[theonewithin]

Please note that depending on which conditioner you use, some are inverting.

So as voltage of the sensor goes up the conditioner actually goes low.the wiki has details on the correct settings for conditioners.

[miker]

its entirely possible the code is wrong but i've started with your tunerstudio setup,

Required fuel calc - you've got a 350cc engine with 30cc flow per min. This should be more like 1400, 1600, 1800 or 2000cc with much higher flowing injectors.

looking at the largest log file i don't see the MAP value change. I'd have thought this would have moved when you started cranking and the sync loss are being generated.

To view the log files past the following as a header directly above the data (each word is a different column & its over 3 rows),
#
PriLevel SecLevel Trigger Sync RefTime MaxTime ToothTime Time
Flag Flag Flag Flag ms ms ms ms

When i paste that i can look at the data. What I see is either an engine cranking with the spark plugs and struggling or something very strange. I'd expect to see a gentle wave pattern as the engine rotates and compression builts (hence why i suggest cranking without the spark plugs to remove compression) except i see a fast pulse then slow pulse, fast, slow, fast, slow etc which when combined together create the gentle wave. I don't see any cam signal.

[miker]

Just gone back to look at your images, you definitely do see cam signals that when I look at the data file I don't see. Very strange.

Looking at the image on the scope it is broadly what I expect, but I think it's confusing perhaps due to inconsistent engine speed. It would be good to try again with a freshly charged battery and no plugs.


Also just realised, I think you're using a max vr conditioner, is anticipate your trigger offset is 10 degrees out. Have you checked thinking on a strobe light?

[Trevor Getty]

Nice ideas. I have brought out a new acid gel battery and will use it for my next logging.
Thanks for the tip on the log file - I will be able to fix and load mine in future... much appreciated... - Any idea why they are broken, when written out?

I will concentrate on the cam signal - I know it is lower, and the fact it is now being seen in the logger, does mean it wont fire the injectors.

Regarding the injector volumes - sorry I had setup a t16 setup fully, then when I moved over to your fork of the code, I started a new project - and I have obviously missed resetting these - good catch !!

Finally I will get the timing light out, set the cranking angle to 0 temporarily and make sure the inj1 pulse lines up with TDC, and if not you recon there may be a 10 degree difference in the MAX VR and the previous one I used, so I will confirm this tonight.

Finally I had a thought that maybe the cam wires were inverted, so that also is something for me to check tonight, to see if they are inverted just on my scope image, or if they are inverted on the speedy inputs. I will move my probe leads to the speedy input side,instead of the cam sensor side of the loom - just to make sure everything is as it should be.

I will get back with some findings later, and newly captured logs.

[theonewithin]

Timing is based on IGN not INJ.

Confirm using actual spark output from a coil.

[Trevor Getty]

Timing is based on IGN not INJ.

Confirm using actual spark output from a coil.

All my scoped images, and logs have been taken using an HT capacitive pick up probe on Spark plug 1.
I see where the incorrect assumption came in my comment above where I mention INJ1 incorrectly. - but all logs where correctly taken from a spark plug.

We had been discussing above that we actually were missing output on the injectors, so i wouldnt' of been able to get any timing from them, but it was my mistake typing the wrong acronym and not seeing it.

Mike - good news from tonights results.
So the CAM sensor inputs, where inverted, so it was not correctly reading the 3 tooth section.
I have now inverted the cam leads correctly into VR+ / VR- and the tooth logger has no spawned into life. :p

We now see the 17teeth and gap in the wheel although it shows as a double height bar for a gap - I assume this is correct.(pic and log attached).

I took out the spark plugs to get the RPM and cranking speed higher and was able to take better measurements, and now have my TDC timing correct - which ended up being exactly 360 degrees. 0 Degrees was exactly one crank rotation out, so would of been fine in wasted spark - but when running sequential ( so I could correctly see spark 1 ) I seen we were one rotation out.

Ive attached named pics of me logging at various points, 0 degrees, 90, 270, 340, 350, 360 as I get closer to the correct result. I also included a -359 but its not what I expected. Do you happen to know what 360+5 degrees would be so if I wanted to shift the timing back a little more, would it be minus 10? I tried -359 but it was 360 degrees out at crank. For now running with 360, and then adding a cranking advance of 5 is nearly perfect.
I am comparing against how it was on original ECU running at 10 degrees retard at idle (pic included for reference).

I then was able to go to my main logger and show that I know get RPM showing ( although its only 210rpm) and I do see some small MAP shifts, but most importantly now we are in sync I see injection dwell time showing now!!!

I plugged back in my sparks, and attached the fuel relay again - but found my battery was dead from all the testing, so will have to wait till tomorrow now - but I think we are super close to it firing up now.

Have a wee look at the logs attached, and pictures, and see if you think the code is actually 360 degrees out of sync at the crank, as we discussed maybe a setting of 10 for the MAX would seem correct based on theory, but it was a full crank rotation off for me.

[Trevor Getty]

Attaching main log, and the tooth logger output, as could only have 5 attachments in the last post.
It wouldn't let me attach the mlg file - so I have renamed with (xxx.mlg.txt) on the end, so just drop the .txt and you can load into MLogViewer.
Thanks t.

If you need anything else before I test tomorrow please let me know.

Fingers crossed it will spring into life tomorrow with a charged battery now the rest is sorted out.

[miker]

Will read all this properly later but 360 degrees out is me guessing the phasing incorrectly. It's a small code fix (invert the revolution flag)

[Trevor Getty]

No problem Mike, I thought the 360 out would be an easy fix. ( you had briefly mentioned the crank revolution assumption before ).

The actual *.mlg file looks the most useful now its almost running.
I can see MAP getting lower during crank, injection dwell increase, and rpm sit around 200 rpm.
The good thing now also is that we dont seem to be loosing sync during crank (at least not with sparks out )

These old engines, have a big heavy flywheel and do crank over slowly, I might get a diesel battery and put it into the boot ( no room at the front ) and see if I can kick it a bit more under load.

Ill go out to the car (after work) and see if it fires up (im like a kid at xmas waiting to see if she runs now.. )

[miker]

Not had time to test but i've done a quick update to hopefully fix the rotation & have updated github - fingers crossed its correct

any complaints about not following good development practice for testing before submitting to git will be met by a loud raspberry and a potential delay of a few days for any further fixes (might need to rebuild my laptop so figured id get it in quick)