CD/DVD drives

I have six CD/DVD drives and all of them work on some machines. I have tested them on the following machines and am reporting the results here, as suggested on the downloads page for 5.28 compatibility:
Machine T:
; Titanium RISC OS 5.24 (16 Apr 2018)
; CDFSDriver 2.41 (24 May 2012)
; CDFS 2.63 (08 May 2016)
; drive connected by SATA
Machine A:
; ARMX6 (kit) RISC OS 5.23 (29 Jun 2016)
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.05 (08 Aug 2015)
; CDFS 2.63 (08 May 2016)
; drive connected by USB
Machine B:
; ARMX6 (kit) RISC OS 5.27 (12 May 2020) – OS update 18 beta low vector
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB
Machine C:
; Pandaboard ES RISC OS 5.27 (10 May 2020)
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Drives:
1 – maplin A58NG DVD-RAM UJ8E2
2 – samsung SE 208GB
3 – maplin MD9105-U2 DVD-RAM UJ8G2
4 – CIG DVD-RAM UJ820S
5 – CD-RW CRX835E (eject button broken, drive no longer works)
6 – ATAPI Pioneer DVD-RW DVRTD11RS

Results as follows:
T6 – mounts, reads disc, reports capacity all OK
A2 – mounts, reads disc, reports capacity all OK
C2 – mounts, reads disc, reports capacity all OK
B2 – appears to mount, reports capacity, but reports either ‘compact disc is faulty’ or ‘CD-ROM drive empty’ when doing *Cat
B1 – mounts, reads disc, reports capacity all OK
C1 – fails to report capacity, gives ‘no hardware at this device ID’
B3 – mounts, reads disc, reports capacity all OK
C3 – fails to report capacity, gives ‘no hardware at this device ID’
A4 – mounts, reads disc, reports capacity all OK
B4 – hangs when asked to report capacity until drive unplugged
C4 – mounts, fails to report capacity, hangs when asked to do *CAT until drive unplugged
drive 5 fails unilaterally!
drives 1 ro 4 work on PC no problems.

The items in bold are my usual standard connection. I have therefore not yet updated machine A (an ARMX6 kit) to a later OS than June 2016 as later OS updates might cause the CD to fail. Machine A is my standard exhibition machine with a PiFi board doing wireless connectivity. Don’t fix what ain’t broke! I also now bring a Pi 4 fitted with a WiFi Hat which copes with some exotic hotel free WiFi. Hotel? What’s a hotel?

ARMX6 (kit) RISC OS 5.27 (12 May 2020) – OS update 18 beta low vector
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Hi Chris, yes I get the same issue on my ARMX6 in that the previously working cd/DVD drive fails to mount a CD and gives CD Drive empty message with the latest ROM supplied as Beta by RComp.

CD/DVD drive is attached via USB to ATA bridge and is using a JMicron device like in the Pandaboard.

I’ll try my other machines that have working CD drives with the latest ROM’s over the course of today hopefully.

SCSIDevices read out below:

*scsidevices
Device Type Capacity Vendor Product Revision
0:0.0 Direct-access 224 Gbytes SanDisk SDSSDA240G U210
0:1.0 Read-only Unknown TSSTcorp CDDVDW SH-224FB SB00
0:2.0 Direct-access 58 Gbytes SanDisk Cruzer Blade 1.00
0:3.0 Direct-access Unknown Generic STORAGE DEVICE 0903

OK did some testing with my Pandaboard and put the 17th May 2020 ROM on it.

I had been helping Rob Sprowson look at an issue where the latest ROM’s would not let the old Pandabaord DVD drive work but in the end we put it down to a failing drive which may not be the case.

I installed a new LG drive that works using the later ROM’s on the Pandaboard.

As the ARMX6 has different connections for the DVD drive I attached the USB to Sata bridge from the Pandaboard to a USB port on the ARMX6 and supplied power to it as well.

I then increased the number of CD/DVD drives to 2 and tried to access the LG drive and it works with the latest RComp Low Vector 5.27 (12-May-20) ROM.

