20 replies to this topic

[Entaran]

Anyone who has ECMlink application can view these:

Check Directory:
http://www.killcount...chive/datalogs/

04 is a simple datalog of the car coming upto temp (takes a bit over 5 minutes) with a little bit of reversing out of the driveway at the end.

06 is me punching it around the block with it cutting out like crazy and CELing me all over the place (Check knock counts) @ 12psi

Comments welcome.

Edited by Entaran, 29 January 2010 - 03:48 AM.

[Entaran]

Also can someone tell me if my O2 sensor is dying or if that's normal?

[Entaran]

Added 3 more.

Looks like the knock sensor is picking up on mechanical noise between 2600 and 3200 rpm? Isn't that a harmonic point for these engines? Which means... crank pulley?

[brisvr4]

Is the knock sensor in good condition Lorrie?
What injectors do you have in there mate?
It looks like Phantom knock, I have my sensor disabled under 2500rpm and 30% tps.
Try that and see if it helps mate.
Have you put a couple of these logs up at Ecmtuning for Tom and Dave to have a peek at?

I also noticed that you never got any higher than 75%throttle in log #6 mate. I take thats because the cel was going off it's tree?

Edited by brisvr4, 29 January 2010 - 08:01 AM.

[Entaran]

I pretty much put the chip in, got the computer working (few hassles there)...

Did the TPS adjustment thing they tell you to do and then let it warm up (logging), stopped logging, took it for a drive (logging), tried flooring it, copped knock and yeah went bad from there.

Also, the knock sensor worked fine 4 days ago on the standard ecu? I'm thinking the ECMlink has a different set of code? Or something broke on my car while in the shed for 4 days turned off whilst not being touched.

It's DEFINATELY phantom knock, there is absolutely no pre-ignition or detonation as the timing is nowhere near far enough advanced or retarded for either of those circumstances to occur. There's a crazyass rattle in the engine, I had a local auto parts store urgently order me an ATI Super damper today so hopefully that'll be here monday or tuesday. See what it does then.

Also replacing the water pump, alt and ps pulleys and using new hardware to secure them, so if they are loose or rattling will fix that too. Might replace timing belt pulleys and idler pulley while the belts off fitting the ATI Super Damper (it comes with custom timing gear apparently) and refit the AC Compressor while I'm there. Day on the hoist coming up.

Garage queen for the weekend though. Not worth driving when it's ill.

Still shits me.

[brisvr4]

I know the feeling mate, mine is sitting in the garage with 2 smashed gearboxes!

[Entaran]

Managed to fix at least half maybe 3/4's of the phantom knock. The bolts in the AC Comp cover pulley had sheared and the comp cover was belting against the chassis creating a really tinny metal noise constantly. A couple of screws and some silicon later it doesn't rattle anymore. (Will replace comp completely at some point).

still got some phantom knock at full throttle along with a weird noise coming from teh throttlebody area sounds like metal rattling against itself. Maybe have a few loose bolts banging around near the knock sensor somewhere, dunno. I think the crank pulley is also creating some harmonic noise that the sensor doesn't like.

Pity I can't find anywhere to simply turn the fucking sensor off.

[MDK87]

The bolts in the AC Comp cover pulley had sheared and the comp cover was belting against the chassis creating a really tinny metal noise constantly.

I'm confused. I've played with alot of VR4's but I can't think of what you're referring to??

Edited by MDK87, 30 January 2010 - 06:14 AM.

[brisvr4]

Pity I can't find anywhere to simply turn the fucking sensor off.

Well you could just tell the computer to activate the knock sensor after 9000rpm and over 100% tps. It would never register knock then!
NOT something I would do though mate.

[bazeng]

You have forgies right?

Piston slap?
Any idea on your piston to bore clearances when it was rebuilt?

I know my previous build was rather loose and it rattled at around 3000rpm especially on decel.

[ozyvr4]

I had some massive knock when I first started playing. If you go to the ecmlink.com & downloads, you can download different maps. I used the evo8 mod1 & all my knock went away. It was great place to start.

If you higher compress piston & big boost on the standard timing map , you will never get rid of the knock on pump fuel. The 1g map is set for a car running 7.8:1 complression on 10psi. It is very aggressive.

Edited by ozyvr4, 30 January 2010 - 10:14 AM.

