Rather than say that Deep's car is overboosting now I'de say it's running the right boost The boost targets are around the 1.4 bar mark, and Defis always read higher than the ECU, typically between 0.05 to 0.1 bar... it was struggling before to get over 1.35. The load axis on Deep's car should be fine up to 1.45 to 1.5 bar, as should Grant's (of course it is VE dependant as the load is MAF based).

Good to see that Grant's car is behaving much better with another brand of fuel! With regard to the occasional ping, it's interesting to look at the K/L at the point it happens. If it "comes out of the blue" then the ECU could not possibly have prevented it because there was no indication that it was about to happen. The only way to counter this type of transition det is to pull the timing so far back that it feels a little lacklustre the rest of the time.

Much more worrying is when the bar graph increases steadily and then it knocks.... this can happen around 6000 RPM as the ECU stops listening to the knock sensor, and the advance curve is extrapolated based on events lower down the rev range.

But as Bob has already said, the final judgement should always be done with a set of det cans, the ear and brain are much more adept than the ECU at picking up det!

Cheers,

Pat.

Edited to remove ambiguity in one statement.

[Edited by pat - 4/3/2003 9:32:56 PM]

 

On the M25 on Sunday,about to have a little fun with an M3,dropped from 6th to 5th floored it and Kl lit up like a Xmas tree.Lifted off and let the M3 pass(much to his satisfaction).Tried WOT again twice(once in 2nd aswell) and all lights lit.Obviously drove very carefully after that.
Just before the journey I topped up the tank with about 33 litres of Optimax.The residual Optimax in the tank already had about 1.5mls/litre of NF and I added another 50mls for the extra 33 litres I added.So about 2mls/litre of NF on average.Never seen all the lights before.Could it just be a bad batch of fuel,despite a decent helping of NF.Any ideas?
STi7/Ecutek/Full SS exh inc upipe/uprated fuel pump.

 

Have you fitted a FMIC yet?

 

Deep as nothing has changed recently it does sound like poor quality fuel. I'm sure that someone said about Optimax losing its additives over time and to avoid garages that don't sell much of it. How you tell which garages don't sell much is beyond me.

 

Hi Moray,no not fitted yet.They did'nt send the bloody instructions in the box! Was sunny but not that hot on Sunday and had been cruising for over half an hour so tmic should have been nice and cool?
Pete,yes thats what I'm hoping.I'm sure my local Shell where I buy from does'nt sell that much,its just not that kind of area.Saying that I've been filling up there twice a week for a year and this is the first time I've noticed all the lights.Anyway I'll soon know once this tank finishes.If its bad fuel it just shows that even with a decent dose of NF its properties remain very substandard.Without the KL and a motorway 'play' with the BMW M3 on Sunday I could have had a melted engine!!!!!

 

I recently had a similar experience although nowhere near as drastic as all lights, only the 2nd green, although on part throttle but well in boost. This may not sound a problem but I don't usually see any activity with the amount of performance I was using at the time. Like you, I was running Optimax (no OB) and had a prior feeling the car did not feel quite as willing as usual. I topped up with BP plus a little drop of Millers but have not had the chance to re-try as I had to swap the Link board for the original for an MOT and have not yet changed back.
I don't think "old" Optimax would drop to the octane level of NUL or below? What other resons would deem it to be "a bad batch" eg NUL in the SUL tank.
Do you normally see any Knocklink activity ie greens?
JohnD
PS 99MY Link, S/S downpipe, PE sport cat centre, Magnex b/b

[Edited by JohnD - 3/31/2003 7:05:15 PM]

 

Deep, with that much NF the ecu should have been able to pull itself down enough if it was below the knock control threshold, did this happen above 6000 rpm or below it ? Did you by any chance notice what boost you pulled ? Its the time of the year when overboost rears its head, cool and dry so VE is high.

 

Deep,
SNAP
Have had the same problem.Spent 2.5 hours with Pat on Friday evening trying to sort out the same thing.We found that by retarding the ignition by 4 degrees it cured the problem.This also has the effect of reducing torque by about 10lbft per 1 degree.
When we mapped it originally we had no problem and the ambient temp was about the same.
We thought it may have been a bad batch of Optimax so run the car nearly dry and refilled at another garage.Unfortunatley no difference.I tryed altering the ignition using Delta Dash and found that 3 degrees of retard cured the problem.
There is one other factor that may have contributed to my problem.
I have been getting a warning light comming on reporting a failure with the variable valve timing selonoid on bank 2.Until it is replaced cannot be totlay sure that this is not the root of my problem.
If you have delta dash would be worth giving it a go in the mean time for safety sake until you can get to see Pat.
Let me know what happens.

