That relay supplies current to the ECU and associated sensors/switches via the adjacent fuses.

Hello all - problem appears to have been solved. I bought a new relay (part no 1K0 906 381, numbered '458' on the top, price about 14 uk pounds) which goes into the R4 position in the fusebox in the engine compartment. I've been doing simple substitution tests. When the original relay is warm, there are no problems but when it is cooled to about 14C, the brake light nonsense occurs. Not so with the new relay. A very noticeable difference is that, with the old (cold) relay, switching off the ignition leads to silence from the engine area. With the new relay, there is a click, a whine (presumably of a solenoid) and a loud click. Anyway, the brake lights don't come on with the new relay - cold or warm, and starting is no longer a problem. At the start of all this, there was an exhaust control warning light which seems to have been independent of the starting/brake light problem, and has come and gone in a seemingly random fashion. With the new relay in place there has, so far, been no sign of the warning light.

Incidentally, in case the problem was the fuse/relay baseplate, I enquired about its cost. I was agreeably surprised - one hundred and thirty eight uk pounds - A replacement circuit board for our German cooking oven cost over three hundred!

So, huge relief and my grateful thanks to all you kind folks who have gone the extra mile to help. In the end it has been a combination of thinking and intuition rather than trying to interpret the vagcom codes. I *will* take the car back to the repairer, who has the test gear, and it will be interesting to see if some of the several fault codes they found have now disappeared.

Hello all - problem appears to have been solved. I bought a new relay (part no 1K0 906 381, numbered '458' on the top, price about 14 uk pounds) which goes into the R4 position in the fusebox in the engine compartment. I've been doing simple substitution tests. When the original relay is warm, there are no problems but when it is cooled to about 14C, the brake light nonsense occurs. Not so with the new relay. A very noticeable difference is that, with the old (cold) relay, switching off the ignition leads to silence from the engine area. With the new relay, there is a click, a whine (presumably of a solenoid) and a loud click. Anyway, the brake lights don't come on with the new relay - cold or warm, and starting is no longer a problem. At the start of all this, there was an exhaust control warning light which seems to have been independent of the starting/brake light problem, and has come and gone in a seemingly random fashion. With the new relay in place there has, so far, been no sign of the warning light.

Incidentally, in case the problem was the fuse/relay baseplate, I enquired about its cost. I was agreeably surprised - one hundred and thirty eight uk pounds - A replacement circuit board for our German cooking oven cost over three hundred!

So, huge relief and my grateful thanks to all you kind folks who have gone the extra mile to help. In the end it has been a combination of thinking and intuition rather than trying to interpret the vagcom codes. I *will* take the car back to the repairer, who has the test gear, and it will be interesting to see if some of the several fault codes they found have now disappeared.

A little more:

The new relay has definitely sorted the problem. No spurious brake lights, no starting problems and no exhaust gas control warning light. I opened up the old relay and found that the relay contacts were badly pitted, as one would expect from an arcing contact, leading to the lights flickering. Perhaps the transfer of material across the contacts led to there being a permanent (but sparky) connection between them. What remains is why this happened in the first place and why a warm relay was less likely to be troublesome than a cold one. Looking at the detail of the relay (after removing the cap), it is easy to understand how a warm relay would suffer the problem (because of copper expansion) but difficult to see why the converse happened.

The final question, now, is why this fault occurred - is there a gremlin further up the 'chain of command' which will repeat the problem in due course. Because it appears to be impossible, without infringing copyright, to get a copy of the wiring diagram, I will have to leave it there and deal with the situation by buying another, spare, relay in case it happens again. I am not happy that it is so difficult to get a wiring diagram but that is the way with manufacturers, these days, I fear. When will Haynes do an Octy II manual, I wonder?

Faults with ECU power relays are very common on VAG cars (although I haven't seen flashing brake lights when one goes before). On Mk.1 diesels, a faulty relay 109 is very common.

Faults with ECU power relays are very common on VAG cars (although I haven't seen flashing brake lights when one goes before). On Mk.1 diesels, a faulty relay 109 is very common.

Wouldn't you know it - my fault *is* in the web forum records. Knowing what I know now, I googled 'VW Skoda relay 458' and found at:

http://www.ilexa.co.uk/forum/index.php?topic=25377.0

the very fault we have been discussing, but in a Golf 2.0. It's perfectly true that as long as you know what to plug into Google, it's out there somewhere!

I'm at a loss to understand what led this mechanic to remove the 458 relay - his explanation seemed a little thin in view of the forest of code information he was wading through but there is no substitute for a little intuition at times. Unfortunately, the mechanic doesn't appear to have searched for a cause for the relay's failure.

Maybe it is as you suggest - just an unprovoked failure in a type of part which commonly fails on VAG cars.