Maximum HDD Size for RISC OS 3.7
robheaton
Can someone please tell me what the maximum hard drive size is for
RISC OS 3.7?
(This is for a SA Risc PC, using the onboard IDE)
Many Thanks,
Rob.
Rick Murray
> Can someone please tell me what the maximum hard drive size is for
> RISC OS 3.7? (This is for a SA Risc PC, using the onboard IDE)
Older RISC OS, internal ADFS... you'd be looking at 512Mb with no
partition support. That was the main reason I got myself a Simtec board,
four drives and partitions (inc. >512Mb) possible.
Best wishes,
Rick.
David Holden
Sorry, but that answer's very wrong. That's the max size for RO 3.1 and 3.5
The max drive size for RO 3.6 and better is 128GB.
You'll find the LFAU gets pretty big after 40GB for RO 3.6 and 3.7 so it
will waste a lot of space if you have mainly smallish files. Also you can't
use any modern drives over 40GB on the motherboard interface because of
hardware problems.
I'd suggest if you really need a large hard drive upgrade to RO 4, which
will give you the later filecore with more efficient disc usage, long
filenames, etc. Alternatively fit one of our IDE interfaces so you can
partition the drive to reduce the LFAU and also use a modern (hence cheaper)
drive, plus it's a lot faster than the on-board IDE. Or, of course, do both.
--
David Holden - APDL - <http://www.apdl.co.uk>
robheaton
Thanks for the advice!
David, I'll probably call you over the next day or so, as I'm after a
couple of bits and bobs.
Rob.
Dave Plowman (News)
<ced47bd2-96da-4ad0...@j31g2000yqe.googlegroups.com>,
robheaton <robjh...@gmail.com> wrote:
> Thanks for the advice!
> David, I'll probably call you over the next day or so, as I'm after a
> couple of bits and bobs.
I can certainly recommend the Blitz card. Not only does it allow larger
modern HDs to be used, but it seems more tolerant of CD ROM and writer
types than any other I've tried. (I have a Unipod on the other machine).
It's also noticeably faster than the IDE buss. However, I am using 4.39.
--
*Money isn't everything, but it sure keeps the kids in touch *
Dave Plowman da...@davenoise.co.uk London SW
To e-mail, change noise into sound.
Rick Murray
> Sorry, but that answer's very wrong. That's the max size for RO 3.1 and 3.5
Oops, my bad. I'll have to go re-read PRM5 as punishment. ;-)
> You'll find the LFAU gets pretty big after 40GB for RO 3.6 and 3.7 so it
> will waste a lot of space if you have mainly smallish files.
Yes, there's always going to be a trade-off between the physical size of
files, and the limitations on how finely you can address specific parts
of the disc.
There was discussion I saw a while back where a new harddisc would
present 256Kb "sectors", which is small problem if you stuff the drive
full of movies and music, but could play hell with operating systems
that hold a sufficient number of small text files for configuration, not
to mention fifty copies of "COPYING.TXT" each rounded up to 256K apiece...
> Also you can't use any modern drives over 40GB on the motherboard
> interface because of hardware problems.
Why? Just interested, as 40Gb is, if my maths is correct, within 45 bit
address, it seems strange to cut off at that point.
> Alternatively fit one of our IDE interfaces so you can partition the drive
> to reduce the LFAU and also use a modern (hence cheaper) drive, plus it's
> a lot faster than the on-board IDE. Or, of course, do both.
I don't have one of your interfaces (mine is Simtec), but I would
completely agree. The only things on my internal IDE are the boot drive
(left as installed) and the CD-ROM. All the serious work is done through
the podule, which talks to two drives as a Master - one 4Gb drive
formatted into 6 partitions of 512Mb, and one 8Gb drive formatted into
two 4Gb partitions. That used to be where I'd build up ISO images (like
backups of the other partitions), but now I have a big drive on the PC
plus a DVD writer, it all takes place there now. [and my RISC OS drives
don't change enough these days to warrant regular backups]
Either way, I think the motherboard IDE is sufficient to get the machine
booted and hold the basic set of tools (MakeModes, etc). Anything else
would benefit from an IDE podule.
Best wishes,
Rick.
Steffen Huber
> On 21/06/2011 07:55, David Holden wrote:
>> Also you can't use any modern drives over 40GB on the motherboard
>> interface because of hardware problems.
>
> Why? Just interested, as 40Gb is, if my maths is correct, within 45 bit
> address, it seems strange to cut off at that point.
It is not size-related, but manufacturing-date-drive-electronics
related.
I have successfully operated 80 GB, 120 GB and 160 GB (formatted
down to 128 GB) drives in my humble Risc PC. The newest is
a Western Digital WD1600AAJB, which works 100% reliable. You
need to run HForm twice, first run ends with a strange error,
but other than that, it just works.
This WD drive should still be currently available.
Other drives (e.g. various 2.5" drives) however don't work
on the internal IDE. I wonder whether this is really hardware
related or if it is ADFS related. The chip used in the RPC
for IDE was also widely used in PC world.
Steffen
--
Steffen Huber - http://www.huber-net.de/
hubersn Software - http://www.hubersn-software.com/
Theo Markettos
> Other drives (e.g. various 2.5" drives) however don't work
> on the internal IDE. I wonder whether this is really hardware
> related or if it is ADFS related. The chip used in the RPC
> for IDE was also widely used in PC world.
The reason is likely to be the /IOCS16 line on the motherboard IDE. This
controls the data bus buffer to select whether an 8 or 16 bit transfer is
expected (most registers are 8 bits, except the data register is 16). It's
generated by the drive, and operates logic outside of the FDC37C665GT I/O
controller (which doesn't implement any IDE bus apart from address
decoding). Only it was already deprecated or otherwise on the way out when
the Risc PC was released in 1994. So it's not surprising manufacturers have
finally given up implementing it. IDE podules don't use this line, and
hence have no problems.
Indeed, there's no real reason to have it either: there's nothing wrong with
all transfers being 16 bits. I've just realised something else. SATA-PATA
adaptors don't work in Risc PCs, do they? It might be due to the same
problem. You can probably modify one to force the /IOCS16 pin low, in which
case it might allow the attachment of SATA drives.
Anyone feel like trying it?
Theo
Alan Adams
Theo Markettos <theom...@chiark.greenend.org.uk> wrote:
> Theo
If I understand the above correctly, the simplest way might be a short
cable with a plug and socket, between the motherboard socket and the
drive cable. The required change could then be done by cutting and
splicing bits of the cable.
--
Alan Adams, from Northamptonshire
al...@adamshome.org.uk
http://www.nckc.org.uk/
Chris Evans
We'll try it.
Just to be sure are you suggesting forcing the /IOCS16 pin low whilst
connected to both Motherboard and the drive or on just the motherboard leaving
that pin on the Drive floating?
Pulling up is normaly done via a resistor, does forcing down need one? Value?
Chris Evans
--
CJE Micro's / 4D 'RISC OS Specialists'
Telephone: 01903 523222 Fax: 01903 523679
ch...@cjemicros.co.uk http://www.cjemicros.co.uk/
78 Brighton Road, Worthing, West Sussex, BN11 2EN
The most beautiful thing anyone can wear, is a smile!
Dave Higton
Chris Evans <ch...@cjemicros.co.uk> wrote:
> Just to be sure are you suggesting forcing the /IOCS16 pin low whilst
> connected to both Motherboard and the drive or on just the motherboard
> leaving that pin on the Drive floating?
>
> Pulling up is normaly done via a resistor, does forcing down need one?
> Value?
There has been a tradition for decades in digital design that pull-
ups use a resistor, usually between 1k0 and 10k; and pull-downs are
short circuits.
The above is for inputs. If the electronics can ever drive that
point, it's necessary to look at each case individually.
Dave
Theo Markettos
> There has been a tradition for decades in digital design that pull-
> ups use a resistor, usually between 1k0 and 10k; and pull-downs are
> short circuits.
>
> The above is for inputs. If the electronics can ever drive that
> point, it's necessary to look at each case individually.
There's a 1.2K pullup on the motherboard. Pin 32 is /IOCS16 on early
interfaces, but now not-used/not-connected. If you know your interface
doesn't drive it (eg a SATA-PATA adaptor), you can short it to ground. If
you might connect it to a drive that does implement it (eg when modifying an
IDE cable), best to short the motherboard end to ground and leave the drive
pin unconnected.
Theo