Its for my 20b 750 primary 1680 secondary.

The car has a wandering idle, stalled when the fan kicked in. AFR all over the place, etc...

please check out may be there is something I'm missing.

Thanks

 

There were some problems when I was loading it (I haven't used MS3, so I didn't have an existing project set up for it in TS), so some of the settings that I'm looking at may be messed up, but some things that stood out are:

Engine Constants - squirts/cycle should be 2 for a 20b, running sequential injection. As I understand it, for the rotary setting, a cycle is 2 revolutions, and each rotor intakes air once per rev, so you need 2 squirts/cycle so that each intake event gets a squirt. This would definitely affect idle/low-load a lot and cause a bouncing AFR, since one intake/firing event would be rich since it just got a squirt meant for two, and the next would be lean, alternating like that for each rotor.

General Settings - Unless you have a lot of noise on the TPS & MAP, I'd use a higher filtering value (IIRC a higher value means less filtering, so any changes in TPS & MAP will be noticed more quickly and can be responded to without a stumble). Also, the overrun fuel cut parameters seem a bit loose, especially the TPS lower than. If you can, I'd keep that at 1-2% (as long as that's above any noise on the TPS channel when you're off the gas), so that you don't get into a low-load abrupt on-off type situation around town.

Just to start, those (especially the first one) seem like they could be causing your problems

 

I'll check that out thanks

 

squirts per cycle doesn't really do anything in seq mode on an ms3. We internally redo the calcs to get back to the base req_fuel.

Ken

 

Thanks Ken, I'll leave that part alone.

 

Oh, it doesn't? Ok, so putting it in "rotary" engine type automatically implies that 2 injection events/channel/cycle (2 revs) is necessary, or does it just define a cycle as 1 revolution? Just out of curiousity?

 

It automatically implies that fuel will squirt from each injector when it needs to. I'd have to check the code to be sure, but I believe 1 cycle = 1 rev in rotary mode. So really it'll be 1 squirt per rev per injector in sequential mode, timed to the timing you enter in the table.

Ken

 

I just looked at my injector timing table and every value is 360. It that a problem?

 

You'll probably want to tune that. The generally accepted method is to tune the VE the way it is, then adjust the timing to the point that makes the engine run the richest, then lean it out in the VE table again.

That will give you the best efficiency. You can also make adjustments for power, etc.