Ok, so ive been thinking about this....

The megasquirt II has an ignition input and the ability to modify the output pulse, but it only has one coil output. Why shouldnt it be possible to modify the code slightly to allow for 4 ignition outputs using some of the spare unused outputs? A transistor could be used to sink enough current to charge a coil from the output signals.

The way I see it, the leading needs 3 output pulses, 120, 240, and 360 degrees. (+_ adjustment factor for rpm/load advance, which the ms already has a table for)

the trailing could be fired wasted spark with all 3 plugs on one channel, and a split table could be added to the code to control this channel to fire at all three leading events +- the referenced split value.

am I way off base on how the ms is set up, or is this something I should think about trying? Im pretty good with programming and electronic type stuff, and i really dont want to spend 2k on an ecu for my 20b project.

any advice you cna give me is appreciated.

thanks
Pat

 

you don't need to use ms2 for this, as the MS1 could give you leading only RIGHT NOW... and I could easily add trailing to it as well.

As for running trailing wasted spark, you can't. In fact you can't run anything wasted spark on a 20b. You'll have trailing plugs firing at the wrong time for different rotors and will be causing yourself all kinds of nasty problems.

As for the MS2, I'm not sure it has enough spare outputs to do everything you'd want to do on a 20b... I think it's only got maybe 2 or 3 spare pins left.. you could probably get it to the point where leading-only would work, but I doubt you'd have enough pins to use a stock CAS, and still have 6 ignition outputs. Maybe if you cut teeth out of a CAS and used it that way...

 

so does the original ms have more spare outputs than the ms2? and enough for 6 ign outs and still a couple left over for boost control and crap like that?

how many inputs does the cas require? i thought just 2...?

also cas is not necessarily the way to go, im willing to screw around with other types of inputs if the ms cant use the cas, or for that matter to modify the cas so that it can...

correct me if im wrong, but 20b has 3 events per eshaft revolution. so thats basically 120, 240, 360 base for leading and assuming 20 degrees of advance maximum, and 20 degree split, that has trailing firing between 100-140 and etc depending on load and rpm, so why is that a problem for wasted spark?

thanks
pat

ps thanks for the quick reply

 

the original MS chip has more available... on the MS2, 2 are used for CAN bus, and several more are used for idle speed stepper motor control, etc....

The CAS only requires 2 inputs, that's correct.

First, I wouldn't assume 20 degrees maximum, on cruise and off boost, you'll see in the 30's or even 40's for advance. In some cases you might fire up to 5 atdc too if the 20b is anything like the 13b.

Second, depending on advance, etc... you could end up firing one of your wasted trailing sparks in the next rotor's unburt fuel mixture.

IT's the same as with the 13B. They don't run trailing wasted because in some situations, you'll fire a trailing plug too late (on the wrong face of the other rotor) causing the air/fuel mixture to ignite before it's supposed to.

All of this disreguards the fact that your trailing spark won't even be accurate... it'll be accurate on the one rotor that it was supposed to be firing on, but you'll have spark firing on two other rotors at a time that you didn't set.

For the 20B, your best bet is to run leading only on the ms1 chip with msns-extra, and if you really want trailing, I or someone could implement it using the other three spark outputs on the MS1.

I'm not sure which features you give up when you're using all those pins as spark outputs though, but I'm sure you have to give up some things.

 

trailing is pretty important for turbo rotaries, it causes a hotter exhaust with more expansion, so you get a much better spool time.

as long as full ignition control is possible with the ms1, i will go ahead and buy the stuff and put one together, and see if i can write the code myself. im sure i will need at least some help, but id like to figure out as much of it as i can myself.

if youre absolutely sure its possible, let me know and i'll get the pcb and parts ordered tomorrow.

thanks
pat

 

Of course it's possible, like I said, RIGHT NOW you can run leading only on a 20b. I think one other person as actually had it working.

In the MS, you'd just set it up as a 4.0L V6. There are three more spark outputs that can be set up, but right now there is no Coil-on-plug rotary support for 13B or 20B, and that's what you'd need.

I don't know anything about the ignitors/coils that come with a 20b either... perhaps you could tell me more about these before I tell you exactly what you'd need?

 

I wont be using those, so I can use pretty much any setup I want. I was planning to use a transistor as an ignitor and just run bosch coils.

 

I'd suggest using vb921's although they're a bit hard to find... I think you can get them from diyautotune.

I think other than that you could get it working with a little time and patience.

