Dave-H

Thanks, but there are no restrictions set up in Group Policy. I've done many other installs which have had no problem, so why that particular one has an issue I have no idea. I did copy the files from the CD to the hard drive and run it from there as well as running from the CD directly, and the result was the same. IIRC there was some mention in the installer log that something wasn't signed, but usually if that's the case you get an option dialogue box to ignore that. Not in this case.

Thanks. What I will do I think is put the 3TB disk into one of the internal caddies that I use for one of the other drives on the system. That is connected directly to the motherboard and is permanently powered, so it will just appear as a normal internal directly SATA connected disk. If that continues to have the corruption problem, I would have thought that the number one suspect has to be the Paragon GPT driver, as surely it should work fine natively in Windows 10.

Please don't blame yourself! It was a perfectly reasonable thing to do, to try other drivers to see if they made any difference to the issues. I'm really not at all sure why trying to install the 3.2.0 driver caused so much problem, it says quite specifically that it's XP compatible, and the installer certainly didn't say otherwise. What I should have probably done is completely uninstall the existing driver and clean the system of it before installing 3.2.0, rather than trying to install 3.2.0 over the top of it. I still don't understand why the driver on the disk that came with the card won't install because of a 'software restriction policy'. That is very puzzling. I did try changing that BIOS option BTW, but it seemed to make no difference to anything. I do think that the important thing now is to track down the cause of this file system corruption that I'm seeing when going between XP and 10. If that can't be explained, the whole thing is useless anyway, so there's no point in playing with different drivers and cards.

@jaclaz Do we now think that this file system corruption when going between operating systems is an intrinsic problem? What I could do is to try temporarily connecting the disk in the enclosure directly to the motherboard, and make sue it's powered all the time that the system is on, so it will be just like it was a fixed internal drive. If I use it like that and see if the corruption still occurs, that should at least prove whether the problem is being caused by the connection via the card or not.

I've been trying other drivers on XP, with no success I'm afraid. The version 1.1.7.110 driver wouldn't install at all. If I run the setup, it runs briefly and then just closes, with no messages on screen. This might have been because there was already a later version of the driver installed. Version 3.2.0 was even worse, it apparently installed, but then there was a BSOD, MACHINE_CHECK_EXCEPTION. On reboot, the machine started OK, but nothing seemed to have changed. In Device Manager, the Asmedia card now showed driver version 3.2.0.0, but on inspecting the driver files it was still using asahci32.sys version 2.0.3.0001 as it was before. Trying another update from the INF file just produced a system freeze. I then tried reinstalling the original 2.0.4 driver. That said another stalled installation needed to be rolled back, which it did and the system froze again! It rebooted OK, and I then manually cleaned the registry of the previous driver installations and started from scratch again with the 2.0.4 driver, which then installed fine. I don't think I'll be trying that again! There was no evidence at all that version 1.1.7.110 was 98 compatible, in fact the INF file specifically says that it isn't, so I'm not even going to try!

This is what I have in my BIOS. The ROM Scan Ordering is set to 'Onboard First' as you can see. The other option is 'Addon First'. There doesn't seem to be an 'Option ROM Scan' option in my BIOS, but the BIOS options can change of course depending on the configuration of the machine, type of processor etc. I haven't tried changing the option. I did try as an experiment connecting the Blu-ray drive to the eSATA port on the other Silicon Image card in the machine, the one I used to connect the external disk enclosure to. To my surprise, it wasn't recognised at all, it was like it didn't exist! No sign of it in the BIOS still, and not in the XP Explorer or Disk Management. I didn't try in Windows 10, but I guess the result would have been the same. I tried updating the driver in XP using the one on the disk which came with the card. It went through the motions and then said 'Windows cannot open this program because it has been prevented by a software restriction policy.' I don't know what to do about that.

