tach spikes? MS1 tuning issues.

[jazig.k]

hi guys,
iv got my ms1 up and running, pretty well when it gets up and moves, but i know theres more left to tune out of it.

im having trouble with getting it to idle well just after starting. usually when cold, but sometime when warmed up too. iv been playing alot with the after start enrich and warm up enrich, but cant quite get it to idle right. my following log has it in the first start up.
it idles fine if i throttle it for a second and then it will settle down like nothings wrong.

also, when up to temp and cruising along there is a little stutter in the engine. im not sure what it is, but i think its tach spikes? while driving and keeping tunerstudio open i see the rpm gauge flash red and the needle jump.

iv made a log, which ill try and up load.

the car is a alfa 75 [verde], ms1, using the MAP instead of a TPS.

also, does the intake air temp seem high [ambi. temp. was about 20deg.cel.]? iv mounted the sensor just after the throttle body in the plenum, instead of just behind the air filter. is this ok?

[Matt Cramer]

You'll need to filter out the tach spikes before you can do any real tuning. What main board do you have?

[jazig.k]

v2.2
so they are tach spikes?

[Matt Cramer]

Yes - the RPM is jumping from 800 to 13,300 and back. Try putting a 20 to 24 volt zener diode inline with the wire to the coil negative, banded end pointed at the coil, and see if that helps.

[jazig.k]

thanks, using a 20v zener diode worked a treat.

[jazig.k]

right-o,
since fixing the spikes, its revealed a missing feeling at light cruise and light accel. the WB 02 is reading about 13.5~14.5 when it happens and sometimes seems to jump to 19~20 for a split second, most of the time it doesnt stay long enough to read. happens from 1300 to 2000rpm ish.
up higher in the rpm, its fine, and pulls strong without the 'missing'. half throttle or more from any rpm and it doesnt miss a beat.

iv got a pretty big cam in the car, barely town drivable on the original ecu, even when it was tuned [well, 'tuned' for a l-jet system], but now its pretty tame with the megasquirt. am i running too rich for off cam? the megamanual said somewhere that running too rich will trick a wb02 into thinking its full lean.

im having a ball learning about all this, regardless how annoying it can be.

[Matt Cramer]

That isn't anywhere close to too rich; my guess would be an ignition missfire there.

[jazig.k]

a log i made this morning to work, the rpm range is between 1500 to 1700 and a split second before the 'miss' [or leaning out of the 02] rpm spikes downwards to 800, or 500 if its a bigger 'miss' [bigger miss = leaner spike]. looks like tacho drop out? seems to follow my problem quite regularly.

[Matt Cramer]

Unless you have downshifted very abruptly and then revved up just as quickly, that certainly sounds like a tach dropout to me. Maybe you need a smaller voltage zener - or a higher current one if yours is overheating.

[jazig.k]

cool, i thought id try some different sized seners. ill give that a go and get back to you's

[jazig.k]

heres a log i made on the way home, from mum and dads, tonight. should show my problems.

[jazig.k]

hi again, been awhile since iv had the alfa going due to cam problems [now fixed, i hope... ill find out in about 1000kms. last one wernt hardened properly , the second ones wernt even for the correct engine ].
iv still got the stutter around 1500~1700, and a smaller stutter at idle, but ill do some logs and fix what i can. gotta start again because iv forgotten where i was at last time i had the car going, which sucks [its been 8 or 9 months fixing the cam problem with the supplier ].

i know my tps isnt set correctly, so ill set that and check my logs first off. ill probably convert to a pot type tsp, instead of the 3position switch type, shortly too.

oh well, ill be back soon with some questions, im sure.

[jazig.k]

heres my most recent log, taken yesterday.

look at around record 1290 onwards, see the spikes or drop-outs there?? could that be noise? i never had this issue the last time the car was running... it shows up a couple of times through out the log.

what are the "mark 001 reset"s? there is 4 of them in this log [001 002 003 004], and i presume its the beeping that megatune makes at me while im driving?

on the RPM setting for the VE bins, i was reading elsewhere and i think iv got those set wrong for my engine. factory idle is specified at 900rpm, should i have one bin around 600~700, one 850~, and another around 1000~ to cover the idle rpm range?

i also notice at the end, when i turn the ignition off the rpm rises up to 25000. normal??

who in australia can dyno tune megasquirt? i know im not up to that yet, but id like to know if there is anyone that has MS experience.

i feel real lost about all this now, so im reading through the megamanual, hoping to jog my memory...

thanks for any help
zavier

[Matt Cramer]

A "reset" mark either indicates the processor has reset or the communications have dropped out. You can tell which is which by looking at the SecL trace. SecL does one thing only - it counts from 0 to 255, drops back to zero, and starts doing this again (note: MS2 has a much higher maximum value than 255). If SecL drops back to zero before it hits 255, you've had a processor reset. If there is a spike in SecL and then it goes back to where it was, you've had a glitch in the serial communication. Looking at SecL is a good way to test other channels as well - if SecL spikes when everything else does, you've got serial noise.

Given that the vast majority of tach dropouts or spikes are accompanied by a SecL spike, I'd say your problem is actually glitchy serial communications and nothing more.

[jazig.k]

thanks matt,

problem is i dont only get issues witht he resets, i get them throughout the log, and where the spikes are affect the engine, i can feel the spikes like a miss in the ignition or similar. but a miss in the ignition system wouldnt cause everything to spike would it? even my clt sensor spikes...

thanks matt, i appreciate the help, the soon i get this sorted out the sooner i can move onto MS2 and try the spark.

[jazig.k]

would it possibly be a poop ground? i know iv checked them, but im going to check them again, and re run them to multiple places. id guess that would cause the engine to be affected with the spikes.

iv just watched the log with secL, and i do see what you mean. throughout the log, it only spikes, and doesnt lose count, it just returns the where it was at.

[EWflyer]

You're using MS1 which means you're fuel-only, right?

If you are trying to sense the Tach event off of the coil then I may have a fix for you. I also had reset problems in my fuel-only setup.

Look at the picture I posted in this thread. It works great for a fuel-only setup that senses the Tach off of the coil(s). This setup was developed for a four-cylinder motorcycle engine, I don't know if it has been tried on a four-cylinder car yet. Obviously the motorcycle was originally a wasted spark because it has two coils.

http://www.microsquirt.com/viewtopic.ph ... 86#p151772

The picture as posted is for a four-cylinder engine. I had to modify it slightly for my two-cylinder engine.

By its design it completely blocks out all the coil's noise and delivers only a gentle "square wave."

Let me know if this helps.

[jazig.k]

thanks, i bought the parts today, and ill give it a go this weekend... but its the GF's birthday and i already copped alot of trouble for working on the car late into the nights recently... so hopefully ill get to sneek away at some stage, lol.