limp mode but no error codes

[DJones]

Top shelf....

Well done, been watching from a distance as i could not add any assistance.

Nice work...

[S2central.net]

At last I can enjoy the homefry to the full after checking the knock sensors, one was loose. Tightened to 20nm and off down the road, let the engine oil warm up and booted it. Back is the quick spool up and gone is limp mode. What a difference a knock sensor makes. Just clocked a 2.5 sec 50-70mph in 3rd which is plenty quick enough for me.....at the moment:mischeif: With any luck the 2.75" exhaust in July will get me a bit more over this 7A 2.25" system.
Thanks to everyone for the help and advice. One to remember for the future.

Oh praise be - absolutely brilliant... Looks like two torqued up Greens is OK then. Your patience is commendable I must say - and it proves again that it was something stupid.

[4WheelDrifter]

Glad you got it sorted. At least you know the knock sensor is doing its job.

Managed to stay out of limp mode long enough tonight to show a couple of chavs in a peugeot GTi what acceleration in 1st and 2nd is like with 4wd. He sat revving his engine at the lights and left tyres squealing and suddenly found this old Audi 1/2 car length ahead by the time he was out of 1st gear, despite me giving him 1/2 car length head start, into second and leaving him pretty quick now then.....limp mode cut in, but by then I'd proved the point...

That was the give away.
Happened to me when I first had my cq20v many years ago. Always seemed to be when I was mid-overtaking.
It drove me mad trying to find the problem. Times were tough before s2forum. Had to learn the hard/expensive way.

[20Vcqdriver]

Sorted at last

At last I can enjoy the homefry to the full after checking the knock sensors, one was loose. Tightened to 20nm and off down the road, let the engine oil warm up and booted it. Back is the quick spool up and gone is limp mode. What a difference a knock sensor makes. Just clocked a 2.5 sec 50-70mph in 3rd which is plenty quick enough for me.....at the moment:mischeif: With any luck the 2.75" exhaust in July will get me a bit more over this 7A 2.25" system.
Thanks to everyone for the help and advice. One to remember for the future.

[S2central.net]

Sorry Dave - what drawing ?

[20Vcqdriver]

Just looking at the drawing and there appears to be a resister accross T3x1/T3x2 and T3y1 /T3y2. What value should this be just in case it has become disconnected.
Any other tests I can do?

[S2central.net]

Perhaps I'm being sceptical but if Audi see an opportunity to charge more for a slightly longer cabled sensor with a different coloured plug then they'll jump at the chance....
Will retorque and check positions, after that will try and source a new knock sensor (brown one) though going on my last experience I'll need someones reg number and chassis code to get it....

NO chassis number required - just ask for the BROWN knock part number I posted above.

[denniselbows]

i run 2 green knock sensors wish out any faults . Can't see this is your problem.

[20Vcqdriver]

It is more than cable length... Don't you think Audi would have made the same sensor with long enough cable for both locations if that was the case ?

One sensor picks up knock on three cyls whilst the other listens on the other two so they have different sensitivity AFAIK.

Them not being torqued correctly will not help though... All the books say this is important.

Perhaps I'm being sceptical but if Audi see an opportunity to charge more for a slightly longer cabled sensor with a different coloured plug then they'll jump at the chance....
Will retorque and check positions, after that will try and source a new knock sensor (brown one) though going on my last experience I'll need someones reg number and chassis code to get it....

[S2central.net]

It is more than cable length... Don't you think Audi would have made the same sensor with long enough cable for both locations if that was the case ?

One sensor picks up knock on three cyls whilst the other listens on the other two so they have different sensitivity AFAIK.

Them not being torqued correctly will not help though... All the books say this is important.

[20Vcqdriver]

Ahhhhah.... They are critical - different sensitivity on them.

For 3B
BROWN one 077 905 377A (930mm long) - tracks to ECU #11 which is G61 for cyls 1-2-3
Green one is 054 905 377A (740mm long) - tracks to ECU #29 which is G66 for cyls 4 & 5

Checked this across the family album twice and over three other 3B tech resources here.

Was told that other than the length of the leads there was no difference and could be used. Is this not the case?! Anyone else running with the 2 green plugged knock sensors?

[S2central.net]

Ahhhhah.... They are critical - different sensitivity on them.

For 3B
BROWN one 077 905 377A (930mm long) - tracks to ECU #11 which is G61 for cyls 1-2-3
Green one is 054 905 377A (740mm long) - tracks to ECU #29 which is G66 for cyls 4 & 5

Checked this across the family album twice and over three other 3B tech resources here.

[20Vcqdriver]

One thing has occured to me. When I bought the new knock sensors 18months ago during the upgrade, I had to buy two of the green plugged version as this was the only one available without having to go through a spanish inquisition on car reg number, chassis number etc. Difficult when car used to be a KV 10V! Question is which position should the green plug in the loom go to and is it critical, would it cause this issue if sensors were the wrong way round, including affecting how quickly the power comes in (seems to decide what to do for a second before letting me have full power after going WOT as boost rises and then takes off as though it goes to full boost and then provides the correct fuel and timing to match 0.5 secs later....)

[S2central.net]

20Nm is correct... Only electrical test is a resistance check. You *should* be able to drive with them disconnected. That will raise fault codes and retards timing by a good amount but I don't think it goes into limp mode with disconnected knock sensors.

If you are getting false levels of knock detection - it would explain things as first of all the ECU starts to retard timing on detection of knock - it does this a few times to try and cure it (usually bad fuel or extremely hot air) but if that fails then it turns off active boost control ==> limp mode.

I hope its this - I'd be pouring petrol on the back seats soon if its not found soon :-)

[denniselbows]

the knock sensor bolts should be torqued to 20nm i would say only way to check them is to slacken them off and retorque them.