A lot of good suggestions above. Try single wheel missing tooth wasted spark mode to start as this takes the cam sync out of the equation, then adjust your POTs on the main board until you get good signal/sync with the crank sensor. Composite logging is your friend, O-scope would be next. You could even bypass the spark box and try this with one set of coils only to get the car started. Once sorted and ready for driving you can work on full sequential and getting the second set of coils figured out.

Dennis, Jamie,
I didn't get time to do much tonight however I did manage to take some logs. At this stage they don't mean much to me and I haven't really figured out how TS should be setup to log what we really want to see. Anyhow, I have posted them on a thread in the MS forum from where you may be able to download them.

I am pretty sure that the rpm signal is dropping to zero when sync loss occurs.
R52 and R56 are still at their original settings.

Unfortunately I don't have access to an oscilloscope. Getting hold of one would only be half the battle ... then there's the whole learning curve of figuring out how to use it! :eek: I will look into the USB type.

I have made up a short ignition lead that plugs between the coil and the spark plug so I can hook up a timing light. I have checked timing several times and it has always matched the cranking advance setting in TS Ignition Options screen.

Tonight I made up a jumper to by-pass the spark box to be confident that this is doing its job. This means that each spark output was directly driving two coils per cylinder. In this arrangement, there was no firing while cranking. As soon as I plugged in the spark box again, I had firing but not strong enough or consistent enough to start the engine.
Bedtime.

Yes Steve, lots of good suggestions here. Unfortunately I won't have time to work through them properly for a few days yet, but I can see that understanding the logs will be very useful.

Bill.

I would start with the adjustment of the board mounted pots. Here are the composite and trigger datalogs you posted in the MSExtra thread.

http://forums.pelicanparts.com/uploa...1401814095.jpg

the MS unit is clearly seeing your cam signal. The tall spikes are evenly spaced at 342ms corresponding to about a 175rpm starter speed.

What is happening is that the low level filter is not set correctly for the crank input and it is seeing the small spikes. If you look at the tooth signal you will see that the lower side the individual teeth drop out and some of the crank signals are missing. This is why you need to start with adjustment of R52 and R56 on the board. Set those as previously indicated and take the log again.

You can watch the live trigger signal in TunerStudio. Adjust them until the peaks are very even and consistent with no drop outs.

Here is more on the adjustment.

http://www.diyautotune.com/tech_articles/vrsensor/megasquirt_vr_index.htm

And here's what those logs look like in Tunerstudio zoomed in:

http://i458.photobucket.com/albums/q...ps52b1bd87.jpg

http://i458.photobucket.com/albums/q...ps3f59e525.jpg

http://i458.photobucket.com/albums/q...psbec1b40f.jpg

POTS POTS POTS :cool:

They make autoscaling handheld oscilloscopes for not much money. They are worth their weight in gold.

They are stupid simple to use.

I turned the voltage pot one turn clockwise from dead counter clockwise and cleared right up.

Good morning and thanks for your input.
I am starting to think we might actually get there!
Reading these logs seems to be a bit of a black art that I will have to work on.
I am aware of the link to VR conditioning and have read it a couple of times recently, so its significance is now starting to become clear.

Bill, my car died after nearly every light off on cold days and almost every stoplight until I adjusted one turn.

I spent nearly 2 years trying to figure out why it would die. 20-30 seconds with the oscilloscope was all I needed to see. I had a second narrow square wave after the main waves throwing off sync.

Cory. Your welcome. :)

HAHA, yes thanks for the scope. All that time ripping my hair out..... :)

Managed a bit of time tonight to make some adjustments and log the results mainly to get a feel for how this all works. I am still not sure what I'm looking at here and I can't really see any difference between most of these tooth logs.
I didn't try a start tonight, just trying to figure out the logging and what it means.

Adding a 12K resistor into the crank sensor output (replacing jumper at JP1) didn't seem to make any difference, but maybe the tooth log is the wrong one to show its effect.

Maybe I need to keep cranking up R56 and R52?

Log 1: R56 and R52 fully CCW
Log 2: R56 1 turn CW, R52 fully CCW
Log 3: R56 2 turns CW, R52 fully CCW
Log 4: R56 2 turns CW, R52 fully CCW, 12K resistor at JP1 (crank sensor output)
Log 5: R56 2 turns CW, R52 1 turn CW, 12K resistor at JP1 (crank sensor output)

http://forums.pelicanparts.com/uploa...1401974172.jpg

http://forums.pelicanparts.com/uploa...1401974188.jpg

http://forums.pelicanparts.com/uploa...1401974199.jpg

http://forums.pelicanparts.com/uploa...1401974210.jpg

http://forums.pelicanparts.com/uploa...1401974222.jpg

Spent a bit more time today trying a few changes with some confusing results.

This log is without spark plugs, i.e. no compression load. Spike intervals are 58 and well defined. The lack of compression obviously smooths out the log compared to those in the previous post.
http://forums.pelicanparts.com/uploa...1402143701.jpg

The only change from the log above to the one below is that the plugs are installed (but not connected).
The spike interval decreases from 58 to 36. Anyone got any clues on that?
http://forums.pelicanparts.com/uploa...1402144218.jpg

Before running these logs, I added another two turns to pot R56 (now 4 turns CW). I also changed the ignition settings to wasted COP and single wheel with missing tooth to simplify things, but this shouldn't have any effect on the logs that I can see.
I did try a start just after the last log but all I got was a mixture of occasional firing, back firing and exhaust flames.

I just noticed that the last log above has secondary spikes at approximately 55 intervals. Why 55 and not 58?
Maybe these are the missing teeth and they are getting overwhelmed by the spikes due to engine slowing down on compression? But then why are those peaks at intervals of 36?

http://forums.pelicanparts.com/uploa...1402146810.jpg

Have you verified the sensor/wheel gap and/or rigidity of the sensor mount?http://finanziellede.de/hu6d.jpg

Yes, sensor is rock solid (standard 964 mount) and I rechecked clearance (1mm) yesterday.

So far I have turned up R56 by 4 turns and R52 by one turn CW.

How many turns are available on these pots?
Are more turns of R56 likely to reduce the effect of the high cranking pulses?

I guess it won't do any harm to try, but it would be handy to know how many turns are available so I know if I have reached the end of adjustment. Yes I know there is supposed to be a feint click at the end of travel, but I couldn't hear or feel it last time I played with these pots.

Mine don't click so you really can't tell where you are. I seem to remember something about 6 turns but can't swear to it. Have you confirmed your coil and injector wiring? The reason I ask is, despite being what I thought was very careful, I still managed to get 2 coil outputs switched. Also, how did you confirm initial timing given your COP setup?

Did you go full CCW until the "click"? I think they turn 8 times CW before reaching full clocking, but not 100%.

Dumb question, but are all of your grounds terminating at a single point? Sorry if it's been addressed already.

Steve, I am 100% sure that wiring is OK. I did actually have the engine running a few months ago but had to take it out to fix FOD before I got the tuning sorted out. Also, with the spark box in action, in TS test mode you can clearly see the coils' LEDs firing in the correct sequence.
To confirm timing with a timing light, I have a short lead that I run between the COP and the plug. I have done that several times and it always reads exactly the same as the Ignition Advance setting in TS.
When I get home today I'm gonna crank that R56 to the limit and see what effect it has.

No dumb questions in this scenario!
Yes, ALL the grounds are at one point on the crankcase at the old distributor mount.
If you are bored, check out the wiring diagram which is posted on the MS forum here.