Gggrrrrr. TDI conversion problems.

[Baxter]

How do.
Not been here for a while.

I have a problem with a customers car that has be scratching my head.

It is a Vanagon with a AFN in it. [I did notdo the conversion, but it is quite tidy and according to the owner has been reliable]

It came as a none runner having broken it's oil feed pipe for the turbo on the motorway, oil everywhere, customer stopped and had the van recovered to me.
We've had it here and corrected the oil line, we fired it up, it idled and didn't leak and we didn't think much of it and moved on to get on with the rest of the work it needed.

Came to take it off the ramp and the throttle pedal has no effect.
Great, has a look and see if we have dislodged anything while we've changed the pipe, nothing obvious.
Plugs VagCom in and we get "00777" throttle position sensor - implausable signal", tidy, sensor has failled we'll try a known good one, no go, still kapput.

So, heres what happens.
Van will start and run, albeit at idle.
You can gently rev the engine till about 1200rpm okay, over this threshold it will go into some form of limp mode where idle speed is increased to about 1000rpm and the accelerator pedal now becomes inoperative.
Now, if you are in VagCom, measuring block group 2 field 3 there is a set of binary figures, the first I believe is the AC raised idle signal, the second is the idle switch and unsure of the 3rd.
When this problem happens then the fields got to "110" rather than the usual "010" setting it should have at idle.[van doesn't have A/C]
While the problem is happening you can move the throttle and you will get the load going from 0% to 80% no problem so the throttle signal is making it's way to the ECU, just the pedal becomes innefective.

Once the "110" is achieved, then the ECU again registers the fault "00777", the fault will not clear while the engine is running at this raided idle speed.

Now, weird bit.
if you start with the engine off, and the fault cleared and hold the throttle to the floor and then start the van it will rev normally, you can drive the van so long as you do not drop it to idle, once you get to idle it then returns to it's limp state and again the throttle pedal now has no effect and the fault code is lodged.

I have tested the sensor, it appears to check out using test data from ELSAWIN, I have tried and alternative sensor with no effect.

If you read the test data, it says that if a problem exsists with the sensor it will default to a raised idle speed, which is I presume what is happening.
I'm wondering if it does this by internally turning on the increased idle speed via the air con request and thats why I'm seeing "110" ??

Dunno.

Tested continuity between the sensor wiring and the ECU - no fault found.

Wouldn't normally post this here but I have 24 hours to fix it!! 

Oh, customer says he scanned for fault codes a little while ago and did see the throttle position fault code a while ago but it hasn't caused any problems.

Any ideas?
.

[Baxter]

Further to this.
Continuity between ECU and sender - good.
Wiring checked for short to each other, short to earth and short to positive - all good.

[the caveman]

I had similar problems with 2 VWs ,one a Beetle ,the other a Passat, around 2002, albeit gassers, but still drive by wire. Both had bad grounds.
But the signal you are getting in the measuring block sounds like shorted wires or 2 mixed up connections . Real question is whether it's a coincidence or something done while doing the sump repair.
I would also do a ECU memory kill by disconnecting the battery and touching the two big cables together for a minute or two

[Baxter]

With it being a conversion, and a conversion done by some respected boffins and the fact that this van really has been around the world I'm tempted to think that it has been alright most of the time so more inclined to think that it is either something that has failed or something we have "nudged" while doing the oil feed pipe, but I've been over it and checked all the electrical connections in the vicinity and all are good.
I have repaired a few things on the way, oil pressure sender not plugged in, alternator wiring damaged etc but nothing has had any effect.

I couldn't get a wiring diagram from ELSA for the base vehicle, but I did get one from autodata and found the wires to the ECU for the air con, and duly snipped them, no effect.

