MS2/Extra 2.0 beta12 (20071006)

[jsmcortina]

Fixes/changes in this release include:

1) Code change for Neon/420A to use one edge of tooth for spark calcs. Some good feedback, some unsure. More testing required.
2) Actually implemented some of the Miata 99-00 patch (I'd only half done it last time) needs in-car verification.
3) Added COP mode for Neon/420A. The cam input must be the same polarity as the main tach. (i.e. same rules as 4G63)

The beta is here:

http://www.msextra.com/ms2extra/files/b ... 071006.zip

The other pending changes that Ken mentioned in the beta 10 announcement are still on the to-do list.

James

[Tjabo]

Just curious James, regarding the COP mode for Neon/420a, is this a non-wasted spark application then? Where can I find the information you menitoned about the cam sensor rules for the 4G63 application?

By the way, is this Beta ready for a hack like me to jump into the MS2/E testing on a Neon?

Thanks for the ongoing hard work! ! ! ! ! ! !

Thad

[UnaClocker]

They really need more people to try it out. I can only start my car when it wants to be started right now. There is some very strange glitch at low RPM's (like cranking speed) where the MS doesn't seem to sync to the wheel. I couldn't get it to start on the MS2 yesterday morning. Swapped to my MS1 and it fired right up. Swapped back and nothing. Got home from work and managed to get it to fire up after reflashing my MSQ. It fired up about 20 times no problem. Got up this morning and it's doing it again, won't fire up. Tried about 20 times. At best, I got the tach to twitch, which is where it ALMOST sounded like it wanted to fire. One step forward, one step back.. Drat.. If I had a friend handy, I'd try tow starting it, I imagine it'd fire just fine if I could crank it faster than the 196 or so RPM it normally cranks at.

[jsmcortina]

Yes, the code is COP, not 4 coil wasted spark because you could do that already.

Unaclocker, have you tried the latest beta? I made a small change that _may_ help or at least highlight the issue in Megatune more. There was a condition where the code should have been losing sync but it wasn't - it might not help you start, but it should make it more obvious what is happening.

James

[UnaClocker]

Nope, hadn't updated, didn't think anything had changed in that respect. I'll give it a shot. May take me until tomorrow morning (or 5 mornings down the road, if it wants to be a PITA) before I'm really sure it's fixed. Unless it immediately shows itself.. heh.. I'll try it out in a bit.

[UnaClocker]

Problem is still there.
I was able to get it to start, jumped it off my wife's car. With her car running it cranks over fast enough to start. I should have datalogged, because I could see in MT the RPM rapidly going between cranking speed and 0 as it was trying to start.
And new for Beta 12.. The tach out on JS10 no longer works. Oops!

[jsmcortina]

And new for Beta 12.. The tach out on JS10 no longer works. Oops!

It is configured as the cam trigger input. I'll make it so this is only the case when running COP.

James

[muythaibxr]

We'll be doing another beta tomorrow. I wanted to do it tonight but didn't get any time to work on everything I wanted to go in.

Will have time for that tomorrow evening, so hopefully we'll get one out then. One of the things going in will be PID closed-loop idle for steppers if I have enough time.

Otherwise I just simplified the semi-sequential settings, and James fixed a few problems.

Getting closer to a solid release though!

Ken

[UnaClocker]

Did you get a chance to check out the odd stepper behavior on power-up/cranking? Yeah, the closed loop control will take care of the 3,000rpm idle I get once in awhile, but it doesn't help the times that the stepper doesn't open at all and the car won't fully start.

[muythaibxr]

I didn't check for that behavior.

Nobody else has reported that behavior though, or the 3000 rpm problem, so I'm wondering if there is something different about your stepper motor compared to the other people who are using it.

Ken

[karlo]

Any chance that you can take a look at my twin trigger for odd fire patch before the final release of 2.0? I detailed the change in http://www.msextra.com/viewtopic.php?t= ... c&start=20. It was required in order to make the engine run properly with full fuel and spark control.

To be honest, I'm still not convinced about the mapsampling part of the patch. It will inevitably hurt throttle response, but without it the mapdot signal went haywire.

[muythaibxr]

James said he didn't think that the first part of the patch was necessary. I'll take a look at the map sampling part at some point when I get a chance.

Ken

[karlo]

The issue is subtle, but goes like this:

* When trigger 1 is detected, tooth_no is set to 2 and execution is ended.
* When trigger 2 is detected, tooth_no is not incremented, and sync is declared. Execution continues, and tooth_no is rolled over to 0.

This results in trigger 2 becoming tooth 0 while trigger 1 becomes 1. (And likewise when the triggers are encountered in the opposite order).

For odd fire engines this is counterintuitive, since documentation says to input the small angle between trigger 1 and 2. Instead, you will now have to enter the large angle between trigger 2 and 1.

The code works the way it is now, it's just a little more confusing.

EDIT: Another surprise is that with a dual coil setup, trigger 1 fires coil B, while trigger 2 fires coil A.

[jsmcortina]

karlo,
thanks for the information. Basically it sounds like I got the triggers mixed up... Not so obvious to test on the stim, but obviously a real issue on an engine.

James

[karlo]

Thanks! It's a good idea to double check my analysis, I don't trust myself that much.

I can come up with a patch for the twin trigger code to support more than two cylinders as well. I hacked the code to work with my V6, but it should work with odd fire V12 (if indeed they exist).

[Smog Fighter]

Thanks! It's a good idea to double check my analysis, I don't trust myself that much.

I can come up with a patch for the twin trigger code to support more than two cylinders as well. I hacked the code to work with my V6, but it should work with odd fire V12 (if indeed they exist).

Wouldn't that be a Ferrari motor?

[karlo]

65 degree V12 sounds pretty odd fired to me, yes.

[Keithg]

Been running this beta for a while as there have been no fixes/enhancements on later versions to prompt me to upgrade.

I am running 36-1, 2nd trigger COP, eae (no lag comp), 1 squirt simultaneous.

I have noticed an odd occurrence. Every so often on WOT accel, I can get spark drop. At least that is what it feels like. It comes right back, but it can be pretty violent. I think it is spark drop because when I catch a glimpse out of the corner of my eye of AFR being full rich then recovering. I could be wrong, but have not been able to get a log when it does this. If I do, I'll post it. Just wondering if anyone else has experienced this or if I should upgrade to current as something has been fixed.

KeithG

[UnaClocker]

I had those kind of problems with the really early beta's, before the Neon code got hosed. Now that the Neon code is completely redone, I don't have that issue ever.

[pac1085]

Been running this beta for a while as there have been no fixes/enhancements on later versions to prompt me to upgrade.

I am running 36-1, 2nd trigger COP, eae (no lag comp), 1 squirt simultaneous.

I have noticed an odd occurrence. Every so often on WOT accel, I can get spark drop. At least that is what it feels like. It comes right back, but it can be pretty violent. I think it is spark drop because when I catch a glimpse out of the corner of my eye of AFR being full rich then recovering. I could be wrong, but have not been able to get a log when it does this. If I do, I'll post it. Just wondering if anyone else has experienced this or if I should upgrade to current as something has been fixed.

KeithG

I have the same issue running 60-2 wasted spark, map ae and 2 squirts/alternating on my 4 cyl 1.8t. im on beta 14.