Cheers

Grant

 

Four degrees is a shed load, VVT is logged in Deltadash, is that showing a varience bank to bank, if it was stuck advanced then that bank could end up leaner. These cars are very fussy on fueling.

bob

 

Bob,
All I can tell at the moment is that when I logged it there was a variance in voltage between the two banks.The fault seems to be on bank 2 which ever one that is.Actually if I remeber correctly there was no voltage coming from bank 2.
The thing is that the CEL is not on all the time (resets itself after 3 stop and starts of the engine)but the fault is constant.
The car also has a tendacy to idle rough.In the Subaru diagnostics it says that the uneven idle is due to a fault with this part.
Now if the fault does not present itself all the time with a CEL does it still affect the timing.
Your'e right 4 degrees is a shed load and makes an enormous differnce to performance, but as a stop gap if will do until I replace the selonoid.
Before this problem raised its head it produced 300bhp at the wheels using the delta road dyno and it felt it.
Cheers

Grant

 

Funny thing is I have just had a problem with det too, one minute it was fine the next xmas tree.

I have just refuelled with Optimax and a bottle of the finest so I will see tomorrow.

Scott

 

Scary stuff indeed.....who has K/L in stock ?

Cheers,
Ray

 

Hi Bob.Each time it was below 6k rpm.I'm quite sure I was'nt overboosting ie 5th gear wot and Defis showing about 1.4ish(as I'm sure you know they tend to over read a little).So if not to lose 4 degs of advance to cure,and not overboosting what do you suggest is the path to follow?
Grant,I'll keep ya posted.I should fill up a fresh tank tommorrow and I'll do some testing. The fuel c'ant have been that bad that despite a respectable dose of OB its still substandard can it?
All I can say is thank god for KLs.On Sunday I was on my own in the car and was about to have some fun with a cheeky young monkey in a M3 at Motorway speeds,wot and redline revs! Potential disaster!

 

"I have just refuelled with Optimax and a bottle of the finest"

What do you guys recommend?

 

There can be many causes... I think it's a bit of a coincidence that two cars, which have been running perfectly, both suddenly decide to start lighting up the KnockLink, both within a few miles of filling up with fresh Optimax.

It may be that Shell have just changed their fuel composition in anticipation for the coming warmer months. This is normal practice for fuel companies... for example in winter when it's cold you need it to be more volatile otherwise you'de have starting trouble, but in summer you don't want it to be too volatile lest all the goodness escapes out of the filler cap It's possible that the mix has gone a bit pear shaped resulting in lower RON than it should have.

In the meantime I would suggest a good alternative fuel. I've always thought that Esso and BP make a good SUL fuel, with better consistency than Optimax, but lower RON than a good tank full of Optimax.

It may be worth while collecting a sample of the fuel and having it analysed in a knock engine to determine just how good or bad it is. I'de be prepared to meet you guys to collect fuel samples for analysis. If it does fall significantly short of the mark, Shell may have a lot of explaining to do!

Cheers.

Pat.

 

Hi Pat,

Hopefully it is the fuel; however, is there anyway that the mapped ecu could be causing the problem ? As someone stated above...even if the fuel was a bad batch whouldn't all the extra octane booster added by Deep prevent the K/L from going crazy....I could see maybe an extra light or two with bad fuel and lots of octane booster, but for it to light right up under 6,000 rpms (on a 8,000 rpm red line) seems a little extreme ?

Thoughts appreciated please as I have one of your recently mapped ecu's on its way to me

Cheers,
Ray
Nassau, Bahamas

 

Well.. assume that the fuel-pump boys managed to squirt the wrong stuff into the tank under the station. So instead of getting 98.6 octane, Deep, etc. get 95 . That's 4 points to make up, which is a lot of booster (remembering that dose 1 might possible get 3 points, dose 2 gets maybe another 1.5, probably not...).
And the most likely point for knock is 4k5rpm or thereabouts (the area of highest VE), not at the top end. So under 6k would be reasonable.

 

Thanks Nom

I think I need to invest in a K/L.

Cheers,
Ray

 

Taking Pats advice I tried filling up from totally empty (put in 57.8lt)filled up with SUL from a busy Esso/Tesco pump.Then did a ECU reset to be totally sure.Waited until advanced multiplier came up with the help of the latest version of Deltadash.
Have covered 160 miles today and even saw Pat on the A130 for a chat in the rain.K/L red light came on once in 1st at about 6500 then nothing the whole day.Until I found a private road and in 5th at around the same point the red light came up for a split second.
Have ordered the new VVT selonoid and so will have to wait until that is fitted early next week to see whether it is having an effect on the timing.
Have also taken some readings from the valve timing and I am getting diffrent voltage outputs from either sides not to sure whether this is correct or not. If any one else has a chance to monitor them please let me know.

