[theonewithin]

You don't need a running engine to confirm timing...

Pull spark plugs. Hook up timing light. Crank and see what you get.

It's what I did on my bike as I wasn't going to pull an engine cover off to measure anything when this is far easier.

Obviously when the engine is running you check again but it will get you very close.

[alessandromeyer]

Unfortunately I have troubles with my crank sensor signal. It's (or rather got!?) unreliable.
Sometime it remains stable for some seconds and then starts "jumping" again (jumping referring to the needle in TS)
Here a video to illustrate the behaviour, this was at about the worst point in time, usually its a little more stable
https://youtu.be/jcAesoKUK8w.

Here's the composite and teeth logger:
Composite - https://drive.google.com/open?id=1JAmL9 ... NMmbb-QuAx
Teeth - https://drive.google.com/open?id=1AcJYi ... zWfHUzrIQh

What I have tested so far:

- Changed the plugs for resistor type (didn't change anything whatsoever)
- Tried adjusting sensor distance to teeth between 0.5mm-2mm. The sweet spot is really between 1mm-1.25mm, this is noticeable but doesn't cure the problem.
- Went back to the old coil + dizzie setup, the signal might be marginally more stable but generally as unreliable (disconnected the efi-coil)
- Tried with filter set to weak - that didn't really change anything either, maybe 10% more stable, but I guess that is expectable.

Interestingly the TPS signal is totally stable, while it follows the same path.

My trigger config: https://drive.google.com/open?id=17EUdp ... n6rcqh_qep

I had some metal particles on the sensor at some point that were from the bracket I fabricated (should've cleaned it better). I tried removing it with some standard magnet and a lot came off up to the point where i have a hard time seeing any more with my eyes. To make totally sure that is no residue i will be trying again with a neodym magnet.

What bugs me the most is that the sensor worked fine before changing over to the 106b coil pack and that disconnecting it doesn't help now

But other than that I'm stuck, i have no idea what to besides trying with another cherry gs1 sensor?

Any further ideas?

Would the output from an oscilloscope help beyond the tooth logger?

Thanks a ton!

[jrbe]

Are you using suppressor core ignition wires? Solid core wires and non resistor plugs can both cause noise issues.
You may need to increase or decrease the sensor to trigger wheel gap if it's out of spec or if the trigger wheel material is a bit off you may need to adjust slightly out of spec (likely closer.)

[PSIG]

Would the output from an oscilloscope help beyond the tooth logger?

Yes, if they show either activity when you key-on, or you can catch any weirdness during the rpm spikes. Your tooth log looks good (36-1, no cycle signal, right?), although the tooth log does not capture at least two gaps, so we can't count teeth. Again, if it isn't while the rpm is bouncing, then it would of course look fine.

One tiny oddity though, and it's probably not it; but your composite log shows a cycle signal detected. Assuming you don't have one (?), then you may have some weird noise on that input. Put the pullup jumper onto that second channel and jumper to the Hall setting. However, if you do have a Hall cycle signal sensor, I would add a resistor from VR2+/RPM2 (depending on board) over to ground. It should be a small resistor, of about 10x the value of your pullup. Post the msq that you used to take those logs.

[EDIT] Curious - what effect did higher filter settings have?

David

[PJ1678]

If looks are important...https://store.jenvey.co.uk/throttle-bod ... itage-dcoe

[alessandromeyer]

Are you using suppressor core ignition wires? Solid core wires and non resistor plugs can both cause noise issues.
You may need to increase or decrease the sensor to trigger wheel gap if it's out of spec or if the trigger wheel material is a bit off you may need to adjust slightly out of spec (likely closer.)

The wires came with the coilpack (it's been used before) they are stock vw leads and look unprotected. Interestingly i switched over to resistor type plugs (didn't need them before with the mech dizzie) and was hoping for cure. Unfortunately absolutely no difference.

The signal really got worse the closer I got (0.5mm-1mm). Best seems to be between 1-1.5mm, it makes the signal a little more stable but not as much as I've hoped.

[alessandromeyer]

Would the output from an oscilloscope help beyond the tooth logger?

Yes, if they show either activity when you key-on, or you can catch any weirdness during the rpm spikes. Your tooth log looks good (36-1, no cycle signal, right?), although the tooth log does not capture at least two gaps, so we can't count teeth. Again, if it isn't while the rpm is bouncing, then it would of course look fine.

One tiny oddity though, and it's probably not it; but your composite log shows a cycle signal detected. Assuming you don't have one (?), then you may have some weird noise on that input. Put the pullup jumper onto that second channel and jumper to the Hall setting. However, if you do have a Hall cycle signal sensor, I would add a resistor from VR2+/RPM2 (depending on board) over to ground. It should be a small resistor, of about 10x the value of your pullup. Post the msq that you used to take those logs.

[EDIT] Curious - what effect did higher filter settings have?

David

It's actually a 60-2 pulley, i think they are called motronic pulleys because they have been used with my different motronic-controlled cars.
If cycle signal is the one that usually is gathered by monitoring cam then no, i'm running wasted spark currently.

I will look into the cycle-noise. Thanks for the hint i must've overlooked that. I'm currently using the UA4C board.

Will post the msq and hopefully some oscilloscope video - what is the best place to attach the oscilloscope, at the jumper maybe?

Thanks again
Alessandro

[PSIG]

Checking the processor input would be most useful. If your find any weirdness, then test again at the input connector. If you again find odd stuff, then test again at the sensor (signal wire disconnected), in order to perhaps find where the strange stuff is entering the system. You have not yet identified what is actually causing the rpm bouncing. That is the first priority. At this point we are guessing and you are assuming, as it likely is, but it may not even be in the sensor/input circuit.

Post your tune. Your last msq showed 36-1 for trigger settings. You didn't say what stronger filter settings did?

David

[alessandromeyer]

So, i was with the car but unfortunately the battery was almost empty thus it was even harder to start.

Even worse it was my first time actually using the oscilloscope and i figured i need to read how to use it:
https://youtu.be/gq17KYU4lv8 (sorry for the bad quality)( the car runs with the dizzie in this vid)
I set it to 5v, dc and tried different hz ranges, the signal looks ok no major spikes, but it's hard to tell since they appear so out of order.

current msq is here: https://drive.google.com/open?id=1KI4GF ... J7LZ3nRtEm

I tried playing with rising vs falling edge, but that didn't change anything (i know, to be expected).
I tripple checked all wires couldn't find anything off. Weird thing is I don't have the bounces no more, but instead It randomly nulls, meaning that it the gauge shows a null and returns to a reasonable rpm. So similar to before but inverted?

Latest composite looks similar to before. https://drive.google.com/open?id=17QEEI ... r4NbJzdI0P

So much for today.
Alessandro

EDIT: by the way the ecu, lambda probe tps and the cranksensor share a common ground post. And speeduino feeds TPS and the cranksensor with 5v. Is there anything i should be concerned about?

[PSIG]

EDIT: by the way the ecu, lambda probe tps and the cranksensor share a common ground post. And speeduino feeds TPS and the cranksensor with 5v. Is there anything i should be concerned about?

5V and 12V sensor and signal grounds for Speeduino to read ("sensor return") should all go to one GND terminal or wire on Speeduino. Power grounds (e.g., O2 heater, all of the other Speeduino GND wires, etc) should all go to one engine ground point (star ground). For example, this means your Speeduino power and remaining grounds should be at one location, and your WBO2 controller power ground should be at the same location. A VR- is an exception (VR is alternating voltage and current, not DC), and should be to the designated VR- terminal that feeds the VR signal conditioner directly.

David