I'm now ready to write my report... and it's going to be a long one, so draw up a chair and have a beer or two to hand.
I left a new phial of SME FD2013 fluid to dribble into the cleaned damper overnight, and then spent this morning going through every setting of the Series III arm for the first time since I installed it in June 1979! All settings were found to be precisely to SME specs other than the bias compensation which was lower than their Instructions advised. Co-incidentally, Hyperion told me yesterday that it had been recommended by a "reliable" AK member to set the compensation on the Series III arm to around half the figure specified by SME as they'd got it too high. I can now confirm that finding.
My next step was to put on Hi-Fi Sound test record HFS75 and test everything that it covers other than Wow & Flutter (they are measured by metering rather than listening). I tried reducing the cartridge tracking from its existing 1.75 gram to 1.5 grams (the range for the Signet MKIII is 1 - 2 gram) but decided that 1.75 gave just
very slightly better results on a particularly difficult track of the test record (almost certainly no difference at all in most real-life listening).
Using the 'Determination of Tracking Pressure and Bias Correction' tracks on the test record, the changes I made to tracking pressure were each accompanied by experiments with bias correction, and I concluded that the ideal for the 1.75 gram tracking that I'm staying with was SME's 1.0 setting (this being the bias correction position on the arm intended to accompany a 1 gram tracking pressure). I'd had the bias correction set at the 1.2 position for almost 38 years and forgotten about it until now, however I recall arriving at it in conjunction with that very same HFS75 test record. The differences between the 1.0 and 1.2 positions were infinitely small, but 1.0 seemed a tiny bit better for the difficult inner track (see below), so that's where it is now set.
There are three tracking pressure/bias correction test positions on the record - for outer, centre and inner groove positions. The outer position carries three bands which are progressively more difficult to track. Bands A and B were handled easily, whilst the 'super high level' Band C was also handled but with some protest (the test record sleeve carries a warning that "Band C should be attempted
only with the very finest cartridges and arms... and can cause amplifier input overload". The centre test position on the record has a single track, which was handled with ease. The inner position also has a single track, and this was handled adequately but not with the absolute smoothness of the centre band and the outer bands A & B.
At last I could get to the MC versus MM+SUT back-to-back tests.
First I played the opening to Mike Oldfield's 'Tubular Bells' - over and over again, swapping inputs and MC/MM selector, levelling the volume as quickly as I could, and then listening. No conclusion reached. I moved onto other genres, including Direct Cut jazz, classical with and without singing, and easy listening. Finally, and appropriately, I played several tracks from the treasured 'This is the Sansui Sound' LP that came with the amplifier.
There wasn't much at all to choose between the two input modes. Some tracks seemed better in one mode at low volume but in the other at high volume, and the genre of the music (smooth or not so smooth) also seemed to affect the results. Both modes were surprisingly close to each other, even the output volumes were almost identical at the same position of the control. Decisions, decisions!
I eventually decided that feeding the MM input via the step-up transformer was the winner, but only by a nose. The main reason for this choice is that the attack was more realistic when playing high volume, high impact, jazz and orchestral pieces (like the cannons in Tchaikovsky's '1812 Overture'). Putting it another way, the MM+SUT combination gave a better representation of what the AU-919 is about. But only just!
Huge thanks are due to John (Hyperion) for making the comparison possible, and I hope he's not disappointed that I won't be using the modification that he put thought and effort into.