Shaft encoder in HP2631B printer

This is a long shot... On the other hand, there are people here who've done all sorts of odd things, and the only silly question is one that's not asked (or something like that). As I mentioned yesterday (when I asked about zneer diode data), I've got an HP2631B printer that's gone crazy. I had anohter quick look at it today, and it appears that one problem is the shaft encoder on the right hand end of the carriage leadscrew. This connects by a 10 way cable (integral with the encoder) to the 'printer logic' PCB, 6 of the pins on the connector sseem to be used, 2 grouds, 2 +5V and the 2 quadrature output signals. Those 2 outputs are buffered by a couple of sections of a '14, the outputs of that go to testpoints (and elswehere, of course). Anyway, both those testpoints are low all the time (turning the leadscrew by hand, I've not fixed the motor drive amplifer yet!). The outputs from the encoder are high all the time (so the problem isn't the '14, more's the pity), they are genuinely high, not floating. And there don't seem to be any pull-ups on the printer logic PCB< so the 'high' is coming from the encoder. Anyway. Iv'e read the HP service manual (on http://www.hpmuseum.net) and it appears you need special tools to remove and replace the encoder. The encoder disk is epoxied to the neadscrew, you need a special puller to break the bond and an alignment tool to get the encoder disk the right distance insde the encoder housing when you refit it. Of coruse I don;t ahve the tools... Has anyone ever done this? Maybe the same tools, or something like them, were used with THe shaft encoders that HP sold as loose componets. Any thoughts? -tony