Bloody nora why is it so hard to find an ECU wiring diagram! Is it top secret or something?!

The Bosch Motronic ME7.5 on the AUQ is a pain to find a diagram for.

I have a block diagram, but it doesn't have pinouts. I can point you towards a J220 schematic for the ECU as applied to the 5-pot Audis, in case that would make a starting point - the knock sensor pins might be the same, but watch out for the shared earth as that could take a lot of other sensors out too.

http://forums.quattroworld.com/s4s6/msgs/21009.phtml

I've recently been banging my head against this too when trying to diagnose fault codes. I know what the fault code says, but what I really need to know is how the ECU arrived at that conclusion - how are you supposed to diagnose an issue if you can't compare expectations with reality? Is there any documentation anywhere explaining the inputs and process behind each fault code? 

Edited 3 August, 201015 yr by Interphase

I have a block diagram, but it doesn't have pinouts. I can point you towards a J220 schematic for the ECU as applied to the 5-pot Audis, in case that would make a starting point - the knock sensor pins might be the same, but watch out for the shared earth as that could take a lot of other sensors out too.

http://forums.quattroworld.com/s4s6/msgs/21009.phtml

I've recently been banging my head against this too when trying to diagnose fault codes. I know what the fault code says, but what I really need to know is how the ECU arrived at that conclusion - how are you supposed to diagnose an issue if you can't compare expectations with reality? Is there any documentation anywhere explaining the inputs and process behind each fault code? 

Cheers for that, I'd come across that diagram in my searches. I share your pain with diagnosing sensor readings. I'm a sponge to information, but it drives me insane when I can't find the damn information! Anyway...

I bit the bullet yesterday and paid for a days access (14 Euros) to https://erwin.skoda-auto.cz/ which is an official resource for Skoda technical documentation. There's loads of info in there, and it gives you all your car's delivery details and specification from the VIN, then shows you the technical documents based on a VIN. Annoyingly you have to download a bunch of software to view some of the files, but that just takes time. Still, I can't register Oracle, so I can't save the damn documents. Just view and print, which will have to do.

From the official wiring diagrams on erWin:

G61 knock sensor:

Blue wire from 2-pin connector (0.35mm wire) goes to pin 99 of the ECU (at the 40-pin connector)

Grey wire from 2-pin connector (0.35mm wire) goes to pin 106 of the ECU (at the 40-pin connector)

Shielding (0.35mm to 0.5mm when it joins the shielding from sensor 2) goes to pin 89 of the ECU

The pin numbers are labelled on the ECU, so you can see where the pin numbers are once you pull the plug off the ECU.

So my readings were:

Grey wire (sensor plug to ECU plug) = 0.2 Ohms

Blue wire (sensor plug to ECU plug) = 0.2 Ohms

Sensor plug (grey) to sensor plug (blue) = 1M Ohm

Blue wire sensor plug to Earth = 0.3 Ohms

Grey wire sensor plug to Earth = 1M Ohm

The blue wire is obviously the Earth wire, and is a shared earth for both the knock sensors. I have the loom a good shake and poke when taking the readings, and I had no fluctuations. So based on those readings I think the wiring is spot on. The specification for the wires is no more than 1 Ohm of resistance, so at 0.2 I'm looking good. I'll try and get the wiring diagram posted up for others.

I also checked out the intake air tempertaure sender wiring and injector wiring, to have a comparison, and found the results to tally up for Earth and 0.35mm wire resistance.

So, now what?! I might buy another knock sensor, and try again, just incase I did crush this one.

Unless I am actually seeing intermittant knock on cylinders 1 or 2. But if that were the case I'd like to think I'd know about it, in terms of noise and power (the car drives fine with bags of power). I guess I could have a cylinder running stupidly lean, but why, and surely I'd notice. If one of the injectors was playing up, surely I'd get an error code? The injector seals do weep slightly, but nothing shocking. Carbon build up in the head? I'm lost. I'm fairly certain it's not knocking though. She'll still hit 150mph happily etc. I suppose I could inspect the injector spray formation by pulling the rail and injectors off the manifold, and putting some jars under the injectors while getting someone to turn the engine over, and view the volume and spray pattern.

Literally any thoughts welcome! I might ring up Ben later and see if he has any ideas on the subject. I'm off to look for some small jars.......

I take it the 1 Megohm readings are from checking for shorts, and the meter is hitting the top of the scale? I'd be worried otherwise

If it was actually knocking, wouldn't it trigger one a performance threshold error for that cylinder instead? From what I gather it combines the known position of the crankshaft sensor with the timing of the knock sensor to work out which cylinder is in the compression stage when the problem occurs, and reports back precisely based on that. That would imply that any error relating to the state of the sensor itself or the level of the signal would be more likely to be an overall fault than a triggered one.

