Thanks for taking the time to reply again Ivanerrol. I tried a tank of 97 the other day - I was sure I could see the fuel gauge move with a distinctive cartoon-style sucking sound as I was driving! After checking the mileage it was only down to 19 from the usual 21MPG, give-or-take errors in my measuring.

My son's child minder is next door to an copy-parts supplier and considering the long weekend (Happy Easter folks!) I quickly went in and ordered one. If it doesn't work then it wasn't a great loss cost-wise and I'll go to the OEM dealership.

I re-adjusted my idle to get it back to normal, or as it was before, and then took off my air filter and poked around the various CIS sensors (the ones I could locate and get access to) using my trusty multimeter.

The engine temperature sensor is one of those with four-connectors and from reading the forums the the pins are diagonal and the resistance equated to the dashboard temperature reading (within reason) or both pole-pairs. There was also a single-wire temperature sensor ahead of that and seemed to show resistance too.

The air temperature sensor I didn't have a guide for however on testing it outside the car and moving it into the hot-air flow from the radiator fan it showed functionality.

The throttle position sensor (perhaps called something else on the Benz?) showed variation when moved.

At the same time I tested the X11 socket for the ECU's fault code. Since this is a non Lambda car / open-loop car the ECU presents a value representing an error code. This is steady at 50% normally but did show 70% with the engine temperature sensor off, 20% with the air temperature sensor off and either 60% or 90% value (can't recall which just now) with the throttle position sensor off.

Now I was puzzled by these values, since another article stated that failures in these sensors should generate 30%, 80% and either 10% or 40% respectively. Then I realised that these figures, when inverted, worked out right: 100% - 20% = 80%. This may be down to how my multimeter works, but it left me scratching my head for a bit!

Clearly this shows that the ECU sensors are not 'failed' and the ones I tested with references seem to be within tolerances.

Still with the air filter off I checked other components and found a switch on the throttle linkage. It never seemed to engage under normal usage but if I pushed it in by hand the revs would drop down a little bit. Should I somehow 'modify' (read: bend) the switch actuator with the roller to make better contact so it's activated normally? Or is it movable to reposition it?

Furthermore I disconnected a device on the front-side of the fuel distributor and the revs dropped right down from around 8000rpm to 6000rpm. I'm guessing this is the extra air enrichment device, the technical name of which I cannot recall. The engine seemed slightly more responsive when driving with this left off and, although I cannot determine until my next fill-up, the fuel gauge seemed to go down slower then usual.

Would I be right in thinking that this along with the cold-start injector are temperature dependent and therefore the current guess at the thermostat issue would be resolved next week upon fitting the new one? I know I'll find out when I fit it, but as is typical when you're keen to fix something a raft of bank holidays get in the way! (the workaholic in me can't cope with such long breaks off )

I shall keep you updated with progress.

Kind regards,

Glyn B.