin rests against a raised area of the main meter casting.

    The tach hesitates on startup because the coil is getting stuck to the casting by some contamination. If you very closely examine the coil bobbin you may see what looks like a very tiny shiny patch of goop. It's really hard to see. It does not look like it was placed there on purpose - it looks like the folks in the factory were overly generous with the soldering flux when assembling the tach and splatter and/or fumes got where they shouldn't. Over the years the stray flux absorbed moisture and contaminants and turned into a pretty good replica of the adhesive used on Post-It™ notes.

    I made a mini-swab out of the cotton from a regular swab and a toothpick and carefully cleaned the top of the coil bobbin and the raised area of the casting it rested against with some alcohol. The drawing above shows the details.

    After cleaning, both tachs worked perfectly. Even after sitting undisturbed for a couple days, the faintest breath would cause the pointer to swing.

    It appears that TR7 and TR8 tachs are exactly the same except for the redline on the dial and the value of one resistor. Fortunately, making a four cylinder tach work with a V6 or V8 requires reducing the value of that resistor, so another resistor can just be placed in parallel with it and no other hacking is required. The little circuit board in the tach has its resistors silkscreened on, but there are other components soldered to the board that the extra resistor can be easily attached to.

   A few words of caution: The ceramic circuit board may contain beryllia. NIOSH says it's something you don't want to inhale and you should avoid skin contact and wash your hands after handling it. So treat the board gently to avoid breaking it up and causing particles to fly.

    If you are converting the tach to work with a V6, you will need a 91K resistor. This is a 'standard' value. I doubt that using anything more precise than a ten-percent tolerance part is worthwhile; these tachs ain't no atomic clocks, especially after all these years. Besides, you can eliminate any small error by slightly turning the little plastic lever that adjusts the meter zero and adjusting the calibration segments that encircle the meter movement. See below.

    For a V8, you need a 48K resistor. That's not a standard value - I used a 51K and a 820K resistor in parallel. You may be able to use a 47K (standard value) resistor and just recalibrate the meter movement. Once again, see below.

    Hold the tach face down and look at the circuit board from the direction of 6000 RPM. You'll see a black rectangular integrated circuit with three striped capacitors to its left and one striped capacitor to its right. Directly on top of the IC is a blue cylindrical electrolytic capacitor.

    Solder the new resistor(s) to the closest-to-you leg of the blue capacitor on top of the IC and the closest-to-you leg of the striped capacitor to the right of the IC. Use care not to short things together with too much solder, etc., etc., etc.

    After you're done, secure the resistor(s) to the back of the dial with a drop of glue so vibration won't cause things to come adrift some day.

    With the tach reinstalled, you can use a tune-up tach to check the calibration before trying to hit the redline (which may now be different than marked!) or with it still on the bench you can use your PC to tweak it precisely.

    But then I got to thinking. It is the twenty first century, after all... so I did a little experiment.

    I fired up a little freeware program I had downloaded from the Web a while back that turns the sound card in a PC into a function generator. (I just checked to see if it is still available and got a DNS error. It may not exist any more.) There are numerous other free sound card signal generators out there (and plugins for programs like Audacity, too) that would work instead. A quick Google will find 'em.

    I plugged a cable into the sound card output and attached it to the very simple circuit shown above that can be made from a couple bucks worth of parts from Radio Shack (or scrounged from that old tape deck sitting in the basement that you never get around to throwing out).

    I connected a 12V power supply to the tach and fed it (the tach) the signal fom my PC. I set the program to output a square wave and fiddled with the amplitude and duty cycle. I looked at the signal with an oscilloscope and the frequency counter. The square wave has a fair amount of 'ringing' but that didn't affect anything provided the signal amplitude was high enough. I set it to put out a square wave with an amplitude of about 2 volts. That is what the sound card puts out with things turned most of the way up (turn off any tone controls). You don't want it much higher than that anyway, so don't run it through an external amp or anything. The duty cycle had to be set between 40 and 50 percent for proper operation over the entire RPM range.

