So, I finally swapped in my MS2 daughterboard to use the MS2Extra Beta code. After some heartburn, I was able to get the engine idling well on fuel only. After much more heartburn, I can't get a decent idle after spark is added.

Last known good configuration:
MSI chip
V3.0 board
MS1Extra firmware (029q2)
LED 17 and 19 driving external 4G63 transistor
Stock coil pack
Trigger angle = 69
Spark output inverted = no
Cranking dwell = 25 ms
running dwell = 5.5 ms
minimum discharge period 0.1 ms

New configuration:
MS2 daughterboard
V3.0 board
MS2Extra Beta firmware (10/20/2007)
LED 17 and 19 driving external 4G63 transistor
Stock coil pack
Trigger angle = 0
Spark output = going low
Dwell = well, thats the problem

Is anyone running MS2Extra with 4G63 transistor? If not, what are you running and what are the spark settings?

Corbin

'95 ESI-T
Now with more power and fewer leaks



Gimme fuel...Gimme fire...Gimme that which I desire

Try the latest MS2 Extra Beta, as there were some odd changes made in the 420a wheel decoder that used wrong tooth size values. The past few newest beta's only use one edge of each tooth. Set dwell around 3.5ms to start, check the ignitor to see if it gets too hot at idle. Lower dwell if it does.

Are you running a 4G63 ignitor and MS2E? It would be good to know that it is possible. I want to get the thing running with as few changes as possible. Then I can start improving from there.

Corbin

'95 ESI-T
Now with more power and fewer leaks



Gimme fuel...Gimme fire...Gimme that which I desire

Originally posted by Corbin
Are you running a 4G63 ignitor and MS2E? It would be good to know that it is possible. I want to get the thing running with as few changes as possible. Then I can start improving from there. Corbin '95 ESI-T Now with more power and fewer leaks Gimme fuel...Gimme fire...Gimme that which I desire

I've run MS2E only in the garage, as my RS is in "storage" for the winter, but it runs good so far. I use a Crane ignition box instead of the mitsu ignitor, with the MS firing the crane box directly from the 2 LED's. There is no reason it shouldnt work with the mitsu ignitor. There should be no hardware changes needed from MS1 to MS2, other than the 12volt feed to the MS2 chip, which you really only need if you are running a stepper motor. Just note that if you do connect the 12 volt feed, you MUST NOT put the MS1 chip back in, or you will fry it. I just leave that 12 volts disconnected for testing between the two chips for now. I do run my 4g63 Laser on MS2 Extra using the OEM mitsu ignitor without issue using the 2 leds, so it's fine to use for firing the 420 coil pack.

I'm not sure exactaly when the older beta's had the wrong crank wheel info, but i think it was around the time that you have for beta code. There was debate of tooth gap sizes, and the decoder used all 16 edges of the crank wheel. It was known for the smaller tooth sizes of 20 degrees, which is why the code was re-written to only use those 8 edges instead. It fixed many other's issues, which is why i suggest to go to the latest beta(its Beta20 now).

I'm trying now to figure out how to run a 2nd trigger input from the cam sensor to run COP, for either 4 spark outputs or run wasted spark COP, as this will give more triggers to choose from to run semi-sequencial fuel injection, which in my experience with the 4g63, makes it idle as good as, if not better than stock. Currently, without the cam sensor input, i can only choose 2 triggers, both of which dont seem to inject fuel at a most proper time for a perfect idle. You would only noitice this slightly rough behavior running larger injectors and 2 squirt alternating set.

I just loaded the Beta 20 (1/27/2008) code. Same shit, different day. It seems to be running on two cylinders at a very low RPM and pouring water from the exhaust (purdy good sign of a problem). When I swap control of the coil back to the stock ECU, everything is fine. I can also get it to idle when setting MS2E to a fixed duty dwell of around 5% (whatever that is worth). So, I believe that both outputs are making it to the coil. If I have no choice, I will switch back to internal VB transistors and use some Neon MSQ's I found. I don't suppose you could E-mail me an MSQ from the Laser? cbehnken@comcast.net if you can.

Corbin

'95 ESI-T
Now with more power and fewer leaks



Gimme fuel...Gimme fire...Gimme that which I desire

Try a running dwell of 1.5ms instead. I know some reported having to run very low dwell numbers for some reason..it doesnt sound right to me, but it worked for them. Sounds like the ignitor is current limiting or something. Since i run a CDI ignition box, it doesnt rely on any dwell at all, so those settings dont even apply for me. An msq from the laser wont be of any help, as the settings are totally different from the 420a, although i do run 3.6ms of dwell on it.

Also, make sure "input capture" setting is set for "falling edge".

Are you seeing good rpm? No spikes, dropouts, or lost wheel syncs? Running fuel only you may not notice such issues, but running spark, you will notice any and all odd behaviors.

I am really running out of options now. I switched back to internal VB921 ignitors...same issue. Then I switched back to the opto input circuit shown here: http://megasquirt.sourceforge.net/extra/setup-neon.html...same issue. My next attempt will be to switch to the opto input circuit shown here: http://www.msextra.com/ms2extra/MS2-Extra_Neon.htm. Then I will swap back to MS1Extra and see if I can gather more information.

Corbin

'95 ESI-T
Now with more power and fewer leaks



Gimme fuel...Gimme fire...Gimme that which I desire

There are a decent amount of people running the 420a code with MS2 and dont have any major problems, so there is something wrong somewhere of course.. In this diagram:




Remove and jumper R12, its not needed at all in this input setup.
Replace the "150 ohm" 5v pullup resistor to a 470 ohm.
Make sure C30 is removed as well. This should closely mimic the V2.2 board input circuits, which work perfect.

As for the spark outputs, they are identical to MS1 in respect to the LED mods. I never really cared for the VB921's, so i would really try and use a Mitsu ignitor module- Fire the module from the LED's, and then hook up the module to fire the coil pack. Also, make sure you haven't reversed the ignition outputs, or firing will be 180 degrees off, and it wont run.

Did I mention that I am half retarded? I never filled out the "dwell battery correction" chart under "advanced". Some kind soul on the MSEXTRA forum was nice enough to point that out. It runs now. I really wish I had looked at that before chasing my tail for weeks with input circuits, Fairchild ignitors, P&H injector board, etc, etc, etc. I will post my MSQ when I finish tuning.

Corbin

'95 ESI-T
Now with more power and fewer leaks



Gimme fuel...Gimme fire...Gimme that which I desire