15 replies to this topic

[jdscott]

Afternoon,

 

Hope someone can help.

 

Bit of a long one as I've tried diagnosing and included that below.  I've now run-out of diagnostics I can think of.  I think its also slightly strange how its occurred.

 

Car works perfectly, been for a good drive a few days before, I then washed the car. nothing different.  The next day I went to start the car, car starts as normal but get an EML light.  Check with Scan gauge and get P1275 which is Acc Pedal sensor circuit 1 and P1635 which is 5v reference circuit. 

I check both LOD connectors and visible wires along with ECU plugs and pedal plug, all good, clean and dry no breaks etc, I spray ACF50 in the sockets just to be sure.  Try again and still the same codes, I can clear the codes but the engine nearly dies when done whilst running and the codes come straight back.  I can confirm using the pedal does nothing.

 

As the codes are pedal related, I checked the voltage at the pedal I determine that wires 2 (grn) and 5 (wht) are the reference voltages and should be 5v, (let me know if this is wrong) Wire 5 gets 5v, wire 2 get just under 2v. So wire 2 is the reference voltage for circuit 1, (again let me know if incorrect).

I de-pin wire 2 from the black LOD plug, I get just under 2v.  Before stripping this wire out back to the ECU I check the voltage at the ECU by carefully dismantling the ECU plug and checking straight off the ECU plug pin, I get 2v.  I can't check directly on the ECU pin as I can't see an easy way of de-pinning.

 

Has anyone any suggestions on what else I can check?  Can't see on the wiring diagrams anything that's linked, but suppose the ECU might be looking for something from elsewhere.

 

I can't think how a working car that has only been washed does this when the wiring is not damaged or has water ingress.

 

Thanks 

 

James 

 

 

 

 

 

 

[siztenboots]

check the multiplug on the back of the expansion bottle bracket

 

[Johnboyhgt]

So basically there are two potentiometers in the pedal.

So
2x 5v feeds
2x Earth's (though via the ecu)
2x signal wires.

Best thing to do is move the pedal and make sure the voltage doesn't change, if it doesn't then it is the 5v feed.

All our ecu's are now 20yrs old so failure is inevitable occasionally.



Sent from my SM-G903F using Tapatalk

[Johnboyhgt]

Also have you checked for corrosion in the ecu plug or connector ?

Sent from my SM-G903F using Tapatalk

[jdscott]

Siztenboots,

 

Thanks for the reply.

 

That's the ones I've checked, (aren't they the LOD connectors?), x2 and x3, blue and black plugs.  Took a bit of getting to the first time, but can now get it done in a few minutes.  No need to remove clam, or disconnect header tank.

 

Are you saying it could be something seemingly unrelated?  All wires visible and the connectors look good.

 

Forgot to say, I also checked the coilpack, all dry and clean, including connector.  And the TB connector, again all dry and clean.

 

Can't believe the ECU is damaged, nothing seemingly done to damage it.  The engine starts and runs perfectly, instruments work, i.e. rev counter.  Just doesn't seem to output 5v on the correct wire.  

 

The ECU must produce individual 5v supplies, thought it would do it centrally and then split.

 

James 

 

 

[jdscott]

Johnboyhgt,

 

ECU plugs checked, no corrosion, all dry and clean.

 

Didn't check the voltage didn't move, not easy to access pins whilst connected to pedal.  But the wires are 

Circuit 1  wires 1

                         2 this gave 2v

                         3

Circuit 2  wires  4

                          5  this gives 5v

                          6

 

Can't remember the values from the other wires, but from the wiring diagram if wire 5 for circuit 2 is the 5v feed then 2 must be for circuit 1.  

When I checked the voltage from wire 2 at the ECU it was disconnected at the black plug, so no feedback voltage could be produced, so it must be the reference voltage.

 

The signal wires work between 0.5 and 4.5v from research so to get 5v should be impossible. 

