Wednesday, 14 August 2013


My list of jobs to do to completely finish the 120 seems to alternate between forty and fifty items.  Just as soon as I knock one off the list, another appears.  This of course is the result of me taking every opportunity to drive the car rather than concentrating on tying up the myriad of loose ends. The other thing responsible for such slow progress is the long overdue need to spend some time on the 140 fixed head and Frogeye Sprite.
Amongst other niggles, the 120's water temperature gauge has stopped working.  Its the type with a copper capillary tube from the gauge to the radiator and is a specialist repair job.  As it's also the oil pressure gauge I can't really do without it so I'm on the lookout for a replacement to send off for reconditioning.  I think the capillary tube for the XK is a bit longer than the standard item so it needs to be a gauge specifically for a 120.  As I really need to know what my engine temperature is, especially at this early stage of running in, I temporarily fit a modern gauge.  Interestingly, it agrees exactly with the typical running temperature of 80 degrees recorded by the original before it stopped working.

Temporary Water temperature gauge attached with Velcro

This is considerably cooler than the 140 when it was running in, which occasionally went off the clock.  Also, it doesn't have anything like the 140's heat soak problem which leads me to wonder if the 140 engines reconditioning done a few years back by VSE actually included a very thorough block clean. It's only around six weeks now to the time when the 120 will be taken off the road for the winter and then all these outstanding jobs can be attended too.  Now I know that the original cooling system is working to spec, I'll also finish off the conversion to the electric water pump over the winter.  This will no doubt generate a post in it's own right.

The 120 motor continues to run superbly, very quite with loads of torque.  The only criticism would be a very light tappet noise which is intermittent and cyclic, most noticeable at around 75 degrees. At 700RPM it will go from silent to ticking and back to silent every fifty seconds or so.  I can only guess that its an out of true bucket which changes the valve clearance by a few thou as it is rotated by it's cam lobe.

I was initially concerned about oil pressure which was between 20 and 50 PSI hot (85 degrees).  This is around 20PSI lower than the 140, but as its remained constant I have now put this down to the earlier gear type pump.  Every expert I've consulted to date tells me I've nothing to worry about and it's fairly typical for an early 120 so it probably is OK.  My initial concern over this led me to have a full oil analysis test done at 100 miles.  This involves sending off a small phial of oil to a lab where they look for any traces of stuff that shouldn't be there.  Tin, lead, Iron, silicone etc. recorded in parts per million; it all adds up to give an indication of whats happening inside.  As my main concern was oil pressure, I thought the oil may have been contaminated with fuel, reducing it's viscosity; possibly caused over rich running and plug fouling problems, due to the otter switch holding the starting carb on longer than necessary (resolved with a manual over-ride switch).   The oil analysis showed nothing unusual, no trace of fuel and the viscosity was also correct for the Millers straight SAE30 running in oil.  For complete peace of mind, I will however invest in another analysis 500 miles after the next oil change.

You may recall in previous posts that I had twice checked speedo accuracy with someone following and it was hugely incorrect reading approximately 20% less than it should.  It also seems that the mileometer is incorrect by a similar percentage.  Is this something to do with the drive from the new five speed box?  My 140 speedo and mileometer are unbelievably accurate (same box but from Classic Jaguar in Texas) so it's a bit of a puzzle.  An additional puzzle is that at 2500 RPM the cars following, on both occasions recorded 80MPH, 9MPH more than my calculations indicated if fifth gear ratio is as I thought - 0.79

Back in May I had attempted to connect  the rev counter drive but it didn't feel right.  Fortunately I checked that the input to the rev counter was free but it was actually seized solid in it's bearing.  I would guess that whatever lubrication it had, had changed it's consistency to a sort of varnish, gluing the surfaces firmly together.  A little heat and some WD40 soon had them apart and after a thorough clean and light oil all back together as good as new. 

Rev Counter input shaft was seized solid in its bearing

Before checking actual speed with a Sat-Nav, I checked the rev counters accuracy against my Gunson test set and at 1000RPM and 2500RPM it was absolutely spot on.  The Sat-Nav then recorded 77MPH at 2500RPM.  This would indicate that the cars following previously were both 3MPH optimistic which is believable and interestingly, the same error I see in my Audi.  To achieve 77MPH I would need a 5th gear ratio of around 0.73.  An optional ratio for the Tremec box is 0.72, so I can't help wondering if that's what I've got.  In any event, it's an ideal 5th gear so whatever the reason, I'm more than happy.

Back in the thirties, the quality and smoothness of a Rolls Royce engine was often demonstrated by standing a coin on its edge, on the top of the engine.  Yesterday a friend showed me a Seismograph app he had just down loaded onto his Ipad.  I made the point that there's not a lot of call for detecting earth quakes in Yorkshire but you can no doubt guess where I'm going with this.
My feeling is that because of the great care I took building and balancing the 120 motor, it feels smoother than the 140.  Is it really, and also how does it compare to say the 4.2 V8 in my Audi.  Could this free Seismograph app be the ideal instrument to record exactly how smooth an engine is?

XK140  700RPM

XK120  700RPM

Audi 4.2 V8  850RPM

Difficult to say really.  The 120 trace seems the most regular but I had expected the Audi to be far smoother.  In this mode the seismograph is only showing the up / down acceleration (actual movement of the engine) which I thought most relevant.
Probably more interesting is the difference between what it would feel like sat on the engine rather than the seat.  The above two traces were taken with a different much more expensive (69p) seismograph app which simultaneously displays all three dimensions of movement.

three dimensions of movement of engine

and on seat cushion
Needs some better analytical skills I think before it becomes meaningful, but nevertheless I would think some clever lads could come up with an automotive application.

Of course, should fracking become commonplace in this 'desolate' North East part of the UK, then according to those green guys, we'll be detecting proper earth quakes left right and centre - I don't think so.

Next Post early September

Thursday, 8 August 2013


A whole series of events, family happenings, holidays and deadlines for various publications have conspired to slow down this post.  Consequently it's going to be a couple of weeks late, but will definitely appear by latest Thursday 15th August.  My apologies if you've spent time checking if its up.