Never ending problem

[nuk1ear]

Basically im not sure what to do now. Ive tried so many things. I can only think that my issue is related to the Peak+Hold Driver and that i should try inline resistors instead......

As you can see, during cranking i just cannot get it to start properly. I tried anywhere from 0-8deg cranking timing, 0deg seems to work better, but either way, its as if my cranking is too rich or something. Which doesnt make sense to me because i get backfires out the exhaust (pops etc) very easily. Once the car starts it runs like turd. Wont idle by itself, makes the car knock around a bit.

This is all really weird because i started it today at about 20deg c and it started up a little hard but started, then idled fine (PWM warmup) until about 45deg then just randomly died. Its not like my idle values are strange either because i tested the valve (open from 70+ and closed from 30-). I just cant win.

I drove it yesterday because it was running alright, but then at 1 point i got a heap of white smoke coming out of the exhaust for a minute or 2. Then it stopped and i didnt see it again (drove for about 20minutes). Possible head gasket? Maybe one of my piston ringlands is cracked? I have no idea. I wanna do a compression test and leakdown test if i can, but in the meantime what in the world is wrong with my setup to be causing such strange hard starting?

Please see my MSQ and attached log.

gen 2 3SGTE with coil on plugs and 850cc delphi injectors

First thread on this problem:
http://msextra.com/forums/viewtopic.php?f=131&t=44046
Below is the thread which i solved a stuffed transistor which was making injector #1 stay open constantly:
viewtopic.php?f=131&t=45875&start=20

[franksidebike]

good morning

Please see my MSQ and attached log.

in your log I can see you have Sync Lost Reason 11
read this http://www.msextra.com/doc/ms3/syncloss.html
your battery is low when not cranking I see 11.2V
frank

[nuk1ear]

yeh battery is low now since I've been cranking it so many times (battery is 2 weeks old, i bought a new one). I also see that loss of sync count but i don't think it matters if it happens while the car is about to die or immediately at cranking. if it happens when running i would look into it.

either way it doesn't explain how the car would start this morning and idle and run perfectly fine, then later on it won't idle and it won't start unless i crank it for a couple of seconds, then flood clear and it will start. Im ready to part this piece of s*** car out. its been the biggest waste of my time. No wins, just constant s*** problems.

Just tried some other settings, i think the stupid amount of fuel going down the exhaust has destroyed my wideband sensor too and if anything, the way the car acts it sounds like a blown ring land or something, its vibrating insanely. It will start (after a lot of effort cranking and flood clearing), but won't idle or run smoothly whatsoever, will die instantly even with fully open iacv and proper fuel. I can't understand any of this. Doubt the car is worth much as it is, maybe i can make some money back parting it out and wasting more of my time

my other log is over 2meg so i can't upload it

[franksidebike]

lost sync reason can come and go in different situations!
hot or cold sensors? electric nois? bad wire? low or hi rpm? low batery? .....
when you have sync lost your engine can run very rough or rich or no!

when your piston is broken, it's every time broken!
frank

[billr]

Have you tried setting ignition to "wasted COP", see if that makes a difference on starting? Since you have to use "clear flood" to get it started, why not back off on the cranking PW a bit? Do you have a sensor for fuel pressure? If not, might consider adding one and logging pressure. Fuel pressure is every bit as critical as PW in determining mass of fuel that is being injected

[nuk1ear]

tried wasted COP and it somewhat works better yes, but still once started it is all retarded. I have a fuel pressure gauge on the regulator, sits dead of 40psi

[billr]

I think we have been through this before, a fuel pressure gauge is not the same as a sensor that can catch fast transients and log data to graph. But, my guess is that fuel pressure isn't really the problem. I think it is related to the cam sensor. I suggest you take one tooth off of your crank wheel, make it a 24-1, then disconnect the cam sensor and try that way (with wasted COP still, of course). Don't worry about putting the missing tooth back, a 24-1 can work as well with the cam sensor for full-sequential as the 24, maybe even better.

[nuk1ear]

thanks for the reply,
Before i go removing teeth from my distributor, could i just try and understand the reason for that idea? So the error 11 relates to a cam signal coming in at an unexpected time right? Cant i test the signals using composite logger? I dunno, i just keep seeing that the signals appear to log ok but as we can see from the logs, there is some loss of sync error.

