In search of good ignition input signal....SOLVED (mostly)

[EWflyer]

At this point my reset problem is as follows:

----- The resets only happen when the laptop computer is connected.

------Each time it happens the fuel pump relay makes a quick clicking noise as it cycles off then back on again.

------I tried using TunerStudio instead of MegaTune and the resets still happened.

If I try to ride with the laptop computer connected (to datalog) the resets with the accompanying fuel pump relay interruption make the bike unridable. Should I try a different laptop? My daughters each have laptops that are several years older and I think they may even have a serial port that I could use. That would remove the serial-to-usb adaptor from the situation.

[24c]

At this point my reset problem is as follows: ------Each time it happens the fuel pump relay makes a quick clicking noise as it cycles off then back on again.

This is standard behaviour for a MicroSquirt power cycle on off, but have you tried to mimic this fault by messing with the fuel pump relay wiring with the laptop connected, and seeing if you can force a reset? Have you tried it on the bench with a stim and your laptop? What happens if you bypass the fuel pump relay connection to the MicroSquirt, does it reset still when connected to the laptop? Is there a fuel pump regulator, blockage in a return line?

[Bruce Bowling]

Laptops can be a PITA sometimes for resets. Two things can cause this (and others for sure):

1) Use of an inverter on the laptop power supply. These spew high harmonic noise back into the uS and can cause issues or noise pickup on the VR input. Or even having the laptop plugged in with its power charger - try running the laptop on batteries only to see if this affects the reset issue.

2) The serial cabling itself is picking up ignition noise and coupling it back into the uS. A big antenna is what it can sometimes act like - not fun on a bike during tuning. You can verify by changing the wire routing for the serial wire, from the ampseal connector all the way back.... keep it separated from any other wires. You may have to peel the serial cable wires out of the cable bundle to keep it away from the rest.

Also - if you have a timing light hooked up to an ignition coil secondary it can back-feed in noise. I had a situation a while back where an ignition light would cause resets - only when a laptop was plugged in. Separating the wires fixed the problem.

- Bruce

[EWflyer]

I really do appreciate every bit of input you send my way.

I may be your worst nightmare: a guy who's very interested and involved in a MicroSquirt project but has absolutely no background in electronics. I read your replies to other people's posts, when you're discussing things like solutions to MAF sensor selection and how to build the curve for it, and I'm amazed. Your depth of knowledge, and that of some of the other forum members has you speaking a language that's foriegn to me. I so want to understand everything I read here on the forum, but I just don't.

My resets are full-on MicroSquirt processor resets: In the datalog the SecL goes to zero and the fuel pump relay cycles off-then-on.

With respect to my Kawasaki EX-250 project, I will try splitting the laptop serial input out from the rest of the MicroSquirt pigtail as a way of isolating it. I also have 12volt Zener diodes on hand (I wanted a 24volt Zener, but I can't put my hands on one in Spokane, WA). I was planning on wiring one of the 12volt Zeners I have inline with my OPTOIN+ which is connected to my bike's coil negative trigger (piggy-backing off the bike's stock ignition system) to see if that reduces/eliminates my reset problem.

I'll be trying out the 12volt Zener on Monday (Dec 14th). I'll solder it in and try running my laptop with the bike for a datalog. I would love to get rid of these processor resets so I can move forward with my project, dataloging and using the datalogs to tune the bike. Without the laptop connected the bike starts and runs easily, but when I ride it that way it's obviously nowhere near an optimal tune. If I could get this reset situation solved, I could datalog and tune via the MegaLogViewer.

Thanks again,

Greg

[triangles95]

I have a 94 vulcan 500 that is very similar ignition wise. Let me know if you ever figure out how to eliminate the CDI box.

[EWflyer]

Well, I was just planning a "fuel only" application for a while. Once I get it that all figured out I might work on the task of giving MicroSquirt control of the ignition.

I've had it suggested to me that I could simply leave the CDI box in it's place and use it's output as an input to the MicroSquirt. It would be a "known" ignition curve that the MicroSquirt could manipulate.

Here's what 800vtwin said:

but does it really matter if you know the exact numbers? Many people install computer and have a blank timing table to stare at. Ita alot of work to come up with a good timing curve. Now add 3d to the mix. (load vs rpm). Your actually kinda lucky in that the timing curve is already done. But you have the ability to adjust it. (just don't get carried away.)

with the timing table at 0, this would give stock cdi curve. you could advance from the stock curve 2 or 4 degrees if wanted. Since you are working off the stock curve, its hard to mess it up.

You also can retard the overall timing and set the table so its back to stcok. -now you could have a high speed retard. (if you motor benefits from it?) set delay teeth at one and back up with trigger offset.

Just don't forget how the stock cdi works. - it doesn't advance. it retards the timimg. The stock flywheel teeth are set up at 42* btdc. The cdi just retards while at lower rpms. and the cdi is basically bypassed at max advance rpm and above.

I don't know much about the ignition end of the Mega/MicroSquirt setup, but it sounds like 800vtwin is making sense here. It would keep me from having to attempt some sort of modification to the bike's existing rotor and VR setup. It sounds like an easy fix to take the bike's original (simple) 2D timing curve and impose a "load vs. rpm" correction to it.

[EWflyer]

If you've gone through this thread you'll know I first tried to use the bike's VR signal which didn't work, and it was eventually pointed out to me that the rotor is non-conforming, it has one long and three short lugs that the bike's OEM CDI decodes to trigger the spark on the correct cylinder.

Then I moved on to try and trigger off of the #1 cylinder coil negative (CDI trigger) line using OPTOIN+ and OPTOIN-. I tried various configurations that were suggested in other threads, but I always ended up with the Microsquirt repeatedly resetting (apparently due to tach signal loss). The processor would interpret something that was occuring in the OPTO circuit as a zero RPM situation and it would momentarily shut the fuel pump off (and the tach count would go to zero in the datalogs), then the processor would instantly resume functioning (apparently the OPTO tach dropout was very brief) and the ASE would re-start causing an overrich condition until the count expired. This situation left me unable to progress with tuning the bike, so I didn't do much with it over the winter.

But now winter is apparently over and I'm very hopeful that I've now found a good solution.

My current solution seems to work very well, although I've had one or two of the tach count type resets I mentioned in the last paragraph. I'm still using the OPTO circuit with OPTOIN+ connected to the coil negative line with a 12 volt Zener diode on it (the bike's coil negative line actually has 12 volts on it except when the CDI triggers the spark plug) and I've got the OPTOIN- connected to a ground wire with a 4.7K resistor on it.

This setup has given me the best results I've had so far. Maybe I can progress to the tuning stage now and end up with a rideable and enjoyable motorcycle.