I'd imagine 'signal too high' would usually stem from a short to positive, but you'd imagine that'd happen quite often. Could it be that the sensor is somehow triggering for too long, which would take it over the expected time for an individual knock and result in a high reading when one shouldn't exist? If it had been overtorqued that could theoretically stem from an everyday engine movement. My only concern there is that you might expect it to follow a knock error, and it doesn't.

Out of interest, the guy I was working with yesterday regarding my car said he'd seen all sorts of failures from not having that sensor correctly torqued, and that it was often because it's a pig to see the torque reading because of where it's positioned so lots of people just get fed up and do a 'good enough' adjustment 

Edited 4 August, 201015 yr by Interphase

Yeah the 1M Ohm readings are checking for shorts. In theory they would be open loop (infinate resistance), but since you can't have perfect insulation, 1 million Ohms is pretty damn near close to infinate, so it's a good reading.

I didn't know about that cylinder specific performance threshold thing. Interesting stuff, cheers for that.

I went for a 30min hard thrash last night, and no error code. If the code does come back I think I might have to buy another sensor, and borrow a more accurate and calibrated torque wrench from work to do it. Just to be sure. Like you say, it looks like these things a super sensitive to torque!

Since the ECU has to learn from scratch again now, I'm going to make sure I only run 99RON from now on. It already felt a bit quicker last night, but that may have been because the roads were empty for once!

Thanks for all the help mate, I'll try and post up a picture of the wiring diagram from TB to ECU for you. When I get my scanner working!

Sounding good

We'd actually managed to work out where the six wires went and what they did (angle position 1 and 2 outputs, angle sensor + and GND, motor drive + and GND) and we've now patched out the angle sensor 2 signal wire with a fresh tested one away from the main loom. Interestingly, the difference in surface texture between the new and old connectors at the ECU end was surprising, suggesting that they'd all tarnished to a fair degree. The old connector had also opened up by a fair amount so we're hoping the new patch lead does the trick - it's been fine so far

There a seem to be a fair range of error DTCs relating to knock sensors - you've got individual threshold errors and regulation limit exceeded errors for up to 16 cylinders (I believe), errors in the control module, circuit high / low / intermittent / malfunction / range & performance, short to ground... it just goes on  Not sure how many are implemented for this engine, but it seems reasonable to assume that a knock will be detected as a knock, and anything else is related to the performance of the circuit.

Edited 5 August, 201015 yr by Interphase

Does any1 know what way round the wires go into the plug? I own a mk1 octy vrs and after changing knock sensor noticed wires wernt connected. Thanks.

Bump

Hiya Bodge, 

 

I feel a right tit for resurrecting an old thread like this.. but I can't seem to make private messages either, which I'm guessing is because I'm new....

 

I have a 2005 MK1 Seat Leon Cupra R (BAM Engine).
Hope you can help as I'm starting to pull my hair out over this one!

So I bought a cheapo VAG-COM cable (KKL) and noticed I was getting the following error code:
16712 - Knock Sensor 1 (G61): Signal too High
P0328 - 35-10 - - - Intermittent

Here's what I've tried so far:

• Cleaned existing knock sensor & retorqued to 20nm
• Replaced knock sensor
• Had wiring checked (all ok)

Now if I clear the code down, it disappears under normal driving. But under hard driving, it shows up again.... any ideas? 

 

I know the engine you had probably isn't the BAM one, but I'm hoping you may have some ideas!

 

Cheers,

Adrian

 

Edit: I've got a thread over here too: http://www.seatcupra.net/forums/showthread.php?p=4364222&posted=1#post4364222 

Edited 19 February, 201412 yr by AdrianL

Bump. Would someone be so kind to PM Bodge to this thread?

 

Thanks

Adrian

Hi Adrian!

 

No worries resurrecting the old thread. Lots has changed over the past 4 years, but this error code isn't one of them!!

 

My engine is the AUQ (180bhp standard), but the knock sensor system should be the same as the BAM.

 

Sounds like we have the same issue.

 

My car even spent a day on Shark Performance's rollers having a custom remap and no issues were flagged with knock or timing. The code just seems to pop up once a month or so, but with no running issues or other symptoms. Can't seem to link it to anything.

 

I don't worry about it any more as it doesn't actually cause any issues. I'm assuming it's a gremlin in the system. I've checked/changed everything I could think of to fix it.

 

I'm only amazed that I seem to be the only Mk1 Octavia RS owner that has this error code. Seems to be a common 20vT issue that nobody has ever resolved.....

