Erratic rpms Rotax 912/Garmin G3X

Bert Sparrow wrote:

Thanks Jacques, I actually went on amazon and bought 20 of these ferrite things/snap ons, of various sizes for only $8.  They should come today so I look forward to connecting them on.  I was actually thinking of NOT trying a wraparound because that's really wrapping them around very tightly, which I don't think would be good.  I do have the tach wires shielded, and grounded on just one side as per the Garmin directions.  I will keep you all posted.

vr

Bert

Hey I know this is a little bit of an older post, so I assume this issue has been corrected?   

I am having a similar issue but it does seem to favour the left "mag" as the right "mag" causes fluctuations when it's on.  I am trying to narrow it down but it's definitely being caused by an issue before my ignition module so I can cancel out the chance of it being coil or spark plugs/leads

Hi Paul, unfortunately it never did.  I did try two big ferrite surpressors, one near the engine and one further back, plus I added more complete shielding to the generator wires, in fact all the way inside the engine compartment the two gen cables are shielded, and it made zero difference.  Where I am now is here:

I decided a $20 dollar workaround was to by a Tint Tach, wrap it around #1 sparkplug, and run the little readout gauge to underneath my instrument panel where I stuck it with 2-way tape.  So now I get a 100% accurate reading of all rpms at all rpm ranges.  

What I now find with my Garmin in comparison is from idle (1700-1800) up to say 4600 rpms or so, BOTH read identical rpms numbers.  But when I go higher rpms than 4600, the Garmin rpm readout becomes erratic, with constantly shifting rpms, like 5000, 6500, then 6100, then 5200, etc, changing every second.  Of course my Garmin flashes the red X for rpms for anything above 5800, so it's annoying.

But at least my cheapo $20 Tiny tach gives me the correct number. It's part of my pilot visual scan now, lol.  I spent so much for all that work right and it doesn't but oh well.  Maybe one of these days I will wire in a 3rd resistor and see if that helps.  

Bert

Oh, and since I has 20 Ferrite suppressors in the box from Amazon, instead of the two in his pic, I tried putting extra one's on there - like on both sides of the molex connector.  And one going into the firewall.  But they never made a difference.  And I tried both large and small size ones.

Bert

If you can visualise the tacho signal on a scope that would allow you to identify the problem and remedy it. Do you know anyone with a portable scope you could borrow? If not, you might be able to use the MIC input on a laptop and capture the waveform in Audacity (free). You would just need to AC couple the tacho signal with a 1uF capacitor. See the attached circuit, that should work. There's a good article here https://www.build-electronic-circuits.com/sound-card-oscilloscope/

This video has a link in the description to a more appropriate scope app. https://www.youtube.com/watch?v=b9XIjElIgy0

You can see if your problem is due to interference, poor amplitude or bad wave shape. You could try measuring before and after your existing diode protection circuit. You will need to adjust R2 in the attached schematic to avoid clipping of the waveform. Although it's uncalibrated, it will give you a sense of the signal integrity. You could probably calibrate the position of R2 if you use a 100Hz tone from the sound card and measure it with a multimeter on ACV and note the position of R2 with the measured value in Audacity.

Let us know what you find if you get a chance to try it out.

Just had another thought. You could try wrapping a few turns around a spark plug lead and feeding that into one of the inputs on your diode protection circuit before going into the Garmin tacho inputs. Just leave the other input to the diode protection circuit free, there should be sufficient capacitance to complete the circuit.