Thanks

Grant

 

Hi guys.Further to Bobs point ie overboosting.Paying closer attention I am now actually boosting higher than I was when first mapped ie 4th gear, wot, Defis reading 1.45-1.5 bar, when before used to only make 1.35 bar.Would this be enough to make the car det under the right(wrong) conditions?
I've filled up with a fresh tank of Optimax and about 1.5mls/litre of NF today and will test over the next couple of days.
Pat,fmic has arrived but no bloody fitting instructions! Off on hols on Monday for two weeks will fit and contact you when I get back.

 

Deep, that would depend on the extent of the engine load axis and the timing values that are in place, if you overboost by that much then the ecu will be looking further along the axis, if the axis is not long enough then it will try and run same timing .... and may det, check with Pat and get a Deltadash log done to confirm its not that, certainly that would be one explaination.

Whilst any fuel can be "bad" Optimax is no better or worse in my opinion, certainly the two cars are not the same in circumstances, Grants has a suspect VVT solenoid. Deeps is boosting higher.

Grant is Deltadash reporting the same VVT timing from each bank, if you tick the boxes you can log and check, my manual lists about six different codes for the VVT, which code are you getting ?
Tthere are 19 pages of diagnostic proceedure in the manual to cover everything from blocked oil feeds to dodgy sensors, tried printing to file but had problems, do you have this info ? If not I'll try and extract it again. You really need to know what the suspect bank is "stuck" on timing wise.

 

Nothing to do with this, is it? The report says dodgy fuel in the North West, but ....

Click

 

Thanks Bob.Yes a DD is needed as soon as I get back from hols.

 

Bob,
The fault code is 1308.
I took some reading this morning.I overlaid the timing with the output voltage for both banks.
If fact the timing didn't move at all but the output voltage varied greatly between both banks.In fact the right bank stayed the same level at 96ma but the left bank was jumping around like a nutter.
Is it right that I got no variance in valve timing through from 1500 to 7200 rpm.
If someone can put up the readings I took at may help.
What I still cannot understand is the fact the ECU does not seem to recognise any det when I see it on the K/L as it does not vary the timing at all.

Grant

 

There should definately be a varience in valve timing across that rev range, the error code is Variable valve timing solenoid valve 2 low input. You need to check with det cans, the knocklink is great but it must be correlated to real det using det cans, if that is not done you could be misleading yourself (either way as it happens). Too much reliance on numbers can lead to problems, back to basics and listen to the engine is safest.


Diags say to disconnect the solenoid connector from solenoid and then measure resistance of connecting loom wire from ecu to that connector, that checks for a cable break or a connector malfunction further back, there is another bulkhead connector involved.


The solenoid resistance should be between 6 and 12 ohms according to the book.

The error code will manifest if it sees the fault for two consecutive driving cycles.

I have the ecu pin outs concerned if you need them.

 

Deep.try using the Shell garage in Welling at the bottom of shooters hill.Always getting through loads of Optimax.Thats where i always fill up......But not for long!!
Got a 330d Touring Sport on the way next week

Dal

 

Note to self.

Fit knocklink this weekend.

 

Have you noticed from the number of posts on det and knocklinks generally how paranoid people are getting about pre-ignition! Don't get me wrong, it's an important subject and knowledge of the causes, effect and preventative measures could have saved a few imports a while back? As Bob R intimated, the use of det cans every now and then could monitor KL activity and maybe allay a few fears (visions of grim faced Subaru drivers giving it stick complete with headphones! - makes a change from mobiles!) So my question is - Is it possible for the KL microphone to pick up extraenous noise from other than det? (ie within the operating frequency limits of the mic.) How likely is it that a "green" is not nesessarily mild det but background noise? (after all it does say "normal" next to the led)
I was once given the instructions on how to put together a set of cans but i've forgoten the tube I.D. and connection to the headphones. Can someone post some details and dimensions? I'm sure quite a few people would be interested.
JohnD

 

Its worth noting that normal 97 SUL could have caused the ecu to pull the advance multiplier down generally which would, of course, then remove the tendency to det as the overall timing would be lower due to less positve knock correction being run. This would also improve the top above 6k wrt det as the ecu runs the last knock correction value on the timing map above that.
Deeps car has still gained boost over where he said he was, worth a double check on that when he comes back imho.

 

John, its normal for the knocklink to pick up and display background noise, correlation with det cans is essential to determine real det, in general steady green lights are background noise, spikey flickery ambers (and the red of course) indicates varying degrees of det from very light upwards. It is, however, normal for some early engines to generate very large amounts of background noise making steady ambers normal. Det cans are the only way to sense check the levels.