Air Flow Meter Woes... (I think)

[log1call]

I'd suspect that those readings that are out of range are either logging the wrong addresses or applying the wrong formula to the data. The motor wouldn't be running like those figures suggest.

What is the ecu's internal Id number?

What addresses are you using for Knock, O2 and boost?

What formula are you applying to them?

The AFM figures are kinda low at idle, perhaps there is an air leak or it's buggered. Once the revs get up the figures seem about right for IPW and load though, so I'd suspect air leak.

[Guest lan_killa]

You guys dont seem to be listening too well THE LOG ABOVE IS A GOOD LOG... CAR RAN PERFECTALLY DURING THIS LOGGING SESSION NO MISFIRES/LOSSES OF POWER WERE NOTICED

I am trying to get a log showing misfires/power losses to analyse against this 'Good log'

[log1call]

Ha, yes I got that killa. What I am sayng is that your good figures are not right. I'm not saying I've seen a fault with the motor, just the data you are logging.

The knock figure should be near zero all the time, if it isn't then there is something seriously wrong. Your car is behaving correctly therefore the data is wrong.

When the knock is detected it should only be under a load... your's is showing knock detection even at idle.

When knock detection happens the timing gets pulled and the knock detect goes back to zero... your's isn't pulling timing and it isn't going back to zero. The fault will no be with the running it will be the data you are collecting.

Same with O2, it should be between 100 and 900 millivolts and it should fluctuate up and down. Yours is right out of scale and it doesn't fluctuate.

If you let me know what the ecu's internal Id number is I will try and find you some better addresses or formulas that will get you the correct figures. The ecu Id number will appear in the event log box when you start logging. Something like 703315.

[log1call]

Ok, Killa, you should go here...

http://cid-4ca3c3459aaa7f7f.skydrive.live.com/browse.aspx/.Public?wa=wsignin1.0&sa=52053601

and download the file called "Data". Rename the old data file in C:\Program Files\EvoScan\EvoScan v2.5\DataSettings as "old", and put this new copy in there.

While you are there you should download another file called... "Diagnostics"

Start evoscan and try some of the other ecu models till you find one that gets you figures that are correct. When I say correct I mean figures that comply with the figures in pages 64 to 74 of that diagnostics pdf. Hopefully there will be a model that has your exact ecu Id number, that's the number that shows in the "event log" box when you connect with evoscan, but if not there will probably be something that will do the job. Just connect, see what number shows up, record it and go looking through the list for your number then use that.

[Guest]

Hi log1call,

would you mind if I downloaded a couple of things of interest to me off there would you?

[log1call]

No, go for it. We need a place to store files on here really but that site lets you store about twenty-five Gb of data so I stick stuff there for people. Help yourself to anything you want.

[Guest]

Thank you I appreciate that.

[Guest lan_killa]

ok ecu is 744033, have a log from a few minutes ago which has some large cutouts near the beginning and end of it...

will look at those files youve linked to after ive uploaded this log

[Guest lan_killa]

Logfile: http://homepages.slingshot.co.nz/~moliver1/WRXbadlog1710.csv

Log ID 65 - 90 There was power loss aswell

Log ID 330 - 350 there was 2 losses of power in that section

[Guest]

 log1call said:

and download the file called "Data". Rename the old data file in C:\Program Files\EvoScan\EvoScan v2.5\DataSettings as "old", and put this new copy in there.

Usable in v2.6?

[log1call]

yup, works in 2.6

I just put a modofied copy of the ssm data file into that microsoft site if anyone wants it. It's called "select monitor" and you just rename your old copy in C:\Program Files\SelectMonitor_V1.1.0 and fit that in it's place, or edit your original copy if that suits you better.

Have fun, I'm off to enjoy some cactus!

[Guest]

Cactus as in plant cactus fruit?

[log1call]

Er, no. Cactus as in recreational drug.

[Guest]

i was Going to say !MESCALINE! but didn’t want to offend or insinuate that you would be taking things of that form so fruit was better

[log1call]

Ha, yes, I blame my sons. They are a bad influence on me.

Ah well, you have to hve fun.

[Guest]

So did you? Have fun that is anyway killa that log looks similar to the last one apart from some Disturbing entry’s from you IA and KC I would do what log1call suggested and use the new/old file and then log because again no definitive data there to go on

[log1call]

Yes we had fun. Rainy day, cactus, Quinton Taranteno = Interesting.

Killa, at post 67 to 74 or so, the throttle was open, the revs were up, the load was coming on, the ignition timing was good for the revs and the injectors were injecting heaps.... or they were trying to. The airfflow readings though are low. I bet that if the O2 readings had been working as they are meant to that you would have noticed a distictly lean mixture. If A lean mixture was revealed under all those conditions then the obvious conclusion is that there is a fuel shortage. If on the other hand a rich mixture had been shown then there is an electrical miss or the airflow meter is faulty. I'm guessing lean because you haven't mentioned any AFM trouble code.

You need to get the right figures recording.

It's also better to go for a specific test drive, somewhere on an open road where you can drive at any particular throttle opening/speed for a few seconds at a time.

Down at the other place you mentioned it's hard to tell what is happening because the throttle is shut and there is nothing happening really. If you had been able to floor it for a few seconds something would have shown up probably.

[Guest boostin]

Have you monitored battery voltage through a power loss yet?

[Guest lan_killa]

the funny thing is that its just a complete power loss, sometimes in traffic, sometimes while idleing, sometimes while cruising at a constant speed, sometimes while flat up it....

the fault doesnt reproduce its self dependant on load conditions, can be anytime anywhere.

Today i was sitting at a set of lights where it decided to cut out, I had issues restarting it, strong smell of fuel, i kept my foot up it once it started and held it around 2000rpm it was running rough as balls, smell of fuel was getting worse CEL light was on, after about 30-40seconds of idleing it at 2000rpm it ran smooth and the CEL went out..... but of course i wasnt logging... typical

[Guest boostin]

I would be disconnecting every plug, spraying contact cleaner through them all, and reconnecting with electrical grease once they've dried.

[log1call]

Well if it's got a strong smell of fuel it will be the ignition dying.

If you could log O2 you would see the O2 going to 900+ Mv I'll bet.

You should check all the feed to the ignition really carefully for bad/loose connections. Perhaps leave it idling and wriggle all the fuses. wires, components.

[Guest boostin]

It could be a dying ignitor. Its hidden behind the boost solenoid and MAP sensor mounting plate.

[Guest lan_killa]

u got a spare one at ure place?

[Guest]

 log1call said:

Yes we had fun. Rainy day, cactus, Quinton Taranteno = Interesting.

Do we (trust) this guy? : D he could be seeing pink flowers in the data log file: D but I bet he is right to... Just get a good data log guy, and then it won’t be so problematic to determine the purposed problem IMO

[Guest boostin]

 lan_killa said:

u got a spare one at ure place?

Hmmm, maybe. BG5A TT the same?? I don't know where I've put it though....