I've only just started using Supermium, and while I think it's a miracle that a Chromium 121 based browser will run on XP, it's a long way from being a replacement for 360Chome for me as well. The GUI font on Supermium is rough, not unusable but ugly. It still has some stability problems, I find it crashes if I try to use the password manager, for instance. I hate the fact that it closes when you close the last tab. You can stop that with a simple setting in 360Chrome. Of course this can be achieved with an extension, I'm sure, but that's not the same! On my system, which is quite powerful, Supermium is slow to start, and quite slow in use. 360Chrome is much better, and I'm using 13.5.2036, which is probably not as well optimised as v13.5.1030 Redux is now. While it's good to have a fallback available if I run into a site which 360Chome no longer works properly with (there aren't many, but the number will inevitably increase, of course) I still wouln't use Supermium as my default browser yet.

I suspect that, at least on Windows 10, it's being used by Windows Search. It's not included in Windows Search on XP though, so that is a bit of a mystery. If I do make a 2TB partition which can be seen normally in XP, can I then make another 1TB partition in Windows 10, which will only be seen in Windows 10? It still has to be a GPT disk I assume? I will certainly be doing the tests you advise, as I do need to know exactly what is corrupting the file system.

Chkdsk is not running automatically on either OS. Could that be because it's being seen as a removable drive, although I wouldn't expect that to make any difference? I'm wondering whether to change the XP driver, and use the one provided with the card. You said that it should still work.

Chkdsk has not been running automatically. That particular instance was after using the disk in Windows 10, and then booting into Windows XP with the disk still connected. Chkdsk did not run on boot, but I immediately got error balloons in the system tray saying that files on T: were corrupted, and telling me to run chkdsk. That readout was the result. There had been no sign of errors when I left Windows 10. I will check that BIOS setting. It's certainly at the default setting, because I've never changed any of those parameters.