So something has changed that has stopped Sony/TSST drives working on the latest ROM’s?

OK done further testing and I managed to transfer the Pandaboard USB to Sata bridge to work on the ARMX6 so just leaving the CD/DVD drive out of tyhe equasion and it, the ARMX6, fails to work with the RComp supplied CD/DVD TSST Corp drive still.

Then by chance whilst rebooting and forgetting I had a CD in the drive I tried to access the drive and up it came.

So I took the drive out and rebooted and noticed a lack of the CD/DVD drive initilisation clunks whilst the ARMX6 booted and on trying to access a CD I get CD Drive empty again.

So leave a CD in the TSST drive and power off and then on and you hear the CD/DVD drive being accessed prior to RISCOS fully initialising with on screen messages.

Once in the RISCOS desktop you can access the CD that was in when you switched on the drive.

Eject that Cd and place another CD in the drive and trying to access gives the same Cd Drive empty message.

Could this an initialisation/sequencing issue or a module ordering issue?

Looks to me like you need to try CDFSSoftSCSI version 0.05 on the machines other than the armx6. So I’ve made one up – presumably it can be soft loaded CDFSSoftSCSI_0_05.zip

Hi Colin,

I have had some further discussion with Rob Sprowson and it looks as if he was already on the case and has submitted a potential fix for SCSISoftUSB.

Updated information below:

I have six CD/DVD drives and all of them work on some machines. I have tested them on the following machines and am reporting the results here, as suggested on the downloads page for 5.28 compatibility:

Machine T:
; Titanium RISC OS 5.24 (16 Apr 2018)
; CDFSDriver 2.41 (24 May 2012)
; CDFS 2.63 (08 May 2016)
; drive connected by SATA

Machine A:
; ARMX6 (kit) OS8 July 2016 RISC OS 5.23 (29 Jun 2016)
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.05 (08 Aug 2015)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Machine B:
; ARMX6 (kit) OS10a Jan 2018 RISC OS 5.23 12 Jan 2018
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.05 (08 Aug 2015)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Machine C:
; ARMX6 (kit) OS11 Feb 2018 RISC OS 5.23 18 Feb 2018
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Machine D:
; ARMX6 (kit) OS12 Apr 2018 RISC OS 5.25 25 Apr 2018
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Machine E:
; ARMX6 latest beta OS14a RISC OS 5.27 12-May-2020 low vector
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Machine F:
; ARMX6 latest ROOL RISC OS 5.27 19-May-2020 low vector
; drive connected by USB

Machine P:
; Pandaboard ES RISC OS 5.27 (10 May 2020)
; CDFSDriver 2.41 (24 May 2012)
; CDFSSoftSCSI 0.06 (10 Feb 2018)
; CDFS 2.63 (08 May 2016)
; drive connected by USB

Drives:
1 – maplin A58NG DVD-RAM UJ8E2
2 – samsung SE 208GB
3 – maplin MD9105-U2 DVD-RAM UJ8G2
4 – CIG DVD-RAM UJ820S
5 – CD-RW CRX835E (eject button broken, drive no longer works)
6 – ATAPI Pioneer DVD-RW DVRTD11RS

Results as follows:
T6 – mounts, reads disc, reports capacity all OK

A4 – mounts, reads disc, reports capacity all OK
Note A4 = B4 = C4 = D4 but E4 hangs until drive unplugged

F4 – mounts, reads disc, reports capacity OK

P2 – mounts, reads disc, reports capacity all OK
E2 – appears to mount, reports capacity, but reports either ‘compact disc is faulty’ or ‘CD-ROM drive empty’ when doing *Cat

F2 – mounts, reads disc, reports capacity OK

Note A2 = B2 = C2 = D2 = E2

E1 – mounts, reads disc, reports capacity all OK

F1 – hangs until unplugged

E3 – mounts, reads disc, reports capacity all OK

F3 – hangs until unplugged

A4 – mounts, reads disc, reports capacity all OK
E4 – hangs when asked to report capacity until drive unplugged
P4 – mounts, fails to report capacity, hangs when asked to do *CAT until drive unplugged
Note A4 = B4 = C4 = D4

drive 5 fails unilaterally!
drives 1 to 4 work on PC no problems.

