TS help note says
"Either MS3X or LEDs. The MS3X outputs are ready to use logic spark outputs. The 'LED' spark outputs will typically only be used on upgrades from MS2."

http://forums.pelicanparts.com/uploa...1404032729.jpg

Make that two Perth boys with big smiles!

http://forums.pelicanparts.com/uploa...1404032814.jpg

YEAH! I see a fan turning in the picture. What was the problem and solution?

Brown dog lives!
 

Sorry for lack of detail in previous post, but had to take SO out to dinner!
During today I back-tracked through a bunch of posts, notes and manuals and made up a list of things to recheck, reset and retest.

An obvious target was getting the right combination of input capture and sensor polarity, so that was the first on the list.
Log 22 is without compression and falling edge spark capture. This clearly shows the double peaks that indicate wrong setting or polarity.
http://forums.pelicanparts.com/uploa...1404050082.jpg

Log 23 below is same settings as for 22 but with input capture switched to rising. This log is exactly what we want to see, so here was one step in the right direction. :)
I had tried this test last week and each time I ran the log with rising edge, the log seemed blank so I had a closer look at the only one I saved and it seems that something caused a really high spike which dramatically increased the vertical scale of the log and made the real pulses very small, so I didn't pick them up.
http://forums.pelicanparts.com/uploa...1404050213.jpg

Log 24 below is still with rising edge but with plugs installed.
In this log, the single pulses for ignition are visible amongst the other peaks (which are due to cranking speed) and looked sufficiently well-defined to maybe get some ignition action so I thought I'd just give starting a try just in case, so I plugged in the ignition and injection fuses and gave it a crank.
http://forums.pelicanparts.com/uploa...1404050310.jpg

I couldn't believe my ears ... it started! :D
Which brings us to log 25 ... engine running.
http://forums.pelicanparts.com/uploa...1404049467.jpg

Now you would think that that would be the end of the drama, however because I had a whole bunch of tools and parts sitting on the rear window that were rattling around and threatening to fall into the engine bay as the beast rumbled, I shut it off to clean up the area. After that, there was no way I could get it going again.
I even logged it a couple of times to see what was happening and it seemed like the ignition pulse was getting swallowed up in amongst the "compression" pulses as in log 26 below. In this log, some of the ignition pulses are just not there.
http://forums.pelicanparts.com/uploa...1404050581.jpg

I have no idea why this should be and at this stage I was ready to chuck a can of fuel over it and take up smoking, but the words of old geezer neighbour kept haunting me about tow starting. To compound the issue, one of my sons was also pressuring me to tow it up the street as well so I succumbed to intergenerational peer pressure, put the wheels back on, hooked up the tow rope and off we went.
It didn't take too long for it to burst into life and once the tow rope was unhooked and no fuel or oil leaks were evident, I was off for a couple of laps of the 'hood.

The engine is running extremely rich at idle - clouds of black smoke - but seems to drive reasonably well.
Throttle response with ITBs is amazing! Flooring it at 2000rpm in any gear is instant acceleration. The engine revs extremely freely and is going to be a ball to drive when properly tuned.
Now I need to start looking at getting the tuning sorted out.
I stopped and restarted the engine a few times without problems, but the real test will be a cold start.

It seems that the combination of incorrect ignition input capture setting (should be rising) and fluctuating cranking speed due to tight engine (causing pulses that masked the ignition pulse) were the main culprits. How many times did that ignition input capture subject come up yet I managed to avoid identifying it as the real culprit until today!
Along the way, we managed to deal with a few wiring issues related to grounding which probably weren't helping, but I am so grateful that we all got there in the end.
Thanks to everyone who contributed.

My only hope now is that the engine has freed up enough so that it will spin smoothly enough on the starter to actually fire up when it is cold.
Oh, and I forgot to mention ... it is LOUD. The whole neighbourhood knows that this project has reached the next stage. :D

Bill. Congratulations at getting it running. So happy. The tuning part will seem easy compared with this.

Now, when you have it running, open the box and tune the two potentiometers until you have the cleanest signal. This will help drop the noise levels and allow you to start.

Awesome Bill!! Totally happy for you. Was worried you were going to butcher the best MS harness ever made.

It is always something simple. I always learn that the hard way. It's just the high amount of settings and variables that make slow to figure out.

Now the long journey of tuning is where it gets a little hard IMO.

Funny, I think tuning is harder than the install, but Jamie you feel opposite.

It boils down to how big your injectors are and how stable they are at idle/light loads.

Unstable and it's painful to tune. Run stockers and it's a breeze.

Just depends....

Congrats Bill- glad you were able to get some seat time. I find that always helps put some wind back in my sails as well.

If you tweak further then you can try the noise filtering settings under the ignition drop down menu. It's right next to your knock sensor settings/ same drop down as the wheel decoder. Megasquirt-3 MS3 Ignition Settings

I'd also suggest changing the load algorithm in your basic settings to ITB before you start tuning. Manifold pressure doesn't work well with ITB setups and you'll feel pain if you go to different altitudes as things go lean/rich as I notice you have only one MAP sensor active on the board.

You could also try one of these devices on the MS power supply line.
Car Radio 12 Volt Accessory 10 Amp Power Wire Engine Noise Filter Eliminator | eBay

Good luck - let us know how you're doing from time to time as you continue wrapping this up.