@Andalu @jaclaz What I'm thinking of doing now is to buy another Silicon Image or similar PCI card, without an eSATA connection, but with two internal SATA connections on the card. I can then connect the Blu-ray drive to the extra connection on that card (my archive disk is already connected to the single internal connection on the card I'm already using). Then only the 3TB drive would be connected to the Asmedia card, which should be fine. If the new card treats the Blu-ray drive as the other PCI-E SI card did, that is to say not as a removable drive, all should be fine. HOWEVER I've now come up against a much bigger problem! If I use the 3TB in Windows 10 and then go to use it in Windows XP, or vice versa, there are a huge number of file system errors on it! I'm having to repeatedly run chkdsk to clean it up before I can use it. There are a truly massive number of errors found every time, which seem to be all fixable, but is obviously hardly ideal! This is after doing a backup to the drive in Windows 10 (where it showed clean) to then trying to use it in XP. Checking file system on T: The type of the file system is NTFS. Chkdsk cannot run because the volume is in use by another process. Chkdsk may run if this volume is dismounted first. ALL OPENED HANDLES TO THIS VOLUME WOULD THEN BE INVALID. Would you like to force a dismount on this volume? (Y/N) Volume dismounted. All opened handles to this volume are now invalid. Volume label is BACKUP. Attribute record of type 0x80 and instance tag 0x7 is cross linked starting at 0xc68a9c6 for possibly 0xd clusters. Attribute record of type 0x80 and instance tag 0x7 is cross linked starting at 0xc68a9c6 for possibly 0xd clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x7 in file 0x14e00 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85504. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9eb7ec for possibly 0x135 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9eb7ec for possibly 0x135 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e1a is already in use. Deleting corrupt attribute record (128, "") from file record segment 85530. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ebc13 for possibly 0x2ba clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ebc13 for possibly 0x2ba clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e20 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85536. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ec358 for possibly 0x1ec clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ec358 for possibly 0x1ec clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e23 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85539. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ec83b for possibly 0x241 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ec83b for possibly 0x241 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e26 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85542. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9eca7c for possibly 0x23 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9eca7c for possibly 0x23 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e27 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85543. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ecaa0 for possibly 0x21a clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ecaa0 for possibly 0x21a clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e28 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85544. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ecece for possibly 0x1bf clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ecece for possibly 0x1bf clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e2c is already in use. Deleting corrupt attribute record (128, "") from file record segment 85548. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed08d for possibly 0x79 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed08d for possibly 0x79 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e2d is already in use. Deleting corrupt attribute record (128, "") from file record segment 85549. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed107 for possibly 0x63 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed107 for possibly 0x63 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e2e is already in use. Deleting corrupt attribute record (128, "") from file record segment 85550. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed2ee for possibly 0x1a clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed2ee for possibly 0x1a clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e39 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85561. Attribute record of type 0x80 and instance tag 0x7 is cross linked starting at 0xc950dde for possibly 0x19 clusters. Attribute record of type 0x80 and instance tag 0x7 is cross linked starting at 0xc950dde for possibly 0x19 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x7 in file 0x14e3f is already in use. Deleting corrupt attribute record (128, "") from file record segment 85567. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed372 for possibly 0xa clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed372 for possibly 0xa clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e40 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85568. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed37c for possibly 0x7 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed37c for possibly 0x7 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e41 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85569. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed383 for possibly 0x15 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed383 for possibly 0x15 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e42 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85570. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed398 for possibly 0x1f clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed398 for possibly 0x1f clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e43 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85571. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3b7 for possibly 0x1f clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3b7 for possibly 0x1f clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e44 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85572. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3d6 for possibly 0x19 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3d6 for possibly 0x19 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e45 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85573. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3ef for possibly 0x14 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed3ef for possibly 0x14 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e46 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85574. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed403 for possibly 0x22 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed403 for possibly 0x22 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e47 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85575. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed425 for possibly 0x17 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed425 for possibly 0x17 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e48 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85576. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed43c for possibly 0x26 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed43c for possibly 0x26 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e49 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85577. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed463 for possibly 0xb clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed463 for possibly 0xb clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e4a is already in use. Deleting corrupt attribute record (128, "") from file record segment 85578. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed46e for possibly 0x23 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed46e for possibly 0x23 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e4b is already in use. Deleting corrupt attribute record (128, "") from file record segment 85579. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed491 for possibly 0xd clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed491 for possibly 0xd clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e4c is already in use. Deleting corrupt attribute record (128, "") from file record segment 85580. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed49e for possibly 0x27 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed49e for possibly 0x27 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e4d is already in use. Deleting corrupt attribute record (128, "") from file record segment 85581. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed937 for possibly 0xa1 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ed937 for possibly 0xa1 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e60 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85600. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ede8f for possibly 0x34 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ede8f for possibly 0x34 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e72 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85618. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9edec3 for possibly 0x9b clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9edec3 for possibly 0x9b clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e73 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85619. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9edf5e for possibly 0x82 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9edf5e for possibly 0x82 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e74 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85620. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee183 for possibly 0x36 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee183 for possibly 0x36 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e7b is already in use. Deleting corrupt attribute record (128, "") from file record segment 85627. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1b9 for possibly 0x11 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1b9 for possibly 0x11 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e7c is already in use. Deleting corrupt attribute record (128, "") from file record segment 85628. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1ca for possibly 0xe clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1ca for possibly 0xe clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e7d is already in use. Deleting corrupt attribute record (128, "") from file record segment 85629. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1d8 for possibly 0x2c clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee1d8 for possibly 0x2c clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e7e is already in use. Deleting corrupt attribute record (128, "") from file record segment 85630. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee205 for possibly 0x4c clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee205 for possibly 0x4c clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e7f is already in use. Deleting corrupt attribute record (128, "") from file record segment 85631. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee251 for possibly 0x5a clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee251 for possibly 0x5a clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e80 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85632. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee2ab for possibly 0x3 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee2ab for possibly 0x3 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e81 is already in use. Deleting corrupt attribute record (128, "") from file record segment 85633. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee2af for possibly 0x61 clusters. Attribute record of type 0x80 and instance tag 0x4 is cross linked starting at 0xc9ee2af for possibly 0x61 clusters. Some clusters occupied by attribute of type 0x80 and instance tag 0x4 in file 0x14e82 is alread For more information, see Help and Support Center at http://localhost:90/redirect.php. And that's just the errors which were recorded in the Windows log, I'm sure there were many many more! I really don't know what to do about this.

