The cps is located on the front of engine block,near gearbox bell housing.There should be a fly lead and connector attached to a bracket at the end of the cylinder head.You could take a resistance reading but as the fault is intermittant,it is unlikely to show faulty.Also if the sensor is short to earth ,it will still give a correct reading.

Found the cps. Bu66er to get out but got it.

It measured 958ohms (and yes, I'm sure the meter is accurate). Haynes says 860 +/-70, so that puts it outside spec - enough outside to be a problem?

Wiring looks okay, but I disconnected the large connectors at the Trionic and reconnected in case it's an electrical connection issue.

Yes it would seem to be intermittent and it's running fine now with no MIL (I cleared the code).

I'll see how it runs for a few days.

thanks guys for the pointers. For anyone else who might come across this thread from Google (other search engines are available), I found this which is helpful:
http://easyautodiagnostics.com/misc-index/ckp-cmp-sensor-basics-1

I considered the gap issue and was careful to put it back fully home.

I haven't tested the wiring yet (it was getting on for 11pm), just a visual. Will test later when back home.

One of the other things that has happened a couple of times is that the engine has "stuttered" whilst on the move. This has "jerked" the car quite severely on a couple of occasions. If it is an intermittent fault with the CKP sensor, then this would be a symptom.
But how to check?
From my googling I believe that the engine ecu gets its engine speed data from the cpk sensor. Do the experts agree?
This being th ecase I can use my Torque program on my phone to monitor the live rpm stream from the ecu. If the engine stalls (either at idle or whilst driving) this rpm reading should fall off too provided the cause is the cpk dropping out. Agreed?
(at this point it could still be the wiring of course)

The car stalled whilst driving on the way into work this morning (just once at about 30mph in third). So actions so far have had no effect.

If the sensor was just a fiver I'd change it, but they are 60 quid. I'm not a fan of "diagnosis by guesswork", ie changing expensive components on a suckit and see basis. So I'm trying to determine the cause as scientifically as possible (within th elimits of my test equipment and knowledge).

Interestingly, the article above suggests that ckp issues rarely throw a MIL code. This also agrees with my symptoms so far (I've only had the P0337 code once in about 15 occurances)

Your help is appreciated.

I have changed the CKP sensor and all is well again.

The big clue was that when it cranked but would not start, there was no response from the rev counter - so I could tell that the engine wasn't getting it's speed signal, which it gets from the CKP sensor. As soon as the CKP sensor kicked back in the rev counter would show cranking speed (about 100 rpm).

It always started okay when cold and started to cut out when hot, so I'm guessing there was a connection fault inside the unit which became more evident when hot. I may disect the old one to have a look.

Edited by Cyd on 22/11/2014 at 17:55