[Coco] ANNOUNCE: Cloud-9 Website Update

[Mark McDougall]

On 26/08/2011 5:30 PM, John Kent wrote:

> I'd use SD cards if I knew a bit more about how the internal registers
> worked. The description of the SD card registers from my reading some years
> ago used register names that I was unfamiliar with and seemed complicated
> than the IDE interface. I couldn't find a good description of how they
> worked. SD cards seemed to support a number of different serial protocols
> such as SPI and I2C and also nybble protocols.

The standard was closed for a long time - IIUC it is now open. The simplest 
mode - SPI - was always open, and relatively simple, though it only allowed 
relatively slow transfer rates. I used to bit-bang SPI to SD in software on 
Altera's NIOS embedded soft core when throughput wasn't an issue.

> or I2C interface chip to interface to the CoCo, or else bit back an I/O
> port.

I did a front-end interface for SD that looks like an IDE drive; this way I 
was able to use the opencores IDE core for both CF and SD (DE1) when I 
emulated the Glenside/SuperIDE controller for my Coco1FPGA. The (read-only) 
SD core itself was done by a colleague - 4-wire IIRC. It runs HDBDOS with 
either media. Would be relatively simple to integrate into a CPLD and have 
an 'IDE' interface for the Coco that used SD. I really need to get 
write-mode going on SD; CF already writes of course.

> I'm not sure what is required in terms of development
> software to use Rocket I/O. You can get SATA to IDE adapters which might be
> an easier way to interface a SATA drive to a FPGA board or CoCo.

Ditto here. With SD, CF and plenty of old ATA rotating drives around, it'll 
be a while before we really need to go to SATA on our FPGA toys!

Regards,

-- 
|              Mark McDougall                | "Electrical Engineers do it
|  <http://members.iinet.net.au/~msmcdoug>   |   with less resistance!"