== Technical ==
CPCAI magazine claims the device has a WD1010 Winchester controller inside. The ROM It certainly seems to imply this too although the be WD1010's A2, A1,A0 are not mapped to A2,A1,A0 on CPCcompatible.
The WD1010 has different commands and communication method compared to the Dobbertin (which appears to use SASI) and modern IDE (which uses ATA). It looks to be an ancester of the IDE because of the seperate cylinder and sector registers and a single command write to initiate commands but it's not the same as IDE/ATA.
Therefore the software & hardware is *not* compatible with the Dobbertin or modern IDE drivesother interfaces.
Port decoding appears to be:
bit 1 = wd1010 a1
bit 0 = wd1010 a0
bit 8 is 1 or 0, maybe it can't be either, but if it is set it could clash
with other i/o. The ports chosen here don't clash.
bit 10 = 0
bit 2 = 0
are the Except for bit 8, other bits checked for are 1. It is not known which bits are decoded by the interace. Bit 8 is 1 or 0, maybe it can't be either, but if it is set it could clashwith other i/o. Maybe the ports chosen here don's?t clash. Ports used and their function:
FBFA - wd1010 sdh register (read/write)
FAF8 = wd1010 cylinder low register (read/write)
FAF9 = wd1010 cylinder high register (read/write)
FAFA = drive /head select? drive bits match those in sdh. (write) wd1010 documentation says head select/drive select not available as outputs. Furthur details:* MFM
* 256 bytes per sector.
* 32 sectors per cylinder