 

If you ever get rid of it let me know 

Like you're saying in your Seat Cupra thread, it's not actually knocking. It's something else going on.........

Hi Adrian!

 

No worries resurrecting the old thread. Lots has changed over the past 4 years, but this error code isn't one of them!!

 

My engine is the AUQ (180bhp standard), but the knock sensor system should be the same as the BAM.

 

Sounds like we have the same issue.

 

My car even spent a day on Shark Performance's rollers having a custom remap and no issues were flagged with knock or timing. The code just seems to pop up once a month or so, but with no running issues or other symptoms. Can't seem to link it to anything.

 

I don't worry about it any more as it doesn't actually cause any issues. I'm assuming it's a gremlin in the system. I've checked/changed everything I could think of to fix it.

 

I'm only amazed that I seem to be the only Mk1 Octavia RS owner that has this error code. Seems to be a common 20vT issue that nobody has ever resolved.....

 

If you ever get rid of it let me know 

 

Hi Bodge, thanks very much for getting back to me!

 

That's really interesting. I'm actually going on a rolling road day on Sunday so it should be interesting to see if there is a lack of power. I've cleared the fault code again so it will be interesting to see if it pops up after the run. I'm glad to hear it hasn't caused you any adverse effects though. I could only find one other active guy on the Seat forum who has had the issue and he said it just occasionally comes up. Like you said it must just be a gremlin in the system.

 

I'll be sure to let you know if I ever figure it out!

 

Like you're saying in your Seat Cupra thread, it's not actually knocking. It's something else going on.........

Yep. If anything, the next step is wiring straight to the ECU... then failing that, a new ECU... but as it's not knocking and (we shall see Sunday) if it's not down on power, then I'm tempted to just leave it.

Hi Bodge, thanks very much for getting back to me!

 

That's really interesting. I'm actually going on a rolling road day on Sunday so it should be interesting to see if there is a lack of power. I've cleared the fault code again so it will be interesting to see if it pops up after the run. I'm glad to hear it hasn't caused you any adverse effects though. I could only find one other active guy on the Seat forum who has had the issue and he said it just occasionally comes up. Like you said it must just be a gremlin in the system.

 

I'll be sure to let you know if I ever figure it out!

 

Yep. If anything, the next step is wiring straight to the ECU... then failing that, a new ECU... but as it's not knocking and (we shall see Sunday) if it's not down on power, then I'm tempted to just leave it.

 

Yeah I mentioned it to Ben and Mikko at Shark Performance while I was there and they didn't find it to be affecting the performance at all. General opinion was to not be concerned by it.

 

My car made healthy power and torque, so I'd imagine yours will be fine.

 

I had a tussle with a friend in a mapped E92 335d yesterday (test track) and he was impressed with how quick the old Skoda was and how stable it was at 130+! Nothing I could do about his mountain of torque though!

Yeah I mentioned it to Ben and Mikko at Shark Performance while I was there and they didn't find it to be affecting the performance at all. General opinion was to not be concerned by it.

 

My car made healthy power and torque, so I'd imagine yours will be fine.

 

I had a tussle with a friend in a mapped E92 335d yesterday (test track) and he was impressed with how quick the old Skoda was and how stable it was at 130+! Nothing I could do about his mountain of torque though!

 

Hiya Bodge,

So had the dyno day today. Made good performance (219bhp at the wheels and 217 ft/lbs torque). However I asked him to check the air/fuel ratio and he said it's running really badly lean! I've got a VAG-COM cable so I'm going to run some diagnostics on it tomorrow night to see what the crack is. I'm not suggesting you have the same issue... especially as you got the code several years ago... but maybe something to keep an eye on.. heck, the code might be totally unrelated anyway... I may fix the fueling issue and still have the code... not sure yet!

 

Cheers,

Adrian

Hi everybody,

I have Seat Ibiza 2000 year 1.4 mpi, it has one knock sensor, which its causing me a lot of headache.

 

On diagnostic it says : 16712 Knock Sensor 1 (G61): Signal too High

 

Symptoms are losing power and acceleration sometimes and sometimes works perfectly. I take the car to the mechanic and the problem was diagnosed by the computer in the knock sensor. I bought a new sensor and problem was still present. Then my mechanic says my injectors are dirty and need to be cleaned. But problem with the knock sensor is still here after the cleaning the injectors. I bought a new spark plugs, new cables for spark plugs, new fuel pump, new map sensor....but I didn't solve the problem. So if you please could tell me any suggestions what could be the problem with the car, what part should be changed or checked. To mention that when I disconnect knock sensor there was knocking was I assume that there is not a ECU failure because  knocking is really happening in the engine.

 

Best regards and I am looking forward for your reply