 

Also the pedal plug was removed when I checked the pins in the plug to determine the 5v reference, so again no voltage should be on any of the wires except the reference.

 

Only 17.5 yrs old 

 

James

[Johnboyhgt]

Signal wires aren't exactly 0.5 and 4.5v. If you look closely at the diagram there's a resistor in the pedal which changes the output signal voltage.

Regardless you should still have 2x5v references.


Sent from my SM-G903F using Tapatalk

[Johnboyhgt]

Ps the 5v is an input not output.

Sent from my SM-G903F using Tapatalk

[jdscott]

Johnboyhgt,

 

As you say the signal may not be exactly 0.5 or 4.5v, but they should be the bounds of the signal.  The signal voltage changes with pedal position.

 

The 5v must be an output from the ECU, has to be, the pedal cannot produce it.  As far as I know, the pedal just alters this voltage, via a potentiometer and returns a different voltage to the ECU.  The pedal has two potentiometers which are inverse to each other, i.e. one 0.5-4.5v and the other 4.5-0.5v.   The ECU then does some checks and balances and confirms or not that everything is ok. In my case its not happy.  Not sure why one has a resistor.  

 

I've concentrated on the reference voltage as I'm convinced the other error code is a symptom of circuit one not getting 5v, so returning a voltage that is outside of parameters  

 

Will triple check the wires tomorrow, but I'm sure I have the 5v references identified.  Although for my logic to work the 2 pedal potentiometers need to be inverse, as far as I know they are.    

 

Thanks for the input.  I'm sure its something simply.

 

James

 

 

[ayresyy]

   I have seen water managing to ingress past the seal inside the throttle body cover before, causing a similar issue.

[FLD]

Also worth checking the tension in the terminals of the LOD connectors.  They are a pin and spring type and the spring gets weak.  The wires aren't supported either side and they constantly flex the connectors loading the terminals.  If you open the plug and remove the yellow mesh you can look in and see the terminals.  Check the blade part for corrosion.  On the female part check it for corrosion and also check that the springy part is right across inside the square hole.  If they get pushed back you will have a weak connection. I've seen a couple now with this problem.  You can bend the spring back into place with a pin but this fix lasts 6-12m.  Terminals are a few pence to replace!  From my experience the worst culprit for this is the large pink/brown wire on the corner. 

[jdscott]

Thanks for the replies.

 

I'll work through them at the weekend, even if already done before.

 

As the error seems to be 5v reference related, does anyone have or know where there is a list of all the 5v references?

 

Thanks 

 

James 

[blackoctagon]

I can't help better than has already been done, but I can tell you how to re-program a new ECU if thats where you end up as I keep a few N/A GMPT15 handy for emergencies and playing with.

Edited by blackoctagon, 09 January 2021 - 05:35 PM.

[jdscott]

Bit of an update, no further in solving though.

 

With the engine running I get 

Circuit 1  wires 1

                         2 this gave 2v

                         3

Circuit 2  wires  4

                          5  this gives 6.7v

                          6  this gives 4.8v

 

These are taken at the pedal plug, plug removed from pedal.  I have previously checked the pinout that supplies 2 above and get 2v direct at the ECU.

 

Anyone experience of these https://www.obd2shop...on-package.html  obviously not OEM.    Please don't suggest Op-Com, I don't do Microsoft, also everyone that has Op-Com seems to have issues with it not working.

 

Anyone any more ideas?

 

Blackoctagon, If needed would you be willing to lend me an ECU so I could check it's not that?

 

thanks 

 

James 

[jdscott]

Just to close this off for anyone who looks in the future.  The issue has been resolved by disconnecting the EGR cheater.  Will be replacing it with a new one and hope the issue is solved.

 

 

[blackoctagon]

jdscott,
I'm sorry i missed your request for a borrowed ECU - i've just seen your thread, but i'm glad it was an easy fix.

I can tell you how to remove the EGR in the ECU though, but you need to be able to read the code. I can explain if you wish.