darry wrote on 2020-10-04, 04:38:

If there was a way to force an SD or CF card to return a lower capacity than actual, it would probably make my life easier (any suggestions ?).

I don't know the specifics, but judging from the existence of the old scam industry based around hacked cards misreporting their size, I am guessing this won't be too difficult....

Shreddoc wrote on 2020-10-04, 06:17:

darry wrote on 2020-10-04, 04:38:

If there was a way to force an SD or CF card to return a lower capacity than actual, it would probably make my life easier (any suggestions ?).

I don't know the specifics, but judging from the existence of the old scam industry based around hacked cards misreporting their size, I am guessing this won't be too difficult....

I have similar issues with an old Pentium with GA586S mainboard and a CF adapter in a removable IDE drive bay.
Most if not all CF cards tested fail to be recognized by the BIOS, except for a 512MB CF card which works fine.
The other cards are NOT fake, though. They work as intended in USB card readers and digital cameras. I can only guess what's "wrong" with them. 5v tolerance? ATA spec too new? Don't support True-IDE mode? Hm. 🤨

Edit: The other cards tested are between 8MB (no typo) and 8GB.

Jo22 wrote on 2020-10-04, 12:00:

I have similar issues with an old Pentium with GA586S mainboard and a CF adapter in a removable IDE drive bay. Most if not all C […]
Show full quote

Shreddoc wrote on 2020-10-04, 06:17:

darry wrote on 2020-10-04, 04:38:

If there was a way to force an SD or CF card to return a lower capacity than actual, it would probably make my life easier (any suggestions ?).

I don't know the specifics, but judging from the existence of the old scam industry based around hacked cards misreporting their size, I am guessing this won't be too difficult....

I have similar issues with an old Pentium with GA586S mainboard and a CF adapter in a removable IDE drive bay.
Most if not all CF cards tested fail to be recognized by the BIOS, except for a 512MB CF card which works fine.
The other cards are NOT fake, though. They work as intended in USB card readers and digital cameras. I can only guess what's "wrong" with them. 5v tolerance? ATA spec too new? Don't support True-IDE mode? Hm. 🤨

Or possibly, the device being fussy about the formatting / partition table / MBR - this can happen even with PCs and certain cards/formattings. E.g. my latest card would only boot my DOS machine after I first prepared it with Rufus/FreeDOS. Any other common format or partition method would fail to boot properly with that card in that machine. But once it was done I could then partition and format the card however I wanted, using the machine itself, and it was then fine forevermore.

Overall, the various CF specification revisions, multiplied by the various reader/device firmwares, multiplied by the various formats and partition setups that a card can assume, means there are many possible vagary combinations, so I am not surprised that it is common for people to find incompatibilities.

I have an explanation for the VX-880EX always wanting to re-format the FC-1307 based adapters . It's not the VS-880EX's fault or an incompatibility, it's the adapters themselves !

Using Windows 10, if I only create a 1GB or more 1GB FAT16 partitions on an FC-1307 adapted microSDHC card, at power-cycle (unplug and re-plug USB to CF adapter), the partitions are wiped!
If I create an NTFS partition, say 9GB in size and then a 1GB FAT16 partition, no wiping occurs .
If I create an an unformated partition, say 9GB in size and then a 1GB FAT16 partition, the wiping occurs .
If I create a either a FAT32 or NTFS partition that spans the entire card, no wiping occurs .

I have the same behaviour with 2 different FC-1307 based microSDHC to CF adapters and with all card sizes I have tried (8GB,16GB,32GB). I have owned those adapters for several years, so they might just have old buggy firmware .
Maybe they will work OK with the 2GB cards that I have on the way .

I just received a 2GB Verbatim CF card and a pack of 5 generic 2GB microSD cards .

The 2GB Verbatim CF card behaves like the Sandisk cards (freezes at boot).

The generic microSD cards behave like all the other cards in the FC-1307 based adapters I have .

<RANT> Why does this have to be so complicated ? I can't keep buying adapters and cards left and right hoping that something will work . I am getting really close to trying an IDE to SATA adapter along with a SATA SSD in that thing </RANT> .

EDIT : The 2GB microSD card I used tests out fine with h2testw.exe .

The fussy behaviour I noted above with my setups, was also using an FC-1307 controller based adapter.

I guess you just need to find something that works, then stick with that combination.

Low quality flash cards (e.g. from AE) can also complicate matters, so it may be best to avoid those.

Shreddoc wrote on 2020-10-04, 23:29:

The fussy behaviour I noted above with my setups, was also using an FC-1307 controller based adapter.

I guess you just need to find something that works, then stick with that combination.

Low quality flash cards (e.g. from AE) can also complicate matters, so it may be best to avoid those.

The microSDXC card I tried with the FC-1307 were an 8GB Kingston, 16GB Sandisk and 32GB Verbatim . The Sandisk was from Amazon.ca and the 2 others were purchased at reputable retail stores, so unlikely to all be counterfeits .

Having researched this a bit, it seems that FC-1307 partition wiping behaviour is not uncommon :

https://askubuntu.com/questions/1252233/shutd … r-clean-install
https://www.amazon.ca/Optimal-Shop-Digital-Ad … customerReviews (DarkGalXE's comment)

The tests that I have run seem to indicate the issue is related to the partitioning scheme used. Additionally, I have confirmed that as long as the adapter is not power cycled (i.e. powered externally), no wiping occurs, which is consistent with what is reported in the referenced links . If this was ever fixed in a firmware update is unknown to me and even if it was, it's not like firmware version is usually advertised on these things .

