MS2/Extra 2.0 RC1 (20080301)

turbo355 · Post by **turbo355** » Thu Mar 20, 2008 11:16 am

muythaibxr wrote:I'll see if I can reproduce that... Was the last msq you posted your latest?

Ken

Yes that was the MSQ i used at the time it did it i changed nothing in the MSQ i came in the house and saved it.

JeroenBosma · Post by **JeroenBosma** » Fri Mar 21, 2008 3:51 am

I have a funny thing: When I set the RPM input trigger to 132 teeth and a second trigger, the firmware become corrupt. I cannot acces the MS2 anymore.

I know it is a weird number but the Porsche 944 uses the 132 teeth on the gearring togehter with a second sensor.

6040solder · Post by **6040solder** » Fri Mar 21, 2008 3:57 am

The maximum tooth count is 60 IIRC. You have probably overflowed an array or something. That shouldn't allowed to happen, but maybe it is.

JeroenBosma · Post by **JeroenBosma** » Fri Mar 21, 2008 5:39 am

Yup, I was able to reproduce it. Flashed new firmware, choose multiwheel and 132 tooth. MS is dead and needs reinstallation of firmware.

Post by **jsmcortina** » Fri Mar 21, 2008 7:41 am

JeroenBosma wrote:Yup, I was able to reproduce it. Flashed new firmware, choose multiwheel and 132 tooth. MS is dead and needs reinstallation of firmware.

Please define "dead" - what exactly happens.

James

JeroenBosma · Post by **JeroenBosma** » Fri Mar 21, 2008 10:09 am

Dead as in it won't come up after a reset. The feulpump won't blink and there is no response. It needs a firmware upload before I can get it back online on MT and having it accept signals (inputs).

I guess the software just stalls.

After a firmware upload it is good to go.

Post by **jsmcortina** » Fri Mar 21, 2008 6:36 pm

132 teeth doesn't work because it is a strange number of degrees per tooth. 360/132 = 2.727272 deg /tooth. The code doesn't support that.

However, you are correct that the code behaves badly in this situation - I debugged it using BDM and the code is stuck in a loop - this prevents the realtime data from being sent to Megatune and it appears to be broken.

I've modified the code to report this as a config error. It should be in the next Release Candidate (RC) we issue.

James

turbo355 · Post by **turbo355** » Fri Mar 21, 2008 7:10 pm

I tried the revlimiter again tonight and it seemed to work just fine i dont know what i could have done before to make it not work i didreload the firmware this time and it seems to be just fine sorry for the false alarm.

Post by **muythaibxr** » Sat Mar 22, 2008 8:41 am

Thanks for letting me know. That saves me time looking at it.

Ken

drytoast · Post by **drytoast** » Sat Mar 22, 2008 12:26 pm

msextra_2.0rc1b_20080303/ms2_extra_ign_in.c: line 716:

Code: Select all

if (trig2cnt == 1) {
  tooth_no = 0;
} else if (trig2cnt == 1)
  tooth_no = 4;
} else {

I'm not actually using the 99 miata code at this point but it seems like line 1 or 3 should be "(trig2cnt == 2)".

AbeFM · Post by **AbeFM** » Sat Mar 22, 2008 7:16 pm

Heh, sounds like a good call. Is this one of the things Arga changed or suggested be changed?

drytoast · Post by **drytoast** » Sun Mar 23, 2008 1:32 am

I'm not sure. I kind of lost track with what was happening. I was just perusing the code and noticed what appears to be a mistype as the second if will never execute.

eliotmansfield · Post by **eliotmansfield** » Sun Mar 23, 2008 1:50 am

JeroenBosma wrote:I have a funny thing: When I set the RPM input trigger to 132 teeth and a second trigger, the firmware become corrupt. I cannot acces the MS2 anymore.

I know it is a weird number but the Porsche 944 uses the 132 teeth on the gearring togehter with a second sensor.

There is a thread somewhere else about doing this. Essentially its a little board that divides the porche/audi ring wheel down to a lower number so you can use MS with it.

Post by **jsmcortina** » Sun Mar 23, 2008 5:22 am

drytoast wrote:msextra_2.0rc1b_20080303/ms2_extra_ign_in.c: line 716:
Code: Select all
if (trig2cnt == 1) {
  tooth_no = 0;
} else if (trig2cnt == 1)
  tooth_no = 4;
} else {
I'm not actually using the 99 miata code at this point but it seems like line 1 or 3 should be "(trig2cnt == 2)".

Yep, I've slipped up there.. now to decide which one is right?!

James

AbeFM · Post by **AbeFM** » Sun Mar 23, 2008 10:19 am

I'm assuming it's this:

1 tooth on the "first" side of the cam tells you the next tooth on the crank is the one near TDC on cyl one.

The tooth doublet on the opposite side of the cam tells you the next tooth on the crank is the one near TDC on cyl 4.

So it seems like the engine is always in bankfire mode right now. Which, as suggested before, is ideal for starting, but wouldn't let anyone run full sequential fuel or spark. I do have COP, but it's set up for batch fire anyway since I'm output limited.

if (trig2cnt == 1) {
tooth_no = 0;
} else if (trig2cnt == 2)
tooth_no = 4;
} else {

I'd do that. 0-3 is degrees 0-360, and 4-7 is degrees 360-720? I'm making that decision arbitratily, really you need to decide where in the code it would decide to fire IGN1 vs IGN4 - which I assume is as I wrote above.

<edit> Is this during starting? Ah, it should start a lot quicker once this change gets made!!! Awesome.

Also, instead of starting a new topic, I would really like to add a request I made before (I'm assuming it's just a flag to set?) could the O2 meter be read/logged when the motor isn't running? I don't care about cranking so much (though it would be interesting), but when not running I would really like to have a real time readout of the O2 meter.

drytoast · Post by **drytoast** » Sun Mar 23, 2008 2:16 pm

This is in the wheel specific code responding to the interrupt from the crank wheel. I haven't installed the MS yet but I will try and get a reading from the cam and one of the ignitor lines to see if I can help determine which is which.

Post by **jsmcortina** » Sun Mar 23, 2008 2:49 pm

I'd commited a fix to CVS earlier, the next few lines after that code snippet show that AbeFM figured it out too.

James

JeroenBosma · Post by **JeroenBosma** » Tue Mar 25, 2008 1:31 pm

I can confirm the odd-fire code is working.

I have placed a new sensor on the engine and matched a "odd-fire" triggerwheel to it instead of using the distributor. The oscilloscope is showing the right patern for the coil. The code for odd-fire 90/150 seems to be good ! I had troubles with picking up the signal but now it turned out to be a sensor / triggerwheel mismatch.

Post by **muythaibxr** » Sat Mar 29, 2008 6:23 am

Thanks for the update Jeroen! Good to hear it's working!

AbeFM · Post by **AbeFM** » Sat Mar 29, 2008 10:59 am

For those 'watching' this topic, RC-2 is out. New thread!