Instable timing at idle

[woodst]

Hello!
I equipped my 914 2.4l with a MSII V3 board. Runs quite well at higher rpm, but cranking and idle is ... lousy.
I just ruined my 3rd starter and took out my MS unit to check it on an osci with the stim.

Results: At rpm below about 1000, ignition signal timing varies quite a bit. I see a signal with the correct dwell duration but the rising edge (initiating the spark) floats from 0°BDTC to about 25° to 30°BTDC!
Thats what it does in the car too during cranking and idle.
It doesn't matter which calculation method is used - it's always the same observation. Most of the pulses are timed correctly, but after 4 or 5 pulses there some erratic ones. It's getter better with rising rpm.
Prediction settings have no influence. When I reduce the "next pulse tolerance" lots of pulses are missing. The remainder seems to be fairly correct.

My configuration:
Pertronix Ignitor (tried stock points too - same story)
fixed distributor (no advance)
Pertronix Flamethrower (tried stock Bosch and Beru coil too - same story)
MSII V3 rev2.891
Spark advance was 0°BTDC all over the table during testing, no IAT oder cold advance
Testing with stim (constant squarewave signal) and dizzy hooked up to drilling machine (somewhat inconstant speed)

I guess this is why I damage my starters. The engine sometimes fires far too early (maybe starting backwards) putting heavy loads on the gears - and breaking them....

Any ideas?
Cheers!
Malte

[Matt Cramer]

What is your trigger angle? If it's in the 0-20 degree area, try setting cranking trigger to Trigger Rise mode.

[woodst]

Matt, thanks for your help....
Cranking trigger is "trigger rise" since "calculated" didn't work due to the variing rev speed during cranking.
The engine fires up quite reliably, but there are those "kicks" while cranking chipping the teeth of the starter gear - sounds terrible!
And still the shifting ignition timing during idle or at low rpm resp....
Seems to me as if there is some problem with the CPU code.

Why does timing shift between 0° and about 30° BTDC during crank and idle when triggered by a rock solid square wave from the stim?
Soldering spots are perfect and the signal is crisp and clear - but in the wrong moment.

When I choose EDIS or multispark resp., the ignition pulse comes exactly in the right instant - but in this case the MS advance table is useless, since the pulse is fixed to the edge you choose for triggering....

Cheers!
Malte

[Matt Cramer]

What is your trigger angle?

[woodst]

Matt, I can't tell you.
I use a Pertronix ignitor module in the car, which is setup as the manufacturer recommends (about 0.4mm gap).
I got another engine with stock D-jet and ignitor running perfectly smooth.

Using the stim, I would say it's about 50% since it's a square wave....

When I was using stock points in the car it was 46%. I didn't test this configuration with the osci, but had the unstable running with an observable shift with the timing light as well.

Do you think I should install a monoflop between the ignitor and MSII simulating a distinctly shorter trigger pulse?
Are there issues with MSII related to trigger pulse duration?

Cheers!
Malte

[Matt Cramer]

Sorry, some different code versions have different names. The number I need is the Trigger Offset angle, under Ignition Set-up, Base Ignition Settings.

[woodst]

Ok, now I understand....
It's 0°.

[Matt Cramer]

That would put your timing at 0 degrees when RPM is below cranking, and it would make a big jump to the timing table as soon as the RPM goes one RPM over cranking RPM. Could you post a data log of the timing instability along with your MSQ file?

[woodst]

Hello Matt,
thanks a lot for your esteemed help.....
No, I don't have a datalog of the settings, when I detected the timing issue. I only saw it on the scope.
There should not be any jump since the whole timing table contained zeros and any other influence was ruled out (IAT retard, cold advance).
I reviewed my last datalogs when I drove the car - which isn't possible now, due to a broken starter gear.

Interestingly they don't confirm my observation. All stored data seem plausible - from my limited point of view. Nevertheless my timing light showed a distinct shift of timing too. Most visible at low rpm, less at higher rpm (3500 rpm). How can this be? A faulty driver maybe would lead to correct data storage but not produce a correct waveform on the scope, I presume. Dizzy is brand-new. No lash or wear in the driving pinion.

I try to attach the MSQ and a datalog. The datalog is made with a different spark advance table than in the MSQ and knock sensing enabled - sorry for that. But note the first peak in the timing immediately after cranking and prior to the knock retard....

Best regards!
Malte

[Matt Cramer]

Try triggering on the falling edge instead of rising. With the Ignitor and a 12 volt pull up, the rising edge is the start of the dwell, and the falling edge is when the Pertronix triggers the spark. If the Pertronix is adjusting the dwell for some reason, the rising edge will be unstable.

Note - you'll have to change your trigger angle to make this work.

[woodst]

Thank you, Matt
I will try this and report next spring....my car is going to sleep for the winter. No chance to do it now - it's far too cold....

What I still don't understand is why the timing shifts on the stim, no matter what triggering edge I use.

Anyway, a digital system has more conditions than 0 or 1. That's for sure.

Best regards!
Malte