The items in bold are my usual standard connection or other working connections. So drives 1, 2, 3, 4 and 6 all work, just not on all versions of RISC OS.

A tweak from Sprow has got E2 working but at the expense of E1 and E3 which now fail if that ‘fix’ is applied. F4 now works with the ROOL ROM.

One strange thing is if I plug in an empty DVD drive then press CTRL-Break, I get the network 
error 'machine start up has not completed successfully route add net gateway 169.254.24.1 file 
exists'. I have now learnt that this means that absolutely nothing is wrong just RISC OS or 
Linux (WiFi) is confused and requires a power off/on to regain sanity.

Are you trying to write one of those logic puzzles?

Tom has 3 oranges but an friday he has 2 bananas. Jill has apples when Fred has apples except on friday…..

Why is Freds cd rom failing?

The answer is a lemon.

It is a logic puzzle as the outcome does not seem to be linked to the version of CDFSSoft or CDFS. There’s something else…

There’s something else…

SCSISoftUSB. You and Doug are mentioned as testers for Rob’s SCSISoftUSB CD/DVD fix which got merged today, so I’m a bit surprised that there’s been so little mention of it in this thread!

I have had some further discussion with Rob Sprowson and it looks as if he was already on the case and has submitted a potential fix for SCSISoftUSB.

Well I did mention it yesterday!

Also my main testing was against the Pandaboard and it was only when RComp’s OS14 Beta made available on Friday did I get round to testing on the ARMX6 and it took me a few tests to get round to the idea it may be the same issue and hence involve Rob again.

I assumed everything relevant to CD would start CD… Oops. Must pay more attention.

There are two fixes, which seem to interact: Jeremy’s change for power eject (now regressed for the moment) and Sprow’s change to SCSISoftUSB residue values (which should be in today’s rom). The other recent change is isochronous USB and the *ReadEDID command, present in ARMX6 OS updates (to support HDMI audio) but not present in ROOL roms, if I understand correctly.

The problem seems to be a DVD drive which behaves correctly whilst empty but gets itself into a confused state when you insert a CD – it just repeatedly tries to mount the CD over and over again for ever until the drive is physically unplugged. A call to *CDDevices will hang while it is ticking. Hence once it gets into this state power off doesn’t help as it gets into the same state again when you power back on. Clearly the software is borking the drive somehow when you insert a CD as the drive works again correctly on a different version of RISC OS. Different versions of RISC OS fail different drives…

The term ‘borking’ is only to be used when you do not have a clue what is happening.

My ‘Xbox 360 HD DVD Player’ (Microsoft-branded, but made by Toshiba) now works with the latest ROM (it didn’t with 5.24). Well done :)

Gah, Chris, I’m so (sort of) envious… mine blew up (literally smoke/fire) about four years ago when I went to add it to my setup for legacy HDDVDs. I still have the drive here – can’t bear to throw it out!

Post above updated to reflect today’s rom (highlighted).

The score is announced: ROOL 2 R-Comp 2 so one drive requires any R-Comp build except the latest or the latest ROOL build;
one drive requires the latest ROOL build only and two of the drives require the latest R-Comp OS 14 beta and nothing else. I think therefore R-Comp wins by a short head.

Gah, Chris, I’m so (sort of) envious… mine blew up (literally smoke/fire) about four years ago when I went to add it to my setup for legacy HDDVDs. I still have the drive here – can’t bear to throw it out!

I regretted not getting a spare when they were $29 or whatever it was, as mine failed a short time later. I ended up getting a second-hand one plus a copy of AnyDVD HD for Windows, and proceeded to make backups of my 100-odd discs.

My Samsung drives works (CDFSDriver 2.41, CDFSSoftSCSI 0.06, SCSISoftUSB 0.27).