    It worked perfectly. A PC signal generator is extremely accurate. Here's how to set it:

Freq in Hz = 4 * RPM / 60 [for V8]
Freq in Hz = 3 * RPM / 60 [for V6]
Freq in Hz = 2 * RPM / 60 [for I4]

For example:
[V8}
500RPM = 33Hz
1000RPM = 67Hz
2000RPM = 133Hz
3000RPM = 200Hz
4000RPM = 267Hz
5000RPM = 333Hz
600RPM = 400Hz
7000RPM = 467Hz

    Whenever you change RPMs, gently tap on the front of the tach a few times with a finger to get the pointer to settle properly.

    Set the low RPM accuracy first by feeding in a 1000 RPM signal and adjusting it dead-on using the little white plastic zero-adjust lever directly opposite the circuit board.
THIS IS IMPORTANT TO DO FIRST !!!

    Examine that weird looking metal 'fence' surrounding the meter movement. It has eight segments. When the pointer is aligned with the 500, 1000, 2000, etc. marks on the dial, the drive coil is right next to one of the segments. If you are feeding in a signal for, say, 4000 RPM and (after gently tapping) the pointer is reading a little high, bend the fence segment that lines up with the coil out a little. Tap the dial again. If the pointer is a little low, bend the segment in. Tweak each segment until the tach calibration is acceptable.

    If you don't have a 12V power supply, you can use a car battery and put a 1 amp fuse in the line to prevent disaster. Or you can pick up a couple of quad AA battery holders when you are in Radio Shack. Wire them in series to get 12V. Or pick up a 'wall wart' power supply at a thrift store for a buck. Get one that puts out DC at around 9 or 10 volts - they almost always put out a lot more than they are rated for when they aren't loaded down heavily. If you find a 'switching' wall wart or brick (you can tell these because they seem far too light for their power ratings), get one that is rated 11 to 14V. (Their regulation is much better.) You may have to load the switcher down with a turn signal bulb because they sometimes won't work correctly unless enough current is drawn. Whatever you use, carefully check it with a voltmeter before hooking up your tach. If you don't have a digital voltmeter, bite me! Sears regularly has them on sale for ten bucks; get one.

    The diagram above shows what you need and how to hook it up. The resistors can be 1/4 or 1/2 watt or whatever. The transistor is an NPN 'small signal' type - somethingsomething2222A, MPSA05, etc. The plug is a mini stereo headphone cable that fits the jack on your PC. Make sure you plug it into the speaker/line/headphone out - not the input by accident (that's ten minutes of anxiety I'll never get back...) Another 4.7K resistor in place of the 2.2K should work, but I didn't try it.

Plug the cable into the PC.
Turn on the power supply.
Run the program.
(Square wave with 45 percent duty cycle.)
You may also have to adjust your PC mixer/volume controls.
Tweak away.

IF YOU HAVE ANY DOUBTS ABOUT YOUR ABILITY TO DO THIS WITHOUT BLOWING SOMETHING UP, DON'T DO IT!

    The first two traces are a little unusual because the input signal is provided by my special tach generator. It makes a spike similar to that produced by the primary winding of an ignition coil. Some gizmos need a big spike to work - the Wedge tach doesn't. This particular waveform makes the tach stop working above 4000 RPM because the 'off' time is too short. I've got to fix that one of these days. You can see that the input signal (pin 2 of the IC) after filtering is almost gone, but he tach will work with a very tiny signal, provided it crosses the trigger point of SourceVoltage / 2.

    Inside your tach, you may notice that some of the ends have fallen off the striped capacitors. Put a little five-minute epoxy or similar on them to prevent moisture from penetrating.

    The blue electrolytic capacitor on top of the IC can go bad. If the tach seems to be inaccurate in the car and generally acts flaky but behaves on the bench, put another similar capacitor in parallel with it.

    If you use a charger to top up the battery while it's in the car without disconnecting it (an extremely bad idea, especially with fuel injection), you may find that the unfiltered 60 or 120Hz output of the charger sometimes can make the tach indicate hundreds or thousands of RPM with the ignition switch on and the engine not running. That may indicate a problem with the battery, its connections or the charger.