A3010 hard disc interface

Arc/RPCs, peripherals, RISCOS operating system & ARM kit eg GP2x, BeagleBoard
User avatar
danielj
Posts: 6269
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 hard disc interface

Post by danielj » Fri Mar 16, 2018 2:48 pm

=D> I'd put the BD pads along the top edge, so they're nearer to the podule connector where you're going to have to pick them up from . You don't need to worry about the RST to 711, that's already there in the 3010 (check the the 3010 schematics) - I don't think you're completely replicating the HCT14 circuit, just the bits that aren't on the A3010 (i.e., inverting I!3*), juts. You just need to take rst* straight from the podule connector to the ide if I recall correctly!

Just be warned, the IDE interface on these is really flakey - it doesn't get on with most CF cards :( Basically we need to rebuild the ROMs with bugfixed modules in place. No one's got around to doing that yet though.

d.

Edit: Also, you might want to consider the position of the CF slot - I don't think that will fit in an A3010 Adelaide model: Image

steve3000
Posts: 1851
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 hard disc interface

Post by steve3000 » Fri Mar 16, 2018 6:13 pm

Just a thought, how difficult would it be to add wiring and a socket for an extension ROM on the same board? (As in the A5000...because you have all the address and data lines already in place?)

This would make the board a little larger (less so if extension ROM mounted underneath is a possibility) but could allow fitment of Wizzo ROM, which would sort out the IDE interface and make compatible with most CF cards...

User avatar
danielj
Posts: 6269
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 hard disc interface

Post by danielj » Fri Mar 16, 2018 9:07 pm

Just looking, you'd need Rcs1* (IOEB chip), which doesn't look like it exists as a signal from the ARM250... :(

steve3000
Posts: 1851
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 hard disc interface

Post by steve3000 » Fri Mar 16, 2018 10:47 pm

...humm...but then how's it done in the pre-IOEB A5000s? :?:

User avatar
danielj
Posts: 6269
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 hard disc interface

Post by danielj » Fri Mar 16, 2018 11:40 pm

Good point... Schematic?

d.

steve3000
Posts: 1851
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 hard disc interface

Post by steve3000 » Sat Mar 17, 2018 8:03 am

danielj wrote:Good point... Schematic?
Is this not online somewhere?

I've got it on paper. IC21 generates Rcs1*... I'll post up a picture later, but it's labelled ASIC, so one of the bank of custom ICs. Might be possible to work out from the inputs maybe...

User avatar
danielj
Posts: 6269
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 hard disc interface

Post by danielj » Sat Mar 17, 2018 9:09 am

The problem I forsee is what causes the OS to look for a 5th column ROM in that context? None of the other machines have that socket, so perhaps it'll only look for it on an A5000? The alternative would be to rig it up like a podule ROM, but I don't know enough about the minipodule architecture to say whether that will prevent using another minipodule with a ROM :?

d.

steve3000
Posts: 1851
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 hard disc interface

Post by steve3000 » Sat Mar 17, 2018 1:10 pm

danielj wrote:The problem I forsee is what causes the OS to look for a 5th column ROM in that context?
RISC OS 3 will look for the extension ROM whatever machine - just need to replicate the extension ROM select line and link up to the correct address lines (I expect...) remember, the A4 also has an extension ROM :)

Attached is a photo of the non-IOEB A5000 circuit page where Rcs1* is generated (IC21, top left). Would need to decode IC21 to work out the correct operation...but then should be possible to replicate.

The alternative, as you mention, would be to add as a podule ROM. To do this, you'd normally need some address decoding/latching ICs if the ROM size is >2kb. But one thought could be to squeeze Jon (sirbod's) ADFS patch into <2kb, which would fix CF compatibility, and should make it simpler to implement as a patch ROM, since the podule bus is designed with a 2kb addressable window...?
Attachments
20180317_130159.jpg

User avatar
IanS
Posts: 607
Joined: Mon Aug 31, 2009 6:02 pm
Contact:

Re: A3010 hard disc interface

Post by IanS » Sat Mar 17, 2018 1:53 pm

steve3000 wrote:...since the podule bus is designed with a 2kb addressable window...?
Isn't it 4K words, it's just normally split in two for rom and peripherals. So you could use 4K, if you don't need anything else.

steve3000
Posts: 1851
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 hard disc interface

Post by steve3000 » Sat Mar 17, 2018 10:53 pm

IanS wrote:
steve3000 wrote:...since the podule bus is designed with a 2kb addressable window...?
Isn't it 4K words, it's just normally split in two for rom and peripherals. So you could use 4K, if you don't need anything else.
Quite right Ian, hadn't considered that :)

Even so, 4kb is still a bit limiting...but easily enough to patch ADFS. Just need to build a prototype now...

Post Reply