How can tuning possibly be harder! :rolleyes:
I did read that life would be easier at idle and light loads with smaller injectors so I stayed with the standard 22lb 964 injectors. I have some larger injectors if I need them, but I doubt it.

Changing load algorithm to ITB is something I am considering.
I haven't logged it yet, but I did notice that the TS dashboard was showing very little vacuum at idle.
The lack of vacuum was a problem for brakes as well. The pedal was really heavy, so I need to revisit that area. Currently I take vacuum from a small port on each TB to one connection for the brake booster, but it obviously isn't enough. May need to drill larger ports into the intake runners.
Vacuum to the ECU is collected from one TB on each side of engine. Just remembered I have a diagram of that (what a surprise! :rolleyes: ).

http://forums.pelicanparts.com/uploa...1404086195.jpg

Before you do too much more on the filtering, I'd focus on getting your idle mix right. The fact that you were able to push start the car, and it started ok the first time, makes me think your re-start challenge might be mixture related. You can get the idle AFR tuned pretty easily. Do you have a wideband AFR as part of your set up?

My megasquirt experience has always been you spend days chasing a problem with a complex solution and it ends up being something simple and obvious.

I am planning to read up on tuning the mixture tonight.
I flicked through a few of the setup menus in TS but didn't see an obvious way to tackle the idle mixture, so back to the manuals!

There is an oxygen sensor and Innovate LC-1 on each exhaust bank but only one is working properly at the moment. I have yet to calibrate them, so the readings are only a rough guide at present.
The AFR readings will only be used for monitoring the engine, not for driving the fuel injection.

Bill,

With the engine running go to TS and find the tuning tab. Open VE table 1. This will be a 3D table showing your fuel map. There are two cursors. A blue one will show where the engine is currently operating in the load/rpm space. The red one is where you can make modifications. toggle the red cursor so that it is in the same proximity as the blue one. Then hold down shift key and use the up arrow to add fuel (richen) and down arrow to remove fuel (lean). I tune idle for lowest MAP reading. This is usually about 13.1:1 on a 911 motor though it depends on each individual build.

Ha! You make it sound almost simple!!
I am looking forward to getting into this but unfortunately you'll have to wait a few days for any feedback as I am seriously busy at work and car is pretty loud, so night-time tuning is a bit anti-social.

I rushed home from work today just to see if it would start. It did. This was the first cold start since we tow-started it yesterday, so it seems that the engine has now loosened up enough for the cranking speed to be more stable.

Actually, tuning the idle is quite simple. Get it running and start adjusting the VE table in the MAP/RPM range of the idle until MAP is at a maximum. Set idle advance at 5 degrees -- or whatever your engine book calls for in the stock configuration (it will idle much better at higher advances, but you do NOT want to go there since it will degrade the engine's ability to idle smoothly under differing conditions.) In the end, the VE table in the "box" around the idle MAP/RPM should all be equal. The "box" will be 3 boxes square centered on the idle RPM/MAP. Do not try to hit a specified AFR -- take what the engine gives you -- Jamie's is 13.1, my 1980 SC closer to 14.0.

Good luck with usable vacuum from the ITBs. A while back, Bowling and Grippo, on the MS forums, developed an electronics package that takes all cylinders' MAP signals, compares each and outputs to MS the minimum MAP that it sees. They reported that this minimum MAP was more consistent with MAP seen by a non-ITB manifold. My latest info here is about 3 years old -- when I switched to Extra and did not look back.

Your reports of working with the 3.6L has tweaked my interest in the same conversion. What a slippery slope! But I'm retired now, on my very last day of work, and ready for another project!

Brian

I don't think k you can tune it without ITB mode from the limited reading I've done. Jamie knows far more for ITB's, but it makes sense, since ITB's lose vacuum really early, so the MAP thinks you floored the throttle over richening.

You need the TPS to make a call in the algorithm......at least what I've read.

I know only one car that runs speed density with ITB's and it runs like crap. It's like a screen door on a submarine... you just don't do it.

^^^ lol!

Bill, Jamie was working those keys on my car like lightning. He's pretty quick. I too haven't used the hot keys at all, just plugged in whatever number I thought it needed. Totally the inefficient method.

Good to hear she ran again!

Bill,

Glad you were able to get her started. I thought that you needed "rising edge". With MS3 you also have the option of using the "Idle VE Table" which allows you to adjust VE in the idle range without affecting the VE settings in the VE1 table. Just some info for you to read about;).

Oscilloscope turned up today. :rolleyes:
I think I'll put that learning curve on the backburner until I really need it!

I haven't had a chance to make any setup changes yet, but I thought I'd try a cold start again today but no success. The engine wanted to run and was only firing weakly and just couldn't get up enough steam to actually run. No synch loss :D.
Keep in mind that it was 3 degrees C in Perth this morning, which is as cold as it gets.
I plugged in and increased the cranking pulse width in stages up to about 470 and also tried varying the cranking advance. 10 degrees seemed best, but it just seems like it 's not getting enough fuel.

Any suggestions on best settings for cold start.
Cranking fuel looks like this.
http://forums.pelicanparts.com/uploa...1404390853.jpg