Been quiet on the MK14 front in here - just thought you may all be interested in the fact that one of Martin's reproduction Tape Interface PCB's with a BC107 rather than a 2N2926 (hfe 120 rather than 100 so I took a chance) works well (once I found the read LED was faulty...) on my JM Precision repro.
Seems that either the SCIOS save or the load routine do not handle the last byte of a 256 byte block... Unlike a Spectrum the system is happier at a much lower volume. It is also very sensitive to any clicks or noise so important you don't press GO until after the mechanism settles, the first bit is corrupt or the whole data stream slips.
Using the Sanyo M1104X on batteries as recommended by SOC - on this repro board there is no R11 (27k) in series with D1 - I used a 1N4148 as it is not named in the manual. The clip on the resistor is just a handy place to pick up 5V and still allow the expansion connector to be removed easily. The white wire loops SIN and SENSEB. GROUND, SENSEB and FLAG0 are soldered on the connector.
