I have two schematics for this project. This one is the more complicated but more robust version, and the one on the next page is a simplification by Bill Yacey.

The circuit works by rectifying and combining the two primary magneto lines. Typically these are both routed up to the dash to the keyswitch where they're used to stop the engine. Since together they'll produce two pulses with each revolution of the engine, they can directly drive an automotive or inboard tachometer setup for a 4 cylinder, 4 stroke engine. This eliminates a complicated and sensitive pulse multiplication stage. It also has a further advantage as the typical inboard tachometer is only scaled to 6k rpms, which is more useful for older outboards.
The three primary design goals of the black box were construction simplicity, insensitivity to noise and battery voltage, and reasonably low power consumption.

The first half of the circuit is primarily a filter and protection for the second half. The signal coming from the magneto is a negative rapidly decaying ring, starting at well over 100v and ringing at about 15KHz. D10 and D11 rectify the signal from the magneto while R1 and R25 prevent loading the magneto down and weakening spark. C1 filters out the high frequency component and lengthens the pulse for better compatibility with a wide range of tachometers. R5 pulls the line up to battery voltage (remember the signals from the magneto are negative). D9 protects against pin 6 going too high in the case of an arc-over or some such anomaly, or below 0v.

I have D11 and D10 as 1N4006, but really that's overkill. Really anything with about a 300V PIV would be just fine, and I think even 200 PIV would work. D9 is a zener marked 7-30v: anything between those two would be just fine, and 1/4w or so. Say 1N5242 (12v). All resistors should be 1/4w for durability. C1 can be polyester, ceramic or any other superior capacitor. It should not be a tantalum bypass cap. Working voltage should be at least 20v.

The value for the input resistors is a trade off. Higher resistances load the magneto less, but make the circuit more susceptible to noise. Experimentally, I determined that loading the magneto by 10K ohms had no discernible effect, and then I doubled that for the circuit.

This is what the output of the filter looks like at about 1000 rpm. The roof of the signal is about 5.5-7v depending on battery voltage. The wave shape varies a fair bit if the arc fails or if the gap is shorted.

The second half of the circuit is simply a schmitt trigger to reduce the chances of the tach seeing a glitch. Op amp U3b compares the signal from the input filter to a voltage reference provided by D12, and modified by feedback via U3a. The resistors' values have been carefully chosen to be easily found and also to track varying battery voltages sanely. But essentially, R27 and R30 set the centre of the trigger voltage while R29 sets the envelope.
The schmitt trigger is also set up to invert the signal, which presents the tachometer with a signal which looks more like a typical 4-stroke battery ignition signal.

LM358 was chosen because it's durable and easily found. It is a dual version of the quad op amp LM324.

The current through D12 is on the order of 3 or 4 ma, so any type of 3v zener will do fine here.

Here's the output going to the tach.

Bill Yacey simplified the interface box by chopping off the Schmitt trigger and feeding the signal raw into his tachometer. The tachometer he used employed the world's most commonly used tachometer driver IC, so I strongly recommend trying out his circuit first and seeing if it works with your tach. Just leave some extra room on your circuit board. If you observe glitches, add the Schmitt trigger and see if that cures it.

Also, Bill has a .1uF capacitor in addition to the electrolytic one for power supply bypassing. Mine ought to have one too.

Nothing is particularly important about how you construct this circuit. I used a perf board and point to point wiring as you can see in the picture. Since this goes in a marine environment, I strongly suggest you conformal coat the board to prevent corrosion. My poor man's conformal coating is simply a few coats of clear acrylic spray paint.
Any project box will do for this, but again because this belongs on a boat, plastic is best.

I've tried to account for various designs of ignition and tachometer, but be warned that I've only tried this on a few engines, with a few vintages of ignition coils. I cannot guarantee that it'll work in all circumstances. The tach I used was a Faria and the tach Bill used was one from Princess Auto.

Feel free to email me if you have any questions or comments.
paul.moir@gmail.com

This is just a simple revision which eliminates the zener D9 and replaces it with D13 and D14. I have them listed as 1N4006 but anything you have on hand will do, eg 1N914. You can apply the same change to Bill's simplification. Also added is the recommended 0.1uF bypass capacitor. This can be the same type as C1, or a even a tantalum. Minimum WV ought to be 25v for it.

If you have any questions or comments feel free to email me at paul.moir@gmail.com.