Speeduino is now on Github Sponsors (Rather than Patreon): https://github.com/sponsors/noisymime
Help with building your Speeduino, installing it, getting it to run etc.
By Mancino
#49551
This is killing me. I can't seem to pinpoint how to solve the noise issue I have while cranking. If you've read my other threads, I've had some nagging issues with re-start once I get running. More particularly cold start. Thinking it was tuning, I figured I just needed to adjust some things. But after looking into it more, I checked some logs and found my RPM reading would suddenly drop to 0 and return again. So hooking up an O-scope, I noticed the 12v rail, 5v rail, and trigger signals get these spikes in their waveforms. All signs point to the starter being the noise maker. I've tried different starters, but non have helped the issue.

I messed with some ferrite cores, beefed up my main ground to the motor, chassis, and battery, added some more capacitance to the 5v rail. All helped a lot but most effective was moving the coil grounds to the chassis instead of to the motor. But even with all this, I still seem to get the spikes occasionally. I'm fairly knowledgeable, but I'm floundering here. When I think I have it pegged, I'll switch things back to normal to check I've really solved it and sure enough, I get a random spike again. :x

I should add, before I tried to fix the problem, the amount of noise was far greater than what I get now. But I have a hard time believing I can't keep the noise down to a minimum level consistently and not stop the sync loss and RPM drop.

Screenshots are of a running tooth log, and megalog view of the rpm drop. Unfortunately I don't see any data while cranking of the tooth log. And I also added a screenshot of the scope. I'm out of ideas at this point...anyone who can help, I'd be greatful! :D
Attachments
20210411_161536.jpg
20210411_161536.jpg (1.73 MiB) Viewed 679 times
Megalog Cranking.png
Megalog Cranking.png (144.09 KiB) Viewed 681 times
Tooth log Running.png
Tooth log Running.png (114.74 KiB) Viewed 681 times
By Mancino
#49552
Just to add, I tried lowering the "skip revolutions" down to zero, and the rpm drop goes away in the logs, but it can be tough to start. Sometimes taking 3+ attempts. Before it starts the car pops a few times like its firing, but won't catch. Here is a log of that...
Attachments
Zero rev skip meglog.png
Zero rev skip meglog.png (147.89 KiB) Viewed 674 times
By Mancino
#49555
theonewithin wrote:
Sun Apr 11, 2021 8:46 pm
Ignore RPM drops. Look at sync loss.

It's no better at all.
It's definitely not! :lol: I'm a pretty stubborn guy, but I'm about ready to throw in the towel. It's just not worth trying to tune it if it still has this issue. I never know if the settings I changed were for good or not because it's loaded with sync losses during startup. :cry:
By Mancino
#49558
Well in a last ditch effort and using scrap parts from an old job, I installed 100uf worth of caps between the 12v input to the ECU and to ground. And low and behold, it helped significantly!! There is still a random single blip in the 5v and trigger pulses which occurs at the same point in time. Not sure I totally solved it, but it's a step in the right direction.
By Mancino
#49620
Yeah, it doesn't totally rid the problem...depending on the combination of ferrites and capacitance, it gets slightly better or worse. I got an assortment of different caps today. I'm going to fiddle with some different combos tonight. But I still haven't figured it out
By Mancino
#49621
Trying some different combinations, I've found I can get no sync loss when I pull the fuses for the injectors, but as soon as I plug them in and a cylinder gets combustion, I get a random sync loss right after...I attached a scope shot. :cry:
Attachments
Sync Loss #3.png
Sync Loss #3.png (151.78 KiB) Viewed 479 times
By LAV1000
#49622
Might be a bad injector or bad connection to injector.
Injectors are just small coils.
Or maybe a bad injector driver.
Does the injector wiring run close to crank sensor wiring ?
User avatar
By PSIG
#49626
It's awesome that you're logging everything, as many of your clues are hiding in the logs. 8-) However, you graph differently than I do, and I can't see the relative interactions in a screen shot. Not intending to be a pain in your ass, but could you post the log file, rather than (or in addition to) a screen shot?
Long post warning....

The issue of calculating accurate timing with a lo[…]

The m44 cam trigger wheel doesn't work with spee[…]

Project Yamaha fzr 400rr injector

Once again... The last few version of firmware ha[…]

Squirts per cycle whatever works best.

Still can't find what you're looking for?