EAE is awesome. Basically I turned the RPM correction to 100%, as well as clt-based and AE off.

Then while on the highway in 5th I kept stopping on the throttle at different rates and rescaled the tables so the min and max had the best resolution. Then starting with stuck-to-walls, i just kept stabbing the throttle at varying rates and added/subtract correction to keep the AFRs so that they were stable with the application of the throttle so that it was almost seamless when throttle is applied.

Then I worked on teh sucked-from-walls, again in 5th gear, tuning it so that the fuel wasn't being pulled too fast when you'd lift. Sometimes if it was too aggressive, you could lift and pulled too much fuel and it would feel jerky.

Didnt take long to get it tuned and it's been great ever since. I went back to the default rpm-based curve and taking out a little correction at low rpms.


And don't even get me started about having a random misfire. Since fitting the ms3 I have one, only goes away if I swap in an ms-ii cpu...

I thought I narrowed it down to my cop harness being degraded after a few years, i was pretty appalled when I pulled it apart, the crimped contacts weren't holding on a few wires, some contacts could littrely fall out of the connectors and some of my splices were iffy at best. So I completely redid it with factory pigtails snipped of a Toyota. Even ran a bigger ground to the coils and to the MS to no avail.

My car bogs down with no AE if I step on the throttle at any revs. I'm running tpsdot which works fine.

Dimitris

interesting :idea:

for example, I can now play with my injector timing.

Right now I have each injector is driven by it's own controller. So I only inject when fuel when i need to, specifically at -350* TDC. Before, it would inject #1 and #3 and 2&4 at the same time. Now I only inject each at the proper time before the intake valve opens, so it injects less fuel and atomizes better.

I can also change the injection timing in relation to the crank by rpm and load. So at idle
if i need to inject sooner or later since, the mass airflow is slow, to better atomize, i can do this.

When i do seq. ignition, the coils will simple have more time to fully charge since they aren't firing at BDC and only once per cycle. So I should get a more complete spark and longer life from the coils.

I can't wait for the end result.

still having trouble with the coils. I'm pretty sure I have cylinders 1 3 4 firing and running and not #2, dunno why that would be...gotta mess with the output test mode and see if i can get them all to fire in test mode.

Can't wait until they have VVT or VTC (Honda K Series) control on MS3. That will be cool!

The two outputs I have which travel back through the factory wires, cylinder #1 and #2, do not spark through the MS3X board (coil test mode is fun BTW). I have a feeling it's because those outputs are piggybacked to the two mainboard outputs. Of course I had to make it so I could run batch/wasted...

I'll try a kill switch on those coming from mainboard when I'm trying to use the MS3X exclusively and see what happens...

Hrm cheers for the EAE writeup brain, that looks relatively simple, might have a bash tonight...

There's a brilliant post on MSExtra on tuning injection timing, I'll dig it out, but the gist was to have a 30 mile route of all driving conditions, set q 'best guess' for the table and adjust it up/down 10-20 deg after rinning autotune for your route, by saving the final result, (but always using the original VE MAP), you can compare the suggested maps and use the best bits of each timing table as the ideal (best being requires the least amount of fuel).

Oh and I thought your missfire was fixed after tensioning your timing belt?

sorta fixed. The major one that was causing the engine to sound like ass, like it was out of sync, and suddenly lose speed at idle, yes. There was major sloop on the TB and it would cause the spark and fuel to work at odd times compared to the actual position of the pistons. When I finally solved that it run smooth and stable and I even had to redo the trigger angle and reduce timing in higher rpms. The random one that would cause 1 or 2 coils to not fire, no.

I thought I narrowed it down to my cop harness, it was seriously getting bad. I was the first to build one using the toyota connectors and contacts, the thing degraded so bad that you could literally pull each wire out of the back of the connector with ease, some of my crimped wires also fatigued and one actually broke off. I rebuilt the entire harness and upgraded the gauge of the power and ground wire and thought that would solve it also with snipped off pigtails off a toyota.

But even still I'd have an event or two...now that I'm running full seq. ignition I had zero misfire events this morning. So there might be still something with the coil wiring, but I was able to remove even more fuel my VE table over what just seq. injection required!

If I would plug an MS-II CPU onto my mainboard, I would get no misses, I think the MS3 is just more sensitive to any faults in wiring; you're supposed to have 8 grounds on the mainboard and another 5 on the ms3x. I think I have 6 total.

Interesting on the sequential ignition, are you using all the MS3X outputs? Just wondering if they allow the coil to power up and spark stronger than using 2N2222's...

yep, fueling and spark is through the ms3x card circuits, everything else is how it was. I haven't looked at what they use, the probably wont publish schematics for the ms3 and ms3x, but it's probably the same circuit as the microsquirt module:

http://www.microsquirtmodule.com/schm9.gif

Solder and heatshrink that shit so you don't have to worry about it. As long as they don't flex much, you shouldn't have any more problems with them.

Brain, is there any way you could post a screen shot of your Adhere to and Sucked from walls constant settings?

I have other things going on, and won't be able to play with my EAE right now, but I'd love to study it more and see a good working setup in place..

umm yeah. one sec.

http://www.boostedmiata.com/gallery2...cked_walls.jpg http://www.boostedmiata.com/gallery2...1-2/adhere.jpg

http://www.boostedmiata.com/gallery2/d/13142-1/ae.jpg

Wow that's a fuckload of normal AE!

I'm on my iPhone, but mine virtually flatlines think the max is 2.1

Oh wait is MSII different?

yeah msII is percent. drives wonderful.