I would like to present my EFI retrofit project based on rusEFI.
Engine: S-31 type (used in FSO/FSM Syrena model 104, 105, 105L, R20 and BOSTO) -- pretty similar to Wartburg 353 engine
Type: Water-cooled, three-cylinder 2-stroke, piston-controlled intake
Mixture control: float-type carburetor
Fuel: premix 1:30
Capacity: 842 ccm
Rated output: 40 PS at 4300 rpm
Max. torque: 80 Nn at 2750 rpm
Parts used:
- fuel pump: universal 12V 4bar
- fuel filter: with integrated 3bar pressure regulator
- crank sensor: 5WK9 607 (opel astra, vectra)
- CLT sensor: NTC 10k M8 termistor
- IAT sensor: NTC 10k termistor
- MAP sensor: 16212460 (np. Opel Corsa)
- full Triumph 955i ITB body with fuel rail, injectors, IAC stepper motor and TPS
- some laser-cut parts designed by my
- STM32F4 Discovery kit mounted on prorotype board (only ADC input buffer, stepper driver -- mostly based on Frankenso circuit and 6x irgb14c40l IGTB transistors for injectors/coils driver)
Images can be enlarged by opening in new browser tab
My FSM Syrena 105L, 1976 - this car is not well known in the world, mainly its distribution took place in Poland during PRL times
engine bay - currently I have test engine installed in self-made test bench, see next photos
intake manifold adapter - plywood prototype and final metal one
crank sensor adapter
36/1 trigger wheel
Test bench
Engine just before first tests
First successful start attempts videos (as it turned out later with a damaged MAP sensor):
Latest test (with working MAP sensor):
Map sensor is connected only to middle cylinder throttle and has optimized measuring window (between inlet window opening start and wide open) set in rusefi firmware. As you can see in the film, the vacuum pressure varies only between 73 and 100 kPa (general 'problem' of 2 stroke engines). This can be a problem because I want to try with Speed Density first -- this can be difficult to tune well. In case of failure I will try alpha-N. I must also deal with interferences from ignition circuit (mostly from high voltage part) because sometimes there are false signals collected from crank sensor input. I think Coil-On-Plug coils type can help with that. At the moment engine starts, works witch SD but requires further tuning.