[urq] RE : What causes knock ?

[Louis-Alain Richard]

> First time I noticed it was when I got all the intake leaks sorted
out,
> and was *finally* getting some boost.

Exactly like in my case... I replaced the air filter last week and also
replaced some more rubber hoses that looked not perfectly sealed, and
pinging appeared...

> Audi has installed a 'water separator' on the small vaccum line from
the
> manifold to the ECU. It looks like a small gas line filer, with no
> innards, and I found mine up under the glovebox, near the ECU.
However, it > had cracked from old age, and there was no way any
vaccum/pressure signal > was going to make it to the ECU!

Mine looks OK. But in any case, I'll remove it and test it with the
MightyVac.


> I replaced the 'filter', and found out all of a sudden, it didn't pull
as
> well (as the ECU was now retarding the timing, as it should have been
> doing, all along!)

I my case, the story is similar has I've seen as much as 1.9Bar on the
dash gauge, a figure I've never seen before. And the car feels faster
than ever, even with 3 guys + tools aboard heading for parts hunting in
a scrapyard yesterday.

> The next summer, I heard what I though was 'pinging' again. But on
> investigation, could now find no leaks. Which lead me to surmise that
the
> ECU pressure transduce had taken a dump.  Sure enough, when I measured
the
> output from the transducer, it sat a '0' volts, no matter what
happended.   > So timing was being set to 'full vaccum' values, rater
than the '1 bar'
> atmospheric that the vaccum leak previously would put it to.


Since I still have Brady's chipped ECU, this will be an easy test: swap
the ECUs. In fact, I've reinstalled my stock ECU because I was thinking
the chipped one had some issues (long hot start, FV buzzing too hard,
inconsistent idle). Now, come to thing of it, it was probably fine, and
I have problems elsewhere...
 
> No surprise that my head gasket failed in there somewhere too!

That's my biggest fear ! 

> So, short form:
> Make sure the ECU IS seeing manifold signal, and that the ECU pressure
> transducer is working!

First part is easy, with a T and a standalone boost gauge. 
Second part, you may have to help me about this. Is this as easy as
applying boost to the ECU PT (while on the bench) and measuring the
output of the pressure transducer ? Isn't it better to look at timing
(with a timing lamp) while pressurizing the PT, engine running ?

Anyway, thanks Franck, helpful post for me.

Louis-Alain