atchi, playing with a megasquirt'd 3 banger puts you in a rather lonely place as the msforum community never really had any 3 cylinder guys contribute to a conversation.
there are some pitfalls associated with development that i had to address while i have been playing with my turbo3.
first, batch fire with injectors large enough to support 150 hp runs like ass. i would suggest that you jump right to setting up for sequential injection to allow for tuning, especially to get a suitable idle that doesn't wash the cylinder walls down with raw gas. to run an ms in batch fire for 3 cylinders you have to spoof the set up as a 6 cylinder and that immediately affects ignition timing. you get tied to a lost fire spark.
i have sort of built upon turbohull's project work and after he threw a rod and dropped out of the turbo3 development in favor of newer engines, and took a good look at how the megasquirt controller deals with timing.
by far, my biggest concern was the high boost, high load, high rpm demise of turbohull's engine. i tested using a v/r sensor reading a 36-1 trigger wheel and i found that the sensor acted like a generator so that the higher the rpm, the higher the voltage generated. the sensor not only put out a square wave, it also transposed the square wave on a voltage that rose to around 70 volts at 6000 rpm. the typical ms build uses a number of components in circuits that "condition" the timing signal before sending it on to the processor. i found that in exactly the place you don't want a timing fart, the processor would succumb to the high voltage and cause a reset of the ignition timing and fire a plug out of sequence with the pistons' tdc. i believe this situation is what threw the rod in turbohull's engine.
if you care to see my development of hall effect devices on the crank and cam suitable for sequential timing and direct 3 channel ignition you can find it in my blue monster vert project thread. if you have more detailed questions just ask. i use some inexpensive chinese signal processing boards that output a digital/ ttl 4.76/0 square wave that can bypass the ms board's signal processing and connect directly to the processor pins on the daughter board. i tested my set up on the bench to 19,200 rpm with zero glitches.
i also used a quad driver board sold by jperf populated with just 3 channels to act as off board injection and ignition drivers. removing all the big components from the case's heat sink allows the controller to operate at way lower temps.
for my direct 3 channel ignition i used the oem hitachi ignitors triggering ford coil on plug units.
so, i guess what i'm saying is that i think that you should look into the ms extra book and go big with your controls instead of having to go back to make this or that piecemeal modifications as you go.
Login
Register
Gallery
FAQ
Search
