I need to recreate a tachometer without altering the wire that goes to the engine spark plug. I can only modify the wiring intended to close the circuit to turn off the engine. It is necessary to take the data in a microcontroller as a sensor. The engine is a Kohler 4-stroke single cylinder, the kind normally used for emergency generators, mixers, etc.
\$\begingroup\$
\$\endgroup\$
3
-
1\$\begingroup\$ Does putting an inductive clamp on a wire count as "altering"? Maybe there is some useful information for you here: Measuring and conditioning ignition coil secondary signal \$\endgroup\$– ocrduCommented Jan 15 at 7:49
-
\$\begingroup\$ No, but is it reliable? I am not sure which caliper I can use, since I understand that there are some with a reading frequency lower than the number of revolutions that the engine can make. \$\endgroup\$– JB-GBCommented Jan 15 at 18:43
-
\$\begingroup\$ In a 4-stroke engine, the spark on any cylinder happens every two revolutions, so double the sparks-per-minute to get RPM if you sense the spark on a wire to one cylinder. \$\endgroup\$– Peter BennettCommented Jan 15 at 21:55
Add a comment
|
2 Answers
\$\begingroup\$
\$\endgroup\$
The voltage to the spark plug is thousands of volts. It is very easy to use a capacitive pickup, as simple as wrapping the wire end around the spark-plug wire -- you might even want to wrap electrical tape around the plug wire to add some extra insulation.
You'd then need to provide a filter network before the IC:
- Resistor to ground to lower the input impedance.
- Series resistor to IC.
- Capacitor in parallel with IC input to reduce frequency response, so as not to count "ringing" of the magneto coil as multiple cycles.
- Diodes back-to-back in parallel with IC input to limit voltages to ±0.6 V. Forward might be a low-voltage Zener, and back germanium, to limit voltages to VZ and ~0.3 V, respectively... or not.
Determine values with some calculation and some experimentation.
\$\begingroup\$
\$\endgroup\$
2
Assuming it's a magneto ignition then yes, the engine stop wire will connect to the points (or equivalent) and so should give one pulse per revolution. The voltage will be something like 100 V.
answered Jan 15 at 7:17
-
\$\begingroup\$ Sounds like a good solution. However I don't know how to proceed, since I do use the stop cable, I should use something like an lm2907 to collect the signals. Am I wrong? \$\endgroup\$– JB-GBCommented Jan 15 at 18:47
-
\$\begingroup\$ see cdn.velleman.eu/downloads/0/manual_k2625.pdf for an example input circuit. (R1,ZD,R2,T1,R3) do frequency sensing in software. \$\endgroup\$ Commented Jan 16 at 0:29