I would love to try something that works, preferably a CF to SD/microSD adapter or an IDE (44 or 40-pin) to SD/microSD, but nearly everything uses an FC-1307 or something undisclosed (probably an FC-1307 anyway). The only alternative I have seen is an INIC-2051 based adapter and that one is completely incompatible .

My reason for preferring an SD/microSD adapter is parts availability (CF is pretty niche at this point), but I would probably settle for a stack of 3 or 4 compatible CF cards of 2GB-4GB or more (large size does not seem to be an issue) for use and as spares .

I dug up an old SD to SATA adapter in my junk pile. It uses an Fc1306T (no idea if this is related to the FC-1307) for the SD to IDE conversion and a Sunplus chip for the IDE to SATA conversion (performance was terrible, around 7MB/sec max sustained, AFAICR). I also happened to have a spare Jmicron JM20330 IDE to SATA converter handy . I decided to combine the two and see what happens .

So, starting from the VS-880EX :
IDE to SATA (using Jmicron JM20330)
SATA to IDE (using a Sunplus chip)
IDE to SDcard (using FC1306T) .

Obviously, I powered this thing unholy monstrosity externally (I did not want to find out how much current a wire in an IDC cable can carry).

AND IT WORKED !!! It formatted properly and no disk wiping occurred on full power down/power up .

This gave me hope that an to IDE to SATA bridge would work with an actual SSD . So I ordered a 44-pin IDE to SATA adapter (SinLoon) and an ADATA 120GB SSD . I will be able to test it easily on external power and, once my Roland disk adapter/caddy arrives, I should be able to power it internally and, if everything works, actually use the VS-880EX .

This has actually been way more painful than getting the Akai DPS12 working . Ironically, the VS-880EX is newer .

Yes the SD adapters seem to all be the same fundamental design; a design which is not very broadly compatible with different card configurations and partitionings.

The SATA->IDE converters are likely to be based on more robust design standards, thereby enabling greater compatibility.

Shreddoc wrote on 2020-10-05, 01:14:

Yes the SD adapters seem to all be the same fundamental design; a design which is not very broadly compatible with different card configurations and partitionings.

The SATA->IDE converters are likely to be based on more robust design standards, thereby enabling greater compatibility.

SD to CF using the FC-1307 is definitely wonky in certain scenarios . Why it even cares about partitions is a mystery to me .

PCs may not be as picky regarding CF card compatibility, but I feel that I made the right choice forgoing CF and especially SD converted to CF in my retro builds . When CF media finally disappears from the market and people start experimenting more with SD converted to CF, they are going to be in for a world of hurt unless either non buggy FC-1307 implementation become a thing or more reliable alternatives appear. Neither of those scenarios seems likely, IMHO .

I am very very slowly working on an ISA+VLB IO card that can definitely use SATA, USB mass storage and probably SD cards too (and drive emulation), with bus mastering for maximum performance, it would put the end to this sort of storage problem. Alas time and money are low, other commitments must come first...

Tiido wrote on 2020-10-05, 01:43:

I am very very slowly working on an ISA+VLB IO card that can definitely use SATA, USB mass storage and probably SD cards too (and drive emulation), with bus mastering for maximum performance, it would put the end to this sort of storage problem. Alas time and money are low, other commitments must come first...

That sounds like it would be really useful in the long term, especially if its ISA support includes 8-bit mode for XT class machines. It could be a one-size fits all storage solution for anything pre-PCI, which would be a fantastic thing to have . I hope time and finances allow you to pursue this eventually .

I have had second thoughts on that Adata SU655 that I ordered . It is apparently prone to overheating and I cannot find a disclosed power draw for it . I guess I will end up using as a testing device or spare . I ordered an a Kingston A400 120GB that has a max power draw (0.195W Idle / 0.279W Avg / 0.642W (MAX) Read / 1.535W (MAX) Write , according to https://www.kingston.com/us/ssd/a400-solid-state-drive ) that's actually less than half that of some the officially recommended 2.5mm" IDE hard drives (3.5W) for the Roland VS family . So even with an IDE to SATA converter, I should not be taxing the VS-880EX's power delivery abilities .

EDIT : With a max of 4GB usable on the VS-880EX and an endurance of 40TBW on the SSD, I should be good for filling up that VS-880EX about 10240 times . If I did that once a day, I should be good for about 28 years . Considering that I mainly intend to use this as mixer, there should be no immediate concerns about endurance . I do wonder a bit about lack of TRIM, though .

EDIT2: Considering the bandwidth limitations of the VS-880EX, even if lack of TRIM eventually leads to slower max throughput to the drive, it will likely be unnoticeable in this use case .

Another day, another surprise . The Adata SU655 120GB SATA SSD, when connected through a JMicron JM20330 IDE to SATA bridge, has the same behaviour as the Sandisk CF cards, the Verbatim 2GB CF card and the Kingston 16GB CF card (freezes at boot) .

I dug up a SATA 2.5" Samsung ST1000LM024 1TB hard drive, tried it through the JM020330, and it works fine (externally powered as it has a 4.25W draw according to label)!

The VS-880EX just does not like the Adata SSD (some might say that it has good taste), but it has no trouble running through cascaded IDE to SATA and SATA to IDE bridges! If the VS-880EX behaves the same way with a Kingston A400, I may just go insane(r)!

As a last resort, could just source a bunch of cheap, compatible 2.5" spinning drives, and churn through those over time. This is an era when such drives are quite cheap, due to the scrap supply from the past decade+'s notebook computers, but that cheap era