Stihl chainsaws , M-tronics system electronics background ?

[EE-digger]

Seasoned design engineer but lately playing with a handful of chainsaws for myself and a few friends.

One of the new saws uses Stihl's M-Tronics ignition control system.  It promises to do a lot but seems to be a real PITA for a lot of repair people.  I have one but am not ready to pull it apart.

My first thoughts, and may end up being my last, is that it is a simple system and things run open loop.  For instance, there is a solenoid / valve that does the work of a needle valve on the carburetor.  In good electronics, such a valve would need to provide feedback as to its position.  If run open loop, such as by percent open vs current, I would expect a multitude of problems, especially with temperature.

Other service people have found that various pieces are sensitive and will not play nice with replacement parts.

I don't believe that sub-assemblies are on any sort of a bus or can be ID'ed by the micro, but could be wrong.

Not looking to redesign this thing but does anyone know the electronics side of this system?  Is there a single wire bus (I2C or other) or does everything indeed run "open loop"?

There is even one component that truly disturbs me and that is the ignition coil, which I believe also generates low voltage for the logic.  I've read that this one component cannot be replaced without using Stihl's dealer only diagnostics system, which talks via a test port on the saw.

This dependency, if true is very disturbing from a repair standpoint.

Any thought you can share would be appreciated !

[fzabkar]

Might be worth a look:

https://opeforum.com/threads/261-m-tronic-what-makes-it-tick.8918/

[EE-digger]

Thanks, I really need to join that forum to see images.   added - their PCB looks like crap but need to wait for approval to see the full image.  Good stuff.

Stihl has a few images, low res but ok, showing the "microswitch" which "registers the position of the choke".  Unless it's on an I2C or other single wire bus, that's all it is ... a microswitch.

Same for their valve.  A red and black wire so unless they PWM it, nothing to see here.  They do measure temperature, not sure about altitude.

I still suspect that the simplicity of the system is half of its problems.  If they are determining all they need from engine RPM including the claim of "fuel quality", then I have some land to sell them.  It's far away and pretty wet but has great potential.

[floobydust]

It looks like a basic Speed-Density system, open loop.
Knowing throttle angle, RPM, air temperature you model the engine as an air pump to get approximate air volume. Then control fuel to around stoich.
The ECU controls ignition timing, fuel mixture.

I would be looking for a pressure sensor (MAP) and the throttle angle sensor, unless they are doing something else cute here, using a carb port for some of it. Otherwise it's a primitive control system.

Kawasaki KZ-1000 motorcycle first with EFI around mid 80's used Speed-Density system. It did not have a MAP sensor though, they had a complex throttle position vs flow lookup table (it's a 4th root curve) in mask ROM. It was too difficult in the 80's with 8-bit MCU's to do a lot of real-time math.
So the EFI system would not tolerate any engine mods like a MAF system would. Once you change engine displacement (big bore kit) or exhaust headers etc. the lookup tables no longer match the factory engine and the system calibration is off.
These systems also go to shit as the engine ages, the lookup tables are no longer right if your valves are not 100% etc. the throttle plate wears as well.