CEL and no stored codes?

[timf]

Hi newbie here with a question :-)

My daughter has a 2002 VW beetle 2.0L GLS automatic. She put some gas in it and didn't click the fuel cap down - end result was she got a CEL.

No problem I thought - just get out my scanner and clear it.
Here's the thing though - no stored codes showed, so I hit clear and the CEL went out as expected.

Problem is it came right back on when I restarted the engine again with no stored codes.

She's been running it for about a week with the CEL on - so I cleared it again, but instantly came back on with an engine re-start.

Any suggestions ? I'm thinking a trip down to autozone to use their reader in case mine is missing something.

Other than the CEL the bug seems to be running fine - it has 54,000 miles on the clock.

[D2Beetle]

Have Autozone or find someone with VCDS (aka VagCom) to scan it.

[DZLBUG]

Generic OBD scanners (including those at Autozone etc) don't always pick up everything.

VCDS would be best to see what is going on. The light might not be related to the gas cap being loose.

[buzzard217]

Generic OBII scanners are not very good with EVAP code in some platforms. VAG COM would be your best bet. Not saying that it is an EVAP code, but it does sound like it.

[timf]

Luckily I had a OBD II cable from my pocket logger which I was able to hook up to my laptop and run VAGCOM 3.11.

Found that I have a P1650 - 35 code which when cleared turns of the MIL.

This code tells me its missing a message from the instrument cluster.

Not too sure where to go from here - even though I'm an experienced electronics engineer :-) Sounds like a loose connection maybe ?

Also under readiness I found a bunch of failed / incomplete messages for the following:
O2 heating
O2 sensor
Secondary Air injection
Evap emissions
Catalyzer

Readiness 0110 1101

Engine seemed plenty warmed up and none of the above seems to be related to the MIL being lit.

Anymore guidance ?

[Smileybug]

A generic OBDII reader might not pick up all of your systems and codes. That would be my guess as to why those areas failed or were incomplete. The Vag-Com software is different from the Ross-Tech OBDII scanner. The Ross-Tech scanner works best to get all codes.

[timf]

As this started on a cold morning - I'm thinking it's the battery starting to fail - ie during cranking the battery voltage is dropping so low it's upsetting the cluster. I'll measure the voltage during cranking.

Then as this fault is 100% repeatable I'm going to start the engine with my jump start battery hooked and at the same time measure the voltage again.

I suspect the MIL will stay off as the battery voltage will be that much higher during cranking.

I'll report back later....

[timf]

So it wasn't the battery :-( But might be related to the immobilizer as my daughter reports that sometimes it starts then cuts out and she has to try again. Will pursue that line of inquiry.
In the meantime in case its just a loose connection - how do you get the instrument cluster out ?

[timf]

So after a bit more googling I decided to remove the cluster and take a close look at it.

Easy to remove - pull off the front shroud then remove two screws, release the catches on the rear plugs.

Took the rear white cover off the cluster and closely inspected the circuit board - no signs of electrical death.

So I sprayed the connectors on the wiring loom with CRC residue free contact cleaner.

Didn't want to spray the cluster itself in case it messed up the dials.

Put everything back together started the engine, cleared the code from previous start.

Turned off the engine, wait a minute or two, started up again and guess what ? No MIL, yippee !

Performed a few more starts just to make sure and it's still good.

So now I wonder was it just a loose connection or was it the interruption of power which did it ? The ODO reset to zero.

Either way looks like I've narrowed it down a little for next time - if there is a next time.

[simonsi]

Originally Posted by timf

The ODO reset to zero.

Just the trip figure presumably, not the miles-since-new figure???

[timf]

Correct just the trip mileage - miles since new remains the same, obviously must be stored in non-volatile memory.