What if i try switching to the other G sensor (this is a nippon denso CAS, and as you definitely know there are 2 cam VR sensors)?? I used good shielded cable and grounded it ONLY at the megasquirt end. If i do try the other G sensor, would i just need to subtract or something 180deg? Or 360deg from the 1st tooth angle? currently my angle is 271.1 and using a timing light it was accurate.

Would the signal explain all of the issues? You can see that im also getting stupid rich AFR (yet no black gas from the exhaust, must be actual fuel liquid coming out i guess).

[nuk1ear]

could this be relavant?
http://www.diyautotune.com/tech_article ... ippondenso

I had an issue where the car would idle down to about 1200rpm during warm up and once it got there it would just suddenly die. Then i changed the crank signal to raising edge and left the cam at falling. maybe i should try swapping them around and see if any of the combinations works well?

[billr]

I'm going to back away from this, I didn't grasp before that this is 24 teeth in the dizzy and is one of the trickier configurations. I was thinking it was 24 teeth on the crank... sorry.

[nuk1ear]

I'm going to back away from this, I didn't grasp before that this is 24 teeth in the dizzy and is one of the trickier configurations. I was thinking it was 24 teeth on the crank... sorry.

fair enough, thanks for the advice either way. Nippondenso system is known to be "more difficult" generally?

[billr]

That comment was just based on the DIY tech article you referenced; it makes sense what with the edge-triggering and poor phasing of the crank and cam teeth described. Most of my experience has been with a 60-2 crank and polling cam, pretty simple and fool-proof.

[jsmcortina]

Post a fresh composite log.

James

[nuk1ear]

Post a fresh composite log.

James

ill get one up tomorrow, thanks James

[atomic6]

its as if my cranking is too rich or something. Which doesnt make sense to me because i get backfires out the exhaust (pops etc) very easily.

This may be one of your problems, popping in the exhaust, is a clear indicator of rich condition.

If I was you I would get some high impedance smaller injectors in there, get it running on those(high z small injectors will be much easier to get tuned correctly), note the PW at idle, and use that to give you a referance, for the switch to the giant injectors to intend to use.

You could put 20 lb injectors in there and find that it idles at 4 ms, so the switch to the 85 lb injectors would likely dictate a sub 1 ms PulseWidth idle, witch could very problematic,and very difficult to tune.

So sometimes the base fuel pressures have to lowered to get a larger PW at idle(as many injectors are not very linear at small PWs), for easier tuning. The MS3 can easily deal with difficult injector curves, but the tuner behind the wheel will need some experience with this stuff to get thru it.

[nuk1ear]

i understand what you are saying but ive seen cars with over double size injectors running well during idle. 850cc is far from huge. For evos a basic injector upgrade is to 1050 or 1200cc injectors and they still idle fine. On top of this i do have peak and hold driver, which has one purpose of reducing the injector dead time somewhat no?

I dont think richness is my issue here. My Req Fuel calculation is only 3.5ms and im using between 200-100% all the way from 0deg up to 80deg for cranking PW%. There must be something else causing the issue, such as bad tach signal which results the the car not starting for too long.

My questions atm:
-If the car has a bad tach signal, will TS still read SYNC even if its getting sync errors? I think so right?
-If the tach signal is giving errors, at what point will the megasquirt STOP injecting fuel and firing spark? Or at what point exactly will it START? This is something im not sure of yet.

[nuk1ear]

Ok check out the logs. Im pretty confused honestly. The logs look completely wrong, theres equal number of cam signals as crank signals! wtf? how does that work? Or am i reading it wrong? Please note during these logs/captures, i disconnected all injectors. Also the battery is getting low so ill charge it over tonight.

The final log i did with the basic noise filter settings activated.

[nuk1ear]

2nd set

[nuk1ear]

3rd set

[nuk1ear]

Finally please check these! Note that this was my last test and obviously there is a tad bit of fuel in the engine (even tho injectors disconnected) and the engine started to fire up for a second when i switched on the filter. This makes me think i may need to adjust one of the resistors a bit to get the correct signal in TS?

My triggers are wired with a 3 core shielded cable, both triggers and the ground run in this cable. The shield is grounded on the megasquirt side. I used G2 sensor and not G1.

The TS shows SYNCED even tho i get this signal? The errors also stop after a few? What steps should i take first to solve this? Thanks very much for the help once again everyone.