Thanks @Andaluand @jaclaz. One way around this would be for me to connect the Blu-ray drive directly to the motherboard, and instead connect another of my drives to the card, which is in an internal swappable caddy. I have two other drives in caddies which sometimes go in there. As that is a physically removable drive, that would work around the problem, and give me the advantage that I could swap the drive in the caddy for another one without rebooting as I have to do now. The disadvantage of that configuration would be that I would lose access to the internal caddy drive in Windows 98, unless you can tell me that there's a Windows 98 compatible driver for the Asmedia card, which I very much doubt! I guess compromises are always inevitable with a complex multi-boot system! I'm obviously very glad that I can now access in Windows XP the 3TB disk in the external enclosure. @jaclaz is that eSATA to SATA adapter you linked to active or passive? I looks like it's probably passive to me. I may be wrong, but I've always assumed that apart from the physical connector being different, eSATA and SATA are the same thing, the only difference being that eSATA connections go through different electronics to allow hot-swapping, which can't be done with standard SATA connections. As I said earlier, I think the SATA ports on the Asmedia card are actually eSATA connections, with a passive jumper switchover between the two socket types, and my experience now would seem to bear that out. After all there's no reason why a SATA connection should not use eSATA hot-swap capable electronics, even if it's permanently connected. Obviously that doesn't apply the other way around! If it's always treating all connected devices as removable devices, that also seems to bear out to me that the supposed SATA ports on the card are actually not that at all, they are eSATA ports pretending to be SATA ports! The fact is that I don't want the system to treat my Blu-ray drive as a removable and non-bootable device. If I have a problem and need to boot into a recovery environment with a boot DVD, I'm going to be in trouble with that configuration unless I physically reconnect the drive directly to the motherboard, which I can do of course, but it's a bit of a faff!

Thanks @Andalu. The hard drive in the enclosure does seem to be recognised correctly and works fine. It is connected to one of the SATA ports though, not one of the eSATA ports. My Blu-ray drive is connected to the other SATA port on the card. The version of asahci32.sys installed is 2.0.3.1, is that the wrong one? Strange about the DVD/Blu-ray drives. On my system, it does not appear as a bootable device when connected via the Asmedia card. It did appear as a bootable device when connected via the previous Silicon Image card though.

Thanks, that's good to know!

Just a couple more observations about Supermium on 32 bit XP. The GUI font looks a bit rough, I don't know if that can be improved, but it's usable. The good thing is that it happily opens the sites that 360Chome won't now render! I had a bit of a laugh when I came to add an image to a post here, to see this - The file open dialogue looks like something from a 16-bit Windows 98 program! It doesn't work either. Drag and drop does work. The browser is rather slow opening, and a bit slow in use. I won't be abandoning 360Chrome quite yet, but it's still great to have a browser on XP which will open the sites that 360Chrome now has trouble with. Supermium used my 360Chome profile files absolutely fine, which was a big bonus!