[Entaran]

Mine's like 8.8:1, forgies, really fucking tight tolerances. I'll try that map soon thanks ozy, I need to replace the crank pulley with a dampener first and find this rattle that I have and eliminate it.

For now I reset the knock sensor to not come on at all so I can actually drive the fucking car. I know for a fact it's not detonating as the knock was occuring under vacuum which = phantom.

AC Comp Cover... as in the pulley that is on teh Aircon compressor. There's 6 bolts that hold it into the bearing and all 6 sheared so the pulley actually came away from the compressor. I siliconed it and drove a pretty serious metal spike through it so it wouldn't move at all. (Compressor is broken anyway). When I get around to it, i'll replace the compressor. Needless to say there's no belt on it... but I have no aircon and it's summer!

[twdorris]

I was asked to take a look at this thread. I'm not sure if I'll be able to keep checking back in regularly, so you might want to start a thread on our support forums if you haven't already.

I agree all that knock retard at part throttle seems like phantom knock to me. Any time you're getting 13 degrees of knock retard with only 11% throttle and still under vacuum, something's wrong. I'm hoping some of the recent mechanical items you're addressing will help fix that.

What I can't explain, though, is why this would be any different now versus when you had the factory chip installed. Can you re-install that factory ECU and see if the thing drives better? I'd be very interested in knowing if anything is really different. Certainly this is atypical behavior, of course.

The knock retard at full throttle is bit harder to dismiss without some data. I would agree with the previous poster that you should try the EVO8 mod1 tables. If that helps, great. If not, try pull another 8-12 degrees of timing manually using the timing sliders on the ECU Config tab (easier than the direct access table). If that STILL doesn't change knock retard at full throttle, then something non-tuning related is ticking off the knock sensor and you will at least know where to focus your efforts.

Hope that helps,

Thomas Dorris
ECMTuning, Inc.

[Entaran]

Updated my thread over there mate Thanks.

[BMGTZ]

I am using the evo 8 tune and found it to be a good place to start

[vertex_vr4]

Also can someone tell me if my O2 sensor is dying or if that's normal?

O2 sensor is working correctly - its bouncing around in closed loop and ~0.84-0.90 during closed/WOT. It will do the 0.02 thing on coast.

Looks like a little timing needs to come off b/w 3.5k and 4k and also some at 6k but I can't tell you at what load because I'm only seeing the .csv results...

The rest of the problem is pretty much classic phantom knock. If you are driving an ASpec I have a patch for it which disables knock detection under a certain amount of load.

(Load ~ units of O2 per cylinder per revolution)

[Entaran]

TBH I turned my knock sensor off period. Car drives significantly better now.

Will worry about it once I have the external mechanical issues sorted and then let the boys at dougmo's work argue with the knock sensor when it's strapped to the hub dyno.

[twdorris]

I can't tell you at what load because I'm only seeing the .csv results...

You can grab the log viewer off our site if you're interested. It has a live cell trace feature that will clearly show exactly which cells to address as you're scrolling through the log (or running the engine for that matter).

If you are driving an ASpec I have a patch for it which disables knock detection under a certain amount of load.

He's running ECMLink which includes a set of knock controls already. We did not add load-based control simply because TPS tends to be easier for people to estimate off the top of their heads. But if someone were REALLY interested in a load based trigger, I suppose we could add that too.

(Load ~ units of O2 per cylinder per revolution)

I'm SOOOO glad to hear you say that. It's so hard to convince some people in the DSM community of this!! Airflow per cylinder is an excellent indication of cylinder pressure which is an ideal load indicator.

Thomas Dorris

[Entaran]

I'm just glad we have people on this site who are knowledgeable about the situation because I'm rather stumped.

[vertex_vr4]

twdorris - Interesting. I've not really looked at ECMLink but it from what I'm reading its been quite well thought out. Is ECMLink based off the factory ECU binaries or is it a rewrite?

From my experience with eliminating PK on a VR4 - changing the cutoff by TPS would be just as good (and for the reasons you mention perhaps better) as load. So probably not worthwhile.

The OP (Entaran) mentioned CELs - are these coming out because of ECMLink is throwing them on knock or are they real? And is there some function in ECMLink to retrieve the values - and perhaps address another underlying problem?

Cheers,
John