While I was at it I wanted to experiment with CD_SCSIUserOp but got the following error:
Soft-load driver does not support this CD_ SWI

Which is curious because looking at the sources, CDFSSoftSCSI sets a flag in its info block to let CDDriver handle and CDFSDriver checks the flag.

Looking in more detail at CDFDriver’s s.Main, I see that the swi number offset r11 is transformed into a driver opcode but in the code that tests which swis should be implemented by CDFSDriver itself (line 258+) r11 is tested against (XCD_xxx – XC_BaseNumber) – 1 instead of the DRIVER__xxx constants. By chance the DRIVER__xxx = (XCD_xxx – XC_BaseNumber) – 1 for all tested operations except for SCSIUserOp.

I have a Samsung/Toshiba DVD drive and am trying to write dvd discs with varied results.
I have found the modules you are listing but they are a lot older.
CDFSDriver 2.39
CDFSSoftSCSI 0.04
SCSISoftUSB 0.19
CDFS 2.49
CDFSFiler 2.47
Will newer ones help and where can I get them from.
My system is a Titanium board running 5.24

I have a Samsung/Toshiba DVD drive and am trying to write dvd discs with varied results.

The modules being discussed in this thread aren’t directly related to writing DVDs. In the case of a Titanium board the drive would almost certainly be connected via SATA (typing *SATADevices would confirm) and so SCSISoftUSB and CDFSSoftSCSI aren’t in play.

What version of CDVDBurn do you have? Steffen has a handful of versions which extend the range of drives supported beyond the version 2.02b mentioned on the hubersn site, even up to 2.10b, so he’s best placed to advise which you want.

I have to confess that although I have versions 2.02b, 2.09s and 2.10b of CDVDBurn I have not yet got far with testing 2.10b with the Titanium for Steffen. I was sent it on 23 Feb 2017 for testing but other things seem have intervened, not least of which was losing a great chunk of my email when I inadvertently formatted the hard disc (SD drive) using the old version of HDform. I will try to do better!

I am using 2.10b that Steffen sent me late last year.
*SATADevices shows a ATAPI TSSTcorpCD/DVDW TS-H653L drive.

I wanted to use it as a another back up as well as a second hard drive.The last hard drive
developed errors so I thought a DVD once written will be a better long term bet.
I tried flash drives but not had much luck with them.
Will a different make of drive prove more reliable.

Further to my reports on CD drives on ARMX6 I have noticed that A2 = B2 = C2 = D2 = E2 are marginal – whether the drive works or not is dependent on two things – whether the CD is in the drive when powered on and whether external power is supplied to the drive. With no power supplied to the drive (except through the data cable) it tends to work better and seemed to work originally on OS8 but retesting it it now fails (as shown above) unless, it turns out, the CD is in the drive when powered on.

A2 = B2 = C2 = D2 = E2

Lets not be cryptic and state that it is the Samsung SE 208GB drive on all machines that is marginal.

Have play around with a range of DVD/BR devices on Titanium. The reason was that the BR drive via USB was faster than a Sony Optiarc via SATA.
Have try LG (BR, DVD), Samsung, Sony, Freecom (NEC)… all on SATA. All are working but there is a wide range of speed. On paper all are “fast enough” but if I try to watch unprotcted DVD, no Problem with the BR device … slow motion with one Sony Optiarc slimline. The same with reading and writing spped and DVDBurn. An other Optiarc (same device older firmware) is fast…

Are you checking with different CD/DVDs? The reason I’m asking is that I have one of the Samsung drives (they were recommended for RPi use a few years back) and with one particular CD I’ve had issues somewhat like this; on RPi3B+ i/RO Direct it reports “CD empty”, Windows 10 tried to mount it then says “please insert a disk in drive D" while Ubuntu 18.04 keeps trying to mount it, and the CLI eject command doesn’t work—I have to restart the computer and press the eject button before the OS boots up and tries to mount the drive. There’s no obvious damage to the disk—and other CDs are fine.