 

What muy says re the 921s is massively correct - there's a lot more going on with coil triggering than meets the eye. An isolated power transistor won't last very long even if it doesn't immediately fail and/or start on fire.

-Mike

If you cannot find the datasheet for the 921, let me know and I'll make a URL containing it available. I'll have to create the link, that's why I'm not offering now.

 

i found it. thanks, looks like just what i need.

so if i use 6 outputs for coils, how many outputs will that still leave me for other stuff?
things like staged injection, turbo timing, and boost control would be nice to have if possible.

edit: or better yet, is there a link i can look at that describes all the I/O space available on the ms?

edit again: so it looks like if i get the pcb v3 then i have the proto area to add all my other crap, is this correct?

thanks
pat

 

turbo timing isn't a feature that msns-extra supports. IT's just not there.

I'm not sure what pins on the cpu the 6 ignition outputs use, but I'll try to find out tonight, I know that three of them are on the LED's, but I'm not sure about the other three.. I'll try to find out if boost control is one of the things that got nuked by adding the 5th and 6th spark outputs.

http://megasquirt.sourceforge.net/extra

Some of the documentation is a little out of date, so I will probably have to look at code to give you an answer.

the proto area isn't that big, so you can fit some things there, but for things like boost control it is better to grab error*'s msns-extra daughtercard.

 

oh well, i already have the parts for my seperate diy turbo timer anyway. boost control code is uder-simple, and only requires 1 output. sequential injection is the big one that id rather not get rid of.

 

sequential injection isn't going to be a choice... the megasquirt only has 2 injector channels... for a rotary, that means one channel for primaries and one for secondaries.

The *code* for boost control is already there, it's just a matter of whether it was sharing a CPU pin with the 5th or 6th ignition outputs. To determine that I'll have to look at code.

 

looks to me like js2 is used for boost control, and it is not labeled as one of the ignition outputs on the msns-e page

 

yeah, but that page is a bit out of date, it hasn't been updated since the 5th/6th ignition outputs were added, which is why I said I'm going to have to look at the actual code to find out which features are sacrificed when you turn on the extra spark outputs.

 

it had a through f ....

anyway, unless im missing anything I think i'll go ahead and order the v3 kit this afternoon... is diyautotune the best place or should i look elsewhere?

pat

 

yeah I guess they HAVE updated it fairly recently, I typically just look at the code to find out what uses what pin because until recently they've not been updating the docs that often. The 5th and 6th ignition outputs are on pin 10 on the cpu, and JS10 (which disables knock detection)

I believe diyautotune has what you need for the base box, and for ignition, you're going to need to buy 5 more vb921 "modkits" from them.

I guess you can do boost control in the proto area, and cut 2 teeth out of the CAS for wheel decoding (unless you can fit a second vr sensor conditioner AND the boost control circuit in the proto area, in which case you can use a stock CAS).

And then the code needs to be written for 6 COP outputs for rotary leading and trailing. That shouldn't be TOO hard but it's going to take quite a few code mods. Essentially you'd be setting it up as v6 wasted spark, with each of the trailings firing after the three leadings

 

well, i never could resist a really hard project...

can you point me in a good direction for some reading up on assembly? ive got a good background in basic and c, but ive never written anything for assembly before..

thanks
pat

 

Have you ever done any embedded C? If you've at least done some of that, you shouldn't have too much trouble.

The document I used to learn back in June was this one:

http://www.hc08.cz/attachments/gp32.pdf

And I used this as a reference for the instructions until I had the ones I used most memorized:

http://www.megasquirt.info/v22manual/mapp.htm#set

Ken

 

cool, i'll check it out.

bought the msv3 kit and 5 921 thingies. the resistors i can buy at radioshack i figure all the wiring and stuff i'll do myself, so no point buying their harness, right?

pat

 

you can do it yourself, but I recommend buying the wires anyway from someplace like rs-autosport or diyautotune. It does save a lot of time.

Ken

 

yeah but does their bundle have enough wires for all the extra crap im gonna be doing, or am i going to have to add 6 or 7 of my own wires anyway?

 

you can email them. I'm sure they'd send you some extra wire if needed.

 

ok ive been readin some more and now I have a new question: would error*s dauhterboard give me the space i need for all my extra crap? (6 ignition outs and 2 vr sensors, plus boost control and a little room for future additions is what i want)

 

the 6 ignition outs have nothing to do with error*'s daughtercard. You'd use only one of his vr sensor conditioners, the boost control circuit, and maybe some of the spare outputs if there are any left on the megasquirt itself.