MicroSquirt.com read only forum (Jan.15,2018)

Yes, I have noticed that note in the schematic draw. That is why I'm wondering about the actually high trigger level. I use a UGN3120 Hall sensor which has a Ucesat of 150mV and there should be enough room to the threshold of 1.8V. But that question I will investigate later on.

In the current 12V configuration the tach works fine and the injector ticker sounds great.

Next step is to implement the ignition with two dual spark coils life on the test rig.

After that I have to get familiar with the tables as I use the alpha-n mode. Maybe one question for my preparation: I have found the 6x6 Alpha-N MAP table which has kPa values as result. Are those kPa values an entry of the VE table?

Thanks a lot
Norbert

No the map values can be whatever you want them to be. The VE tables are not used unless you use a hybrid option where map ve starts getting phased in above or below a certain rpm range. In that case they use the standard map ve values.

Hi, tested 2.881 on bench before I wanted to load it into car.
I used the original config in flash except the settings for wheeldecoding and dwell:
Single crank wheel 12-1, 0 delay teeth.
On some steady rpm points I got nearly double dwell and wrong advanced angle,thus dwell looks ok in the logs. I did a cross check with reburned 2.873 which works fine with same settings(fresh msq).

On a further try I compiled 2.881 by myself and disabled the new inter-tach time correction method by commenting out the lines 5277-5318. That worked so far.
Here two scope shots(timebase 5ms):


Working rpm/ disabled correction:


abnormal:

MoveX wrote:Hi, tested 2.881 on bench before I wanted to load it into car.
I used the original config in flash except the settings for wheeldecoding and dwell:
Single crank wheel 12-1, 0 delay teeth.
On some steady rpm points I got nearly double dwell and wrong advanced angle,thus dwell looks ok in the logs. I did a cross check with reburned 2.873 which works fine with same settings(fresh msq).

On a further try I compiled 2.881 by myself and disabled the new inter-tach time correction method by commenting out the lines 5277-5318. That worked so far.

I want to thank you for the completeness of this post! You posted the problem and backed it up with data and the section of code that may be causing the issue - excellent! It really helps alot!

I am sure Al is checking into this right now. Question, it seems that the issues is at the point where there is the missing tooth. If you set the advance offset to some other number do you still see it?

- Bruce

There are only a few rpm difference on the two shots. If I rev higher it disappears but on further points it will rise again. I put in some offset to (-5 deg) no improvement.

Same happens with 2 teeth delay:



Here the msq I used:
http://www.movex.de/down/v2.881.msq

Let me echo Bruce's thanks for finding this. I was able to see the problem last night, but it could take a while to fix. This is now my top priority.

MoveX,

I think I fixed your problem last Saturday, but I have spent the past week looking at and fixing other problems in the dual spark code. I tested all sorts of combinations of modes and no. of cylinders, etc, and think I have a decent working version now, so it is attached as version v2.882. No new features were added and no changes whatsoever are required for MT. Anyone who is using dual spark mode should upgrade to this version, and future units will be programmed with it.

Thanks again for the testing.

Al,

I have also put the code here: http://www.megamanual.com/ms2/betacode.htm. The INI and settings file are there as well, for those not upgrading from 2.881 (those upgrading from 2.881 can use the same megasquirt-II.ini and settings.ini files they were using previously).

Lance.