Trying to Format RPC Hard Drive

discuss the archimedes & risc pc, peripherals and risc os/risc os on pi
Post Reply
Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Trying to Format RPC Hard Drive

Post by Rod C » Fri Aug 07, 2020 5:19 am

I'm trying to format a replacement hard drive in my StrongArm RIscPC 600 running RiscOS 3.7 but not having much luck. I'm using HForm ver 2.50.

The drive is a Samsung SP0411N 40GB IDE and the label on the drive doesn't give CHS specs but states LBA 78, 242, 976 40.0GB PANGO.

Here is the seller's description of the drive.
Seller's note.JPG
I was trying to guess CHS specs to use in HForm but it always stopped formatting within about 15-30 seconds with error &108C6 "The disc has no more available space".

I couldn't find anything useful about the drive until I included PANGO in my search and this turned up the Service Manual.
This says to use Cyl 77622, Heads 16, Sectors 63. So I calculate 77622 x 16 x 63 x 512 = 40,060,403,712. That looks good!
BIOS Setup Parameters.JPG
However, when I tried to format using those specs it still failed with the same error message after getting to 649. (Is that tracks?)
Not sure what else to try unless there is a different format program? If it would help, I'd be happy to format the drive to a lower capacity.

The old drive which is unresponsive is a Samsung 4.3GB with Cyl 8912, HD 15, Sec 63. I've tried this drive in my other RiscPC and it doesn't work there either. Attempts to access it brings up the response "Disc not understood - has it been formatted?" I had backed up some of this drive to a set of zip discs as it appeared to be failing a while ago. Now presumably dead but I haven't tried to reformat it just in case there is some other issue. It's probably coincidence but this happened after I replaced a failed CMOS chip and got that aspect working properly again. That wasn't a corrosion issue, just my stupidity in connecting a new CMOS battery with reverse polarity! :shock: :roll: #-o ](*,)

philpem
Posts: 559
Joined: Fri Apr 04, 2014 6:42 pm
Contact:

Re: Trying to Format RPC Hard Drive

Post by philpem » Fri Aug 07, 2020 8:27 am

Hard drive compatibility with the onboard IDE controller is very hit-or-miss, sadly. I remember trying a box of drives and the only one I could get to work reliably was a 6GB Seagate.
Riscos.info suggests that you need to *Configure ADFSBuffers 0 or install the RISC OS ROM Patch with drives larger than 2GB: https://www.riscos.info/index.php/RISC_ ... machine.3F

If you're not dead-set on a hard drive, a Startech CompactFlash-to-IDE adapter and a CompactFlash card is a good solution. I've personally used the 16GB SanDisk Extreme, I've heard others have had good luck with the "Cloudisk" cards from Amazon.

If you have access to a 3D printer, I created an adapter to mount the Startech adapter in the RISC PC's hard drive bay: https://www.thingiverse.com/thing:4546129
Otherwise a couple of foam double-sided sticky pads will likely work just as well.

Cheers
Phil.

Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Re: Trying to Format RPC Hard Drive

Post by Rod C » Fri Aug 07, 2020 3:45 pm

Hi Phil
Thanks for your reply and ideas to follow up. I had already configured ADFSBuffers 0 and that doesn't help with the failure to format. I had another try anyway and got the same result.

The RiscOS.info page refers to the machine hanging when using drives >2GB and it makes me wonder if that was the cause of the problems I had with the earlier 4.3GB hard drive. At that stage the ADFSBuffer setting was not 0. When the computer would hang, I had to just switch off the power and maybe that was a bit harsh on the hard drive if it wasn't able to carry out the parking function. I did transport the computer with me when I was isolating at the beach house so it might have been bumped around too much also. It hasn't seized as I can hear it spin up and the activity light comes on but it just fails to work. Sad :cry:

I would like to get the RPC working again on a hard drive but I might end up using a CF adapter as you suggested. There's always the risk that another old drive could fail anyway.

