I have just completed a repair on a severely damaged IDE header in a RioCar Mk2a unit. The header was originally soldered on at a tilt, such that one row of 22 pins was not completely flat on its pads -- the typical immediate cause of header failure on Empegs.

But this situation was worse than normal, in that repeated flexing of the header has resulted in several pads (the tissue-thin copper traces) beging completely lifted from the main board -- impossible to reattach. This necessitated running jumper wires from the header pins to locations on the underside of the board to reconnect the broken signals. Four (or was it five?) of those in total. Ugly. Fortunately, they were all "data lines", which are easy to test/diagnose with a customized Hijack kernel.

Having completed this repair, it is obvious to me that the weakened header (having four pins that are not soldered to anything) will in the future continue to flex more and more, and eventually fail again.

Why does it flex? Heat expansion? Sure, a little.

But the main cause of IDE header failure (apart from poor soldering to begin with) is the shock-mounted drive tray. Which moves constantly in response to vibrations in the car, or simply carrying the player by its handle, or whatever.

As the tray floats around, the cable connecting it to the mainboard is pushed and pulled upon, transfering forces to the IDE header. Which will eventually fail as a result.

A large culprit in this scenario is the very short (and therefore stiff) length of ribbon cable between the nearest drive bay and the header -- on a stock Empeg cable this is a very straight and short length, with no allowance for flex.

Sooo.. GET TO THE POINT, MARK!!

Okay, my point is this:

(1) in a single-drive player, connect the single drive to the END of the IDE cable, rather than to the middle connector. This will extend the life of the IDE header considerably, as well as reduce ringing on the signals.

(2) in a dual-drive player, a new cable is needed: a cable with enough extra length (2" or 5cm extra) between the header and the first drive connector, so that the cable can extend up beyond the drive tray, and then be looped back down to connect to that first drive. NOT FOLDED down, but rather LOOPED down. So that it can flex in ths extra loop of cable rather than transfering all movement down to the fragile IDE header.

Soo... Stu, are you listening? You are now the primary source for replacement IDE cables for these players. Yes, a few of us (like me) can make our own cables, but everyone else on this side of the pond (and others outside of N.A.) seems to be sourcing their replacement cables through Eutronix now.

These replacement cables need to be constructed with an extra two inches (5cm) of ribbon cable between the header and the first drive connector, so that a nice gentle LOOP of cable can be left there to absorb some of the drive tray movement.

I'm thinking of offering these enhanced cables for sale here, but frankly you are better positioned for most folks.

Go for it!

Cheers