@Andalu @jaclaz OK, back at last! The correct card finally arrived, and I put it in yesterday. The first problem I ran into, which I should have anticipated of course, was that there was actually no physical room to get the riser in! It would need quite a big space between the installed cards to get it in between two of them, and it just isn't there. I looked at mounting the riser elsewhere, and running a cable to it, but I then thought that maybe, as they are just changed over by passive physical jumpers, the SATA and eSATA sockets on the new card are in fact electronically the same, just with different connector types. So, I put the card in where the other Silicon Image one had been, where the eSATA ports are physically inaccessible, and connected my Blu-ray drive and HDD enclosure to the two SATA ports. Fortunately, I did have a backplate with two eSATA sockets on it with cables to connect them to internal SATA ports, so I could use that for the HDD. It works! I'm pretty sure that this would not have worked with the Silicon Image card, as I did try connecting the HDD enclosure to the SATA port on that, and it didn't work as an eSATA port, i.e. with hot-swap capability. With the new card, it seems to be fine. So, all's well that ends well with that, I've just now got two risers which I don't need, but they were pretty cheap, so I'm not too worried about it! So, the physical problem seems to be sorted, but there are still a couple of configuration anomalies (aren't there always?!) I used the recommended v2.0.4 driver on XP. The disk that came with the card claimed to have an XP compatible driver on it, but I didn't use that because @Andalu said the v2.0.4 driver was the best one to use here. I did use the later driver on the disk (2.0.8 from 2014) for Windows 10, which is fine. The problem is that the Blu-ray drive and the HDD are being seen as removable devices, with a 'Safely Remove Hardware' icon now permanently in the system tray on both XP and 10. That in itself is not a real problem of course, certainly not with the HDD, which isn't always present anyway, but the Blu-ray drive is now not appearing in the boot devices list in the machine BIOS, which means that I wouldn't be able to boot from it. That could indeed be an issue if I ever needed to, of course. This did not happen with the Silicon Image card, when the Blu-ray drive was connected to the SATA port on that it appeared normally and was in the BIOS boot devices list. There is an option in Device Manager on XP for the card, which is presumably supposedly to remove the 'Safely Remove Hardware' option. It was enabled by default. I've tried switching it off, as you can see, but it's made absolutely no difference, the 'Safely Remove Hardware' icon is still there! That option is not there at all on the Windows 10 driver. Any ideas? Thanks, Dave.

Just to quickly report that I've now had a go with Supermium on Windows XP 32 bit, and it seems to work very well! Early days of course, but it looks very hopeful.

Yes, I'm sure that's right. I did wonder whether the USB lead connection could carry power as well as the data. I can try it out with my existing card, and I might do that tomorrow just to confirm.

Getting there! I now have two 'risers', and the correct eSATA card is now on its way to me! Just a quick question about the risers. They both have power connectors on them. One has more than the other, and that's the one I will use as it's physically slightly lighter. Am I right in assuming that those power connectors are there because the unit is designed to work with power-hungry powerful gaming graphics cards, which need more power than the PCI-E connector can provide? If so I guess I almost certainly won't need to connect them for my use.

The saga continues, I'm afraid. The people I ordered the card from are still trying to let me know whether their warehouse can actually supply the card I ordered! The PCI-E 'extender' I ordered still hadn't arrived several days after the last quoted delivery date. I reported this to eBay, and sent a message to the seller. It was supposedly despatched on January 31st, but is not recorded as delivered. I heard nothing, until I got a message from eBay saying my money had been refunded. I then ordered another one from a different source. Then, I got a message from the original seller apologising for the fact I'd not received the first one, saying he'd refunded the money, and he was sending another one! Oh dear! So, it looks as if I'm going to end up with two extenders, possibly three if the original one now arrives! I'll get there eventually!

Hi Dietmar. I don't have the power to do that I'm afraid. Tripredacus the forum supervisor may be able to do that, but the attachment quota for standard members is intrinsically limited I think. You will need to delete some older attachments to free up space. Cheers, Dave.

I'm rather hoping that I don't have to do that!

This is the half height bracket that was sent with the incorrect card I received. It should do the job I think, as the 'top bit' is definitely shorter than on a normal bracket, so hopefully the card will hang in the air high enough above the PCI-X socket below it to enable the PCI-E extender to go underneath it. I have spoken to the people I ordered the card from, and sent them a picture of the incorrect card. Hopefully they will come back soon to let me know if they can send the right one. Naughty I know, but if they want the other card back, I'm going to quietly keep that half height bracket just in case I can't find the right card supplied with one!

Ah, OK. I hope the Startech one I ordered will do the job. I'm not intending to try and boot from it. Cheers, Dave.

Is that the one I've ordered?