Tachometer flutter

[Hasse]

Hi,

 

I am doing the first tests with a new Rotax 912 engine in my Savannah S and the engine runs very smooth at all speeds. However, the tachometer behaves very awkvard and the instrument needle flutters markedly at a specific range (3100-4200 rpm). Very strange because the flutter is not correlated to any irregularities of the engine rpm. What may be the cause of this anomaly? Do I have a defect tachometer or may there be some kind of disturbances that can interfere at a specific range of speed? Below 3100 rpm and over 4200 rpm up to max (5400 rpm) the tachometer behaves normally and the needle is as steady as is the engine. I have separate tachometer and hourmeter instruments and not the standard combined Tachometer/Hourmeter.

 

Any ideas from you guys or anyone who have had the same experience?

 

I hope to fly soon

 

Kind regards

 

Hans Marchner

 

 

[cscotthendry]

Hans:

 

I had the same problem. Without going into it too deeply, the problem is that the tachometer does not load the tach coil of the engine enough. The coil produces too high a voltage for the meter and overloads the input. The fix is very simple. Wire a small resistor across the sensor leads going into the gauge. I used a 220 ohm resistor, but I suspect just about any value will do the trick.

 

 

[Hasse]

Great input Scott,

 

I will sure try this asap. Not that I understand why this flutter only appears at a certain range of rpm. Still I will try your medication and report the outcome.

 

Thanks

 

Hans

 

 

[cscotthendry]

Hans:

 

Do you have the Aviasport tacho?

 

The reason it happens at a certain RPM range has to do with resonant frequencies. The coil on the Rotax engine is an inductor and the input to the tacho is an R/C *circuit. At the resonant frequency of the coil coupled with the R/C input of the tacho, the voltage in the circuit peaks and overloads....the internal circuit which is .....if you have the Aviasport tacho, or some version of it (the ones with the coloured bands specifically for Rotax engines) the internal circuitry is a microcontroller that counts pulses from the tach coil and passes a variable voltage to the meter movement in the gauge. When the voltage peaks and overloads the sample input pin, the micro starts to clock internally from that pin rather than its internal clock.

 

All of the Aviasport Rotax gauges have the same internals. The only difference between them is that some measure the resistance of a thermocouple, one measures the voltage and the tacho counts pulses.

 

BTW, most of the gauges also have an output that produces around 4.7V when the gauge is in the yellow or red zone. When it's in the yellow zone, the ouput pulses and when it's in the red zone it stays on. You can connect an LED to this output as a warning indicator, but make sure you get the polarity right. Connect the anode of the LED to this output and the cathode to ground through a 330 ohm current limiting rsistor.

 

I don't know why Aviasport haven't fixed this. It is a known issue for some time now, obscure, but known.

 

*Resistor / Capacitor, not "Radio Controlled"

 

 

[Hasse]

Hi Scott,

 

You're a mind reader! Yes, I have an Aviasport tacho and it looks like you described it. The idea of resonance at a certain frequence also makes sense.

 

It's great talking to you Scott!

 

Thanks

 

Hans

 

 

[Guernsey]

This is what I really like about this Forum, members helping others. .

 

Alan.

 

 

[fly_tornado]

just typed into youtube "bouncing tachometer"

 

 

 

[microman]

I have a slightly similar problem with my Rotax tacho (the one with the coloured bands) - the needle is steady, but it wont read over 5000 rpm - even at full throttle which I checked with a go-kart digital tacho is around 5500 rpm. Would Scott's fix with a resistor cure this problem too?

 

Graeme

 

 

[Hasse]

Hi Graeme,

 

I don't know, but why not have a go? It's not a big job to connect a small resistor in series with the yellow/blue thread and see what happens. Not much that can go wrong.

 

Hans

 

 

[pmccarthy]

In series with one lead? I thought the advice was to wire it in parallel across the instrument.

 

 

[Hasse]

pmccarthy:

 

Well according to the youtube instruction "bouncing tachometer" provided by fly_tornado the resistor is just cut into the lead thread in series.

 

Hans

 

 

[Guest Andys@coffs]

a resistor in series( given the suggested 220 ohms) then given the input of the microprocessor is likely very high impedance will likely change very little... Wired in parallel it will dramatically affect the resonance issue. Try in series first, but be prepare to change it to what the original poster specified across the inputs...if there is only a single input then one end of the resistor on that input the other end attached to earth.

 

If however the gauge input isn't high impedance due to other circuitry that the designers added outside the microprocessor then in series will act as a voltage divider reducing the voltage that the circuit sees. If as the original poster suggests that the issue is due to overdriving the input then such an approach is valid, but may lead to underdriving at other rpm ranges....

 

Andy, an old airforce radio tech.

 

 

[Hasse]

W

 

 

[Hasse]

Thanks Andy,

 

Ok, now I've got several options. I will try them one by one and let you know.

 

Hans

 

 

[cscotthendry]

Thanks Andy,Ok, now I've got several options. I will try them one by one and let you know.

Hans

Hans:

 

The easiest option is to get in behind the gauge and just hand wire the resistor across the input and ground connections. Just wrap the leads around the pins and then go for a short flight. If it doesn't fix your problem, remove the resistor, no harm done. Just make sure the leads are secure. If they're not secure enough, they will make intermittent contact which will look like the problem is not fixed. Try to get at least two or three turns of the leads around the pins of the gauge.

 

Not too sure whether connecting it in series will do the job. Parallel definitely will though.

 

 

[Hasse]

Thanks Andy,

 

I'll do the parallell first. No problems theoretically. In practise however my instrument panel is not easy to deal with. There are a lot of threads that pop up when you open it. Reminds me of the Pandora's Box.