I know this conversion is done and dusted now but I just thought I should update the stock tachometer situation in this thread. As I said in earlier posts, the stock tacho doesn't work on the EDIS signals - I didn't end up doing the hacked relay trick because I was convinced of a better option. I ended up pulling the instrument cluster apart, removing the stock tacho PCB and making my own from scratch using the 555 timer IC chip. The 555 tacho circuit is as old as the hills; tried, tested, simple and effective. I initially had some problems after I made the circuit with it not working, or the needle being very jumpy and vibrating. It took me a long time to work out why - all the circuits for this you'll find online are just what some electronics whiz has dreamed up and what he thinks should work, most of them haven't even been tested on the bench let alone on a car
Here is the circuit diagram I initially followed:
This circuit has a number of problems:
1. The input resistor (15R) is completely wrong, with 15 ohms here the 9v zener is dissipating something like 4-5 watts! With most zeners being 1-2w rated I went through a few zeners before I realized what was happening. A 5k resistor here is much better.
2. The timing resistor and capacitor on pins 6 and 7 are are bad choice (especially for this application) This choice of values gives an output pulse of .001 seconds duration. The end result of this is a shaky/vibrating needle since there's too much off time between the output pulses going to the meter. I worked out the maximum on time I could have for each pulse at 10,000rpm was something like .012 seconds, so I changed the resistor to 9.1k ohm and the capacitor to 10 microfarad to get .010 second pulses. This is when taking a signal from 1 coil only out of 2 so I'm getting half the pulses than if I was reading off both coils.
After this it worked much better. To calibrate the circuit on the bench, I found a 240v to 16v a/c transformer and connected one line of the 16vac to the input, the other line to ground of the circuit. This triggered my circuit at 60hz mains frequency, which on a regular 4cyl car would be 3000rpm. On mine because I'm reading off of one coil, I calibrated it to 1500rpm.
I borrowed the oscilloscope I use at work and it can save screen grabs to USB so:
The blue line is the 60hz mains sine wave (stepped down to 16v) you can see as soon as the sine wave goes negative, the tacho circuit triggers and makes a positive pulse at the output for the set duration. If you count the squares, you will see it is exactly .010 seconds
The constant pulses going to the needle movement is smoothed out by the movement giving a stable reading at the needle. As the input frequency increases (higher revs) the output frequency will increase and the needle will move higher.
If you're interested, this is the signal coming from the ignition coil:
This was with the probe at 10x division so each square represents 10v. You can see the coil sits at 12v approx until the EDIS module disconnects it from ground for about 3 milliseconds. At this time, the coil fires from the secondary to the spark plug but also there is a huge 40v spike on the primary. This is the spike I was talking about it earlier posts, you get this when you turn off any inductive load (coils) e.g relays which is why they have supression diodes. The 5v zener, 1k resistor and .1uf capacitor in the tacho circuit prevents this spike from damaging the 555 IC.
Okay so the last little bit was probably not necessary to explain and maybe too technical but I thought I'd post the scope screen shots anyway, so you guys can see what is really going on when I'm talking about pulses and duration etc etc. Oscilloscopes are great fun