I understand that it's possible to select how many squirts the injectors will put out per cycle, correct? Is it possible to adjust when the injectors squirt?
Thanks.

 

Yes you can select how many squirts and no, it's not possible to adjust when they squirt....

 

I noticed they squirt whenever the middle LED flashes on my setup.

 

What are you using your middle led for?

In most 2 rotor 2nd gen setups, the middle led is going to be the coil-select signal... which isn't explicitly tied to injection events in any way.

The point though is that with the megasquirt, the way the code is written now, you can't configure when you want them to squirt... ie you can't time them to squirt when the rotor is passing the intake port (or on a piston engine when the intake valve is about to open).

For the people who are using staged injection, it doesn't even make sense to configure when the squirt happens because all the primaries squirt at the same time... and then when the secondaries kick in, all the injectors squirt at the same time.

 

Middle LED is used to trigger an ignitor. Actually, all three of my LEDs trigger one ignitor each.

 

Ahh ok, makes sense... leading only on a 3 rotor.

It's just a coincidence that it flashes at the same time the fuel squirts then... there's nothing in the code that ties that led or ignitor output to fuel squirting.

 

In the original design, Megasquirt fires the injectors in time with the tach signal. In other words, injection starts when the coil fires. This squirt event may not occur at _each_ spark event, as you can configure the system to fire 'x' times per cycle, where 'x' must be an integer divisor of cylinder count. For the typical 2-rotor installation, where '4-cylinder, 4-stroke' is selected, you could configure to squirt 1, 2, or 4 times per engine cycle, where a cycle is defined as 2 shaft revolutions. Ken, how does this squirt scheduling occur with the wheel decoder? I'd think it'd be similar, although I haven't dug through the code to check.

 

I believe it is scheduled to check whether or not it's supposed to squirt only when a trigger tooth occurs... at which point it checks to see whether alternating is set or not, and checks to see how many squirts per cycle are set up, and if it's supposed to squirt it sets the sched and inj bits for whichever injector channel is supposed to fire.

then the .1 ms section the next time it runs sees the sched/inj bits are set, and goes about actually squirting the injector...

So it's not going to inject necessarily on a spark event, but close to or just after a trigger.

 

OK, that's sounds the same as the original MS code then. I simplified things a bit with my 'with the tach signal' statement. It should be rewritten as 'up to 100us after the tach signal' to be more precise.

 

Ken,

Could the code be hacked to schedule off some tooth offset from the trigger, rather than trigger itself? If so, it might be interesting as an experiment to play with the timing...in discrete steps anyways.

 

I'm sure it could be hacked to do that... although it would make more sense to do that once we have the support for sequential injection in hardware.

 

just wondering, what EMSs can do this, "time them to squirt when the rotor is passing the intake port"
and what would be the application for it even?

thanks

 

For ports radically different then stock. Like peripheral.

 

It could be used for lean burn and/or stratified charge fuel injection--both of which could help the poor fuel economy of the rotary.

Wish I had unlimited funds to play around with, out of my budget right now.

 

ok, well it sounds to me like there aren't too many EMS out there that can do this?
so not really a major downfall of the megasquirt...

 

A number of aftermarket ems outfits offer sequential injection. I'm not familiar with them in detail, but sequential alone doesn't necessarily mean you can alter the _timing_ of the injection event, but just that the squirt events occur in a sequence rather than all simultaneously. I suspect the benefits of sequential, and also injection timing are primarily found in the area of idle quality and idle/low speed emissions. Those benefits fade quickly at high power settings when the injectors are open 60-80% of the time anyways...

It's quite possible the MS hardware guys will have equipment to do this sort of thing in the near future (6month-1yr??), but if this is really a hard requirement, then I'd say you are outside the scope of a $200 diy efi system anyways.

 

in a sequencial injection,on production cars, the injector opens once per 720 degrees of crank rotation. batch firing, the injector fires twice per 720 degrees of crank rotation. on the sequential injection the duration of injector opening is twice as long as batch firing. supposedly its more effiecient in fuel economy. fuel doesnt accumulate on the closed intake vavles, like they do on batch firing cars.

 

Many consider fuel sitting on the intake valve a positive thing, and actually time the injection for this purpose. The hot intake valve helps to vapourize the fuel prior to entering the combustion chamber, rather than entering in droplet form.

 

What I have learned over the years on sequential

1. On OEM motors its done for emissions at idle and light cruise. Injecting onto a closed valve vapourises the fuel to allow a leaner burn.

2. Vapourised fuel is bad for power as you lose air volume over atomised fuel

3. In terms out outright power then sequential may gain you a few % say 5% at best

4. For a radically ported engine sequential will help fuel economy, as you aren't stuffing fuel straight down the exhaust. Maybe 10-15% in extreme cases.

5. One advantage of sequential is that you gain per cylinder trim. So if one cylinder/rotor runs leaner than the other then you can tweak it up.

Its this last one that is the main driver for a performance application.

BUT tuning a sequential setup is much harder than batch, so unless you are the sort of person that knows why they need a motec, then don't worry about it.

Bill

 

lots of good info, all the points mentioned make sense.

it would be interesting if this were developed for the MS though...i wonder how hard it would be?

 

I believe things like that are on the list of things to do for ms2 with CAN bus once the router board is done.... this is in the 6 months to 1 year timeframe though last I heard.... and will probably cost significantly more than what is available now....

In my opinion it's not worth waiting for.

 

A 4 injector drive system would be very nice for FD applications so you could stage primary and secondary and have trim to deal with the manifold flow differences front to rear. Also for a bp or pp NA rotary some sort of timing would be nice.

However both of these are slightly rotary specfic rather than generic niceties on MS.

 

My point is that while it'd be nice, it probably isn't something that's worth waiting for... and most of the installs I've seen do fine the way they are...

Like I said though, adding this sort of thing is also going to up the cost ... possibly significantly.

 

I've seen enough evidence to say that, for big boost FD installs it would be worthwhile. For us mere mortals, I agree fully with you

 

i originally didn't like the batch fire concept, but i thought about it for a little bit. on a piston motor, the intake valve it closed for probably more than 400 degrees of crank rotation. which lets the fuel just sit on the valve at low revs. i didn't really like the idea of how on each bank, some vavles will be open, some will be closed and others somewhere inbetween.
but on a rotary, the intake ports aren't closed for nearly as much crank rotation(less than 70 degrees or so?), so the fuel will have much less time to puddle.

just thinking, that's all. i have a MS built but no harness yet. i've been messing with it on my laptop just to fiddle. i don't have a clue what i'm doing.