[the caveman]

Do you indeed have a raised idle speed while the fault is there ? If not i would still suspect a bad ground or ECU failure. I doubt the ECU failure , but ...
If you do have raised idle speed then the ECU is almost definitely seeing a failed throttle issue.
I'm sure that you know that most of the connectors at the ECU are ridiculously small and can develop connection problems due to voltage drops across such small surfaces . maybe try a wiggle test at the ECU connector while running

[Baxter]

Do you indeed have a raised idle speed while the fault is there ? If not i would still suspect a bad ground or ECU failure. I doubt the ECU failure , but ...
If you do have raised idle speed then the ECU is almost definitely seeing a failed throttle issue.
I'm sure that you know that most of the connectors at the ECU are ridiculously small and can develop connection problems due to voltage drops across such small surfaces . maybe try a wiggle test at the ECU connector while running

Yeah, raised idle as soon as the TPS fault occurs.
I'm also tempted to think it's an earthing issue, but with this being a conversion someone else has done then tracing earths is a pain in the ring piece!
Customer has been delayed a few days to give me some breathing space so when I get some more time then I'm going to check lives and eaths at the ECU.
All connectors at the ECU seem okay, and it's been plugged and unplugged a few times so any eroneous problems caused by stale contacts should be remedied already.

[vanagonturbo]

Have you checked to see if the ecu is getting a signal from the brake light switch telling it the brake is applied?

[the caveman]

Have you checked to see if the ecu is getting a signal from the brake light switch telling it the brake is applied?

very good point

[Baxter]

Mini breakthrough.

just noticed that when the engine is revved the idle switch reading in VagCom says the contacts are closed.
I've read somewhere that the idle switch contacts should be closed under 17% throttle, and open above 17% throttle.
I'm slowly pressing the throttle, get to 17%, then the air con raised idle speed reading meationed above comes on, but the idle switch contacts appear to stay closed.
if the idle switch contacts are staying closed and the throttle is above 17% then the ECU is thinking theres a problem and putting it into emergency mode with the ineffective throttle and raised idle speed.
I've tested the switch (yet again) and the switch contacts work properly, but the ECU isn't seeing the switch open so now I'm thinking that there is an internal fault in the ECU.

What VW say...

Pedal position determined by G79
A decisive factor for the injection quantity is the accelerator
pedal position, i.e. the driver input.
The accelerator position sender is a sliding contact potentiometer
and includes an idling switch and a kick-down
switch (refer to Function Diagram). From these signals,
the control unit calculates the necessary fuel quantity
using additional parameters.
Substitute function
If a fault occurs, the engine runs at a higher idling speed
so that the customer can reach the next workshop.
The accelerator position sender is then deactivated.

Theres a spare ECU here, part number however is different.

Part number for the working ECU is 028 906 021 GH.

Part number of the spare ECU in the back of the van is 028 906 021 DF

Anyone know if they are intercahngeable?

Earths checked at ECU connector, all good.

Also, looked on RossTechs site about swapping ECU's, and also looked in ELSAWIN. Apparently I'm looking for a 4 digit code? where RossTech say it is, it isn't. Any ideas on interchangeability of ECU's? tips on swapping?

[Baxter]

Fixed.

I snipped the idle switch wire at the ecu and the reading in VagCom went to zero, result [ish].

Van then run constantly in emergency mode as it had detected the open circuit on the idle switch.

While it was snipped I soldered a fresh wire from the sender direct to the ECU and it ran properly.

thank f*ck for that.

Why it din't show up when I tested the continuity of all the wires I cannot explain, probably me being p*ssed off and not noticing it as I've been back and tested the wire in question and there is indeed a high resistance down it.

While the meter was on it we gave it a good pull about and it didn't alter it.

So, wether we caused this or not i can't say, I'm just glad it's fixed.

Roadtest tomorrow [taking tow rope with me!!]

[8v-of-fury]

Congrats

Nothing like working on someone elses bastard child of a swap

GL on the test drive!

[R.O.R-2.0]

Congrats

Nothing like working on someone elses bastard child of a swap

GL on the test drive!

i absolutely will not work on other peoples abortions.. i dont know any of what they did (or did not) do..