Oh, I just remembered! Another pathway could be to put in the SCSI card I have stashed away for this computer and put a hard drive on that.

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Trying to Format RPC Hard Drive

Post by Kazzie » Fri Aug 07, 2020 3:45 pm

philpem wrote:
Fri Aug 07, 2020 8:27 am
Otherwise a couple of foam double-sided sticky pads will likely work just as well.
That's what was provided by Cumana for their (non-ATAPI) CD interface card (stuck down in the space next to the hard drive). :lol:
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

User avatar
helpful
Posts: 634
Joined: Tue Sep 22, 2009 1:18 pm
Location: London
Contact:

Re: Trying to Format RPC Hard Drive

Post by helpful » Fri Aug 07, 2020 7:23 pm

It's been a long time, but I'm not sure a 40GB drive will work on RO3.7. It certainly won't work efficiently - the allocation unit size will be massive! RO4+ was much better for large hard drives.

You could try the latest HForm 2.75 to see if it makes any difference -
https://packages.riscosopen.org/rool/Di ... html#HForm

Bryan.
RISC OS User Group Of London - http://www.rougol.jellybaby.net/
RISC OS London Show - http://www.riscoslondonshow.co.uk/

User avatar
BeebMaster
Posts: 3626
Joined: Sun Aug 02, 2009 5:59 pm
Location: Lost in the BeebVault!
Contact:

Re: Trying to Format RPC Hard Drive

Post by BeebMaster » Fri Aug 07, 2020 8:28 pm

I have a 40GB IDE Seagate drive in my RPC 700. I had the same problem with HForm (2.48) which, by a process of error and further error, I worked out was because it will only accept 77,505 cylinders maximum. So I formatted it to 77,500 and it worked fine. It is very inefficient, the allocation unit is 128K. Currently it tells me there are 12GB used but there is only about 500MB of data on there.
Image

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Trying to Format RPC Hard Drive

Post by Kazzie » Fri Aug 07, 2020 9:20 pm

BeebMaster wrote:
Fri Aug 07, 2020 8:28 pm
Currently it tells me there are 12GB used but there is only about 500MB of data on there.
:shock:
With that level of inefficiency, you could have a position on a Government IT programme!
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

User avatar
BeebMaster
Posts: 3626
Joined: Sun Aug 02, 2009 5:59 pm
Location: Lost in the BeebVault!
Contact:

Re: Trying to Format RPC Hard Drive

Post by BeebMaster » Fri Aug 07, 2020 9:23 pm

I did it a disservice...slightly! Nearly 870MB used, but still taking up 12GB!!!
scrn0.png
Image

Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Re: Trying to Format RPC Hard Drive

Post by Rod C » Sat Aug 08, 2020 8:24 am

Talk about inefficiency, my Large Time Allocation Unit on this problem has grown to a ridiculous size! :lol:

Thanks for the suggestions, but still no success.

I've tried many different configurations with both HForm ver 2.50 and ver 2.75.
BeebMaster wrote:
Fri Aug 07, 2020 8:28 pm
I have a 40GB IDE Seagate drive in my RPC 700. I had the same problem with HForm (2.48) which, by a process of error and further error, I worked out was because it will only accept 77,505 cylinders maximum. So I formatted it to 77,500 and it worked fine.
BeebMaster, when I set 77500 cylinders ,16 heads, 63 sectors it got up to 649 ( I'm pretty sure now this is cylinders) and stopped with the same old no more space error.

Recalling that some of my research suggested the drive has 1 head, not 16, prompted me to try that setting. Indeed,checking the PANGO Service Manual I see 1 disc/1 head.
PANGO specs.JPG
I'm not sure if this is really "true" as it conflicts with the BIOS settings given elsewhere in that document. However, I think I read somewhere that the 16 heads all move together when instructed to read/write and this might be relevant. Anyway that prompted me to try some formatting settings specifying only 1 head.

The most promising was setting 10395 cylinders, 1 head, 63 sectors.
The format process gets to 10394, reports that the file allocation unit will be 1024 then says:
Writing defect list
Creating Map
Writing map
Writing root directory
HForm failed. Disc not understood. Has it been formatted? at line 115 ](*,)
helpful wrote:
Fri Aug 07, 2020 7:23 pm
You could try the latest HForm 2.75 to see if it makes any difference -
https://packages.riscosopen.org/rool/Di ... html#HForm
Thanks, Bryan. I got that and it looked really promising when it reported the following:
LBA size 78,242,976 sectors
Suggested shape 20547 cyl, 16 heads, 238 sectors/track
Trying to format with those specs got me only to 172 then the "No more space" error.

Ahh, did I mention that the PANGO Service Manual is my guide? Here's a sample of the clarity of that manual -

Code: Select all

① Boot Record: Boot record is consists of sector 1, track 0, side 0.
Boot record is the one of most important parts. It has some information about the physical
structure of disk. If boot record is vanished, the disk is useless entirely.
The physical information mean, for example, the number of FAT, capacity of disk something
like that.
Yeah, something like that.:shock:

User avatar
BeebMaster
Posts: 3626
Joined: Sun Aug 02, 2009 5:59 pm
Location: Lost in the BeebVault!
Contact:

Re: Trying to Format RPC Hard Drive

Post by BeebMaster » Sat Aug 08, 2020 12:47 pm

The manual for my Seagate drive had unusual geometry, which made it only format to 7GB initially, but I got the number of cylinders by working back from the "guaranteed sectors" count specified in the service manual, divided by 16 heads and 63 sectors per track.

What is the exact "no more space" error, and is there a line number? If you chain or load/run Hform with BASIC already running rather than double-clicking on the App it might give you a line number. It sounds like the equivalent of 8-bit BASIC's DIM space, or even No room, which sounds like the huge number of cylinders has used up all the available RAM. My RISC PC has the "next task" memory allocation set to 16384K, although at the time I did this it was 4096K. So maybe try increasing the "next" slider in the task manager to at least 4096K and running Hform again.
Image

User avatar
daveejhitchins
Posts: 5913
Joined: Wed Jun 13, 2012 6:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: Trying to Format RPC Hard Drive

Post by daveejhitchins » Sat Aug 08, 2020 3:09 pm

IIRC most hard drives can accept various Heads, Cylinders and Sectors and just maps them to the physical drive. So experimenting sometime gives you the best results - You won't damage anything, taking this approach! - I mean . . . 16 Heads! they stopped using that many even further back :lol:

Dave H.

Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Re: Trying to Format RPC Hard Drive

Post by Rod C » Sat Aug 08, 2020 3:15 pm

BeebMaster wrote:
Sat Aug 08, 2020 12:47 pm
What is the exact "no more space" error, and is there a line number?
This is reported as Error &108C6 "The disc has no more available space".

The other error occurs when I use parameters that get to the stage of writing the defect list, creating and writing the map and then writing the root directory. This error shows the line number which is line 115 in HForm ver 2.50 and line 135 in HForm ver 2.75 as shown below:
135PRINT:PROCL:PROCM;OSCLI("-"+f$+"-%MOUNT :"+STR$p%):OSCLI("-"+f$+"-%FREE:""+STR$p%):OSCLI("-"+f$+"-%DISMOUNT :"+STR$p%)
Looks to me like this error is generated by failure to mount the drive.
BeebMaster wrote:
Sat Aug 08, 2020 12:47 pm
So maybe try increasing the "next" slider in the task manager to at least 4096K and running Hform again.
I increased the memory allocations to:
Next 16512K
System heap/stack 2040K
and set up a RAM disc of 100K to copy !HForm to.

I ran HForm from BASIC using the CHS parameters sugggested by HForm ver 2.75 which is able to read the LBA data from the disc. This didn't proceed any further than when I had much lower memory allocations for the program and generated Error &108C6 as before.

It's all pretty mysterious and frustrating, especially given the likelihood of hardware incompatibility.

This evening I bought a Seagate Medalist 6531 IDE 6.4GB disc for a very reasonable price. It's described as ""Opened – never used: An item in excellent, new condition with no wear". It's a fairly short postal distance away in Melbourne and seems like a better chance than the Samsung 40GB. It should be here soon and I will have my fingers crossed!

Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Re: Trying to Format RPC Hard Drive

Post by Rod C » Sat Aug 08, 2020 3:27 pm

daveejhitchins wrote:
Sat Aug 08, 2020 3:09 pm
So experimenting sometime gives you the best results - You won't damage anything, taking this approach!
That's reassuring! I've seen several comments about no need to format, just initialize the disc but does that apply if the disc had been formatted on a different system such as a PC (not the Risc kind)?

When the LBA size (78242976 cylinders) is converted to "suggested shape" of 20574 cyl, 16 heads, 238 sectors is that just a way of representing a format geometry and not a reference to 16 actual heads?

User avatar
daveejhitchins
Posts: 5913
Joined: Wed Jun 13, 2012 6:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: Trying to Format RPC Hard Drive

Post by daveejhitchins » Sat Aug 08, 2020 4:01 pm

daveejhitchins wrote:
Sat Aug 08, 2020 3:09 pm
So experimenting sometime gives you the best results - You won't damage anything, taking this approach!
That's reassuring! I've seen several comments about no need to format, just initialize the disc but does that apply if the disc had been formatted on a different system such as a PC (not the Risc kind)?
I believe (someone please qualify!) that the Formatter reads how the disc has been formatted - PC/Acorn doesn't matter!

When the LBA size (78242976 cylinders) is converted to "suggested shape" of 20574 cyl, 16 heads, 238 sectors is that just a way of representing a format geometry and not a reference to 16 actual heads?
Correct. You can also format them to be a much smaller drive - The drive itself doesn't care! I know for certain, in the past, that sellers have formatted to smaller capacity to avoid a cluster of bad-sectors at the top end of the drive.

As an aside, I was impressed with the Watford Formatter, in one respect, in that it checks for bad sectors and writes the bad-sector table. Not all Formatteres do that automatically!

Dave H.

User avatar
helpful
Posts: 634
Joined: Tue Sep 22, 2009 1:18 pm
Location: London
Contact:

Re: Trying to Format RPC Hard Drive

Post by helpful » Sat Aug 08, 2020 5:02 pm

Rod C wrote:
Sat Aug 08, 2020 3:27 pm
I've seen several comments about no need to format, just initialize the disc
I think you only need to initialise it. Are the errors you're getting from HForm when you try to do a full format?
BeebMaster wrote:
Fri Aug 07, 2020 9:23 pm
I did it a disservice...slightly! Nearly 870MB used, but still taking up 12GB!!!
> PRINT 870218619/212595
4093.31649

You have an average file size of under 4K on a disc with a 128K allocation unit :lol:
RISC OS User Group Of London - http://www.rougol.jellybaby.net/
RISC OS London Show - http://www.riscoslondonshow.co.uk/

User avatar
BeebMaster
Posts: 3626
Joined: Sun Aug 02, 2009 5:59 pm
Location: Lost in the BeebVault!
Contact:

Re: Trying to Format RPC Hard Drive

Post by BeebMaster » Sat Aug 08, 2020 6:16 pm

Most of the files are exactly 960 bytes, being over 100,000 Teletext pages. Can I force it to use a smaller allocation unit, eg. by bodging HForm not to insist on a higher value? Or would that cause problems with the ADFS file structure?
Image

Kazzie
Posts: 1793
Joined: Sun Oct 15, 2017 8:10 pm
Location: North Wales
Contact:

Re: Trying to Format RPC Hard Drive

Post by Kazzie » Sun Aug 09, 2020 7:05 am

It strikes me that one compromise would be to have two drives: a small one with a regular allocation unit size for general use, and a big one with a larger allocation unit for big files.
BBC Model B 32K issue 7, Sidewise ROM board with 16K RAM
Archimedes 420/1 upgraded to 4MB RAM, ZIDEFS with 512MB CF card
RiscPC 600 under repair
Acorn System 1 home-made replica

sirbod
Posts: 1133
Joined: Mon Apr 09, 2012 9:44 am
Location: Essex
Contact:

Re: Trying to Format RPC Hard Drive

Post by sirbod » Sun Aug 09, 2020 8:35 am

Kazzie wrote:
Sun Aug 09, 2020 7:05 am
It strikes me that one compromise would be to have two drives: a small one with a regular allocation unit size for general use, and a big one with a larger allocation unit for big files.
The route I took was to have a small donor drive using the old FileCore format, it updates FileCore, ADFS etc to support large discs (and vicariously small sectors) and long filenames, then mounts a 2nd drive which becomes the main boot drive. I also added some other bits to allow me to boot many OS versions, but that can be ignored in this instance.

Everything needed is in this post, although I dont recall if I patched the version of ADFS in it to support modern drives, so it might need my ADFS patcher running on it.

Rod C
Posts: 237
Joined: Wed Jul 04, 2018 3:26 pm
Location: Australia
Contact:

Re: Trying to Format RPC Hard Drive

Post by Rod C » Mon Sep 21, 2020 4:10 am

Rod C wrote:
Sat Aug 08, 2020 3:15 pm
This evening I bought a Seagate Medalist 6531 IDE 6.4GB disc for a very reasonable price. It's described as ""Opened – never used: An item in excellent, new condition with no wear". It's a fairly short postal distance away in Melbourne and seems like a better chance than the Samsung 40GB. It should be here soon and I will have my fingers crossed!
This drive was intended as a second try replacement for the hard drive that failed in my StrongARM RPC. As it turned out I had no success initializing or formatting this drive either. :roll:

Not giving up, I arranged to buy a Quantum Trailblazer with options for 840, 630 or 420 MB which the seller kindly tested and reformatted (840MB) in a PC. This was encouraging but it turned out to be not so simple in the land of the RiscPC.

In the StrongARM RPC (RiscOS 3.7, ADFS 3.27) it failed to initialize or format to any of the alternate sizes. #-o

In the ARM610 RPC (RiscOS 3.5, ADFS 2.91) the format program reported "This version of ADFS only supports up to 512MB." I selected the CHS parameters for 420MB and selected Initialize only. This worked and I quickly had a mountable drive. :)

When I swapped the drive over to the StrongARM RPC it wouldn't mount so I returned it to the ARM610 and checked that I could copy some files to it but when I ran verify it threw lots of errors. ](*,) This was a bit alarming but I thought it might be because of the failed formatting operations I had done earlier. Now I ran a full format for 420MB in the ARM610 which completed with no errors. Back in the StrongARM however, the disc still failed to mount but this time when I tried again to format for 420MB it succeeded. :lol: I then mounted the Iomega Zip drive and restored the files I had backed up from the original disc.

Encouraged by this success, I decided to see if anything could be done with the original Samsung 4.3GB (CHS = 8912,15,63) drive that failed in this computer. :roll: I swapped it back in, ran !HForm, initialize only, short test for defects and it reported cyl 8912, head 4, sector 30 defect, defect is beyond end of disc. (I'm wondering how is this so?) It kept trying to write different patterns which all failed and I had to turn off the machine. :cry: I tried initialize again and this time it succeeded and I was able to mount the 4.3GB disc. I ran verify and no problems were found. :shock: :lol:

Finally, I restored all the files from the backup and this drive has been working faultlessly since! \:D/

So, it turned out to be a long and rather tedious journey to get back to where I started from but I might have learnt a little along the way and I now have a few spare drives which might come in handy one day, including one with a functional backup on it. I wonder what happened to make this disc fail in the first place and does the successful verify really indicate that it is without defects? Perhaps time will tell...

Thanks to those who gave advice.

Rod

Post Reply

Return to “32-bit acorn hardware”