Erratic rpms Rotax 912/Garmin G3X

Looking at the implementation of the diode clamp circuit, I would say that it is going to be susceptible to interference. It would be much better to buy 3 new components and cut the leads down to about 5mm length and then solder them directly onto the tacho wires so that it is all in line and you can keep the twist in the wires going around the diodes and just use heat shrink to cover the circuit.

The idea behind the twisted cable is that any noise induced into the circuit will appear equally on both conductors and so it will be cancelled out at the input buffer of the Garmin. The way that the current circuit is laid out makes is more prone to having a different amount of interference couple into each of the conductors.

I'm not saying that this will fix your problem but it is an easy fix. Ideally, as I've said before, you need to visualise the signal on a scope so that you can identify the problem that you are trying to fix.

Bert Sparrow wrote:

No I don't believe I ever tried mag off.  Attached is the pic of the Rotax-specific part that consists of diodes and a quarter watt 300 ohm resistor between wires.  I bought it pre-made from an Aircraft Company I think somewhere in Virginia, and all I had to do was handle the input and output wires.  

Bert

That looks exactly like what I have the difference being that mine is more or less just shrink wrapped and bundled to the wires.   I guess the next, next step aside from getting another one of these would be to just keep it simple and toss a resistor in line like Kevin suggested if the diode could be allowing for interference.  

I have all my wires as far away from anything coming out of the stator section so I really can't do anything in my engine compartment aside from changing the other P-lead wire and going a bit further with the shielding than what the factory did.

I would include the diodes as well as the resistor. This is a description of the circuit operation.

The tacho coil on the back of the Rotax engine generates positive and negative pulses as the timing edges pass it.

The amplitude of these pulses is proportional to the engine speed. The amplitude can be quite high depending on the load presented by the tachometer or EFIS/EMS.

I suspect that the input buffer in the Garmin is looking for positive going pulses with a maximum amplitude of about 5V.

The conditioning circuit that they recommend works in the following way. The first diode blocks negative going pulses allowing only the positive pulses to appear across the 300R resistor. The zener diode is probably a 4.7V device (you would have to read the part number off your existing circuit). This effectively clamps the maximum amplitude of the pulse to just below 5V so that the input buffer isn't damaged.

I'm assuming that the conditioning circuit is definitely connected the right way around? The diodes must be in the RPM 1 line, not the RPM GND_1 line. The signal diode connects to the coil and the zener connects to the Garmin. If the orientation is incorrect then it will be presenting large negative pulses and ting positive pulses to the Garmin making it difficult for it to reliably make a measurement.

Kevin Stewart wrote:

I would include the diodes as well as the resistor. This is a description of the circuit operation.

The tacho coil on the back of the Rotax engine generates positive and negative pulses as the timing edges pass it.

The amplitude of these pulses is proportional to the engine speed. The amplitude can be quite high depending on the load presented by the tachometer or EFIS/EMS.

I suspect that the input buffer in the Garmin is looking for positive going pulses with a maximum amplitude of about 5V.

The conditioning circuit that they recommend works in the following way. The first diode blocks negative going pulses allowing only the positive pulses to appear across the 300R resistor. The zener diode is probably a 4.7V device (you would have to read the part number off your existing circuit). This effectively clamps the maximum amplitude of the pulse to just below 5V so that the input buffer isn't damaged.

I'm assuming that the conditioning circuit is definitely connected the right way around? The diodes must be in the RPM 1 line, not the RPM GND_1 line. The signal diode connects to the coil and the zener connects to the Garmin. If the orientation is incorrect then it will be presenting large negative pulses and ting positive pulses to the Garmin making it difficult for it to reliably make a measurement.

I believe then I do have the correct set up, but I wont be able to actually test it so as this has been slowly creeping in I can't help but wonder then of that zener is maybe the issue if that helps keep it below a certain threshold.   I'll double check that unit with what you have mentioned and compare it

Our issue seems to be resolved, and I thought I might share the results should you be interested.

We had ordered 011-02348-00 from the manufacturer and I don't know if our director ordered new chips or it just came with that; but we did have to assemble this unit ourselves.   Came with some resistors, diodes, zeners and after assembly and testing with a run it appears to be rock solid!  Haven't heard pilots mention otherwise!

I assume these are just going to be finicky and based on temperature during assembly damage can be caused to all of these components and there's a chance they simply degraded due to a myriad of factors which may or may not include damage during the assembly of that Garmin "filter"

I'd be curious if you had another one built (perhaps by Park Rapids avionics or Lockwood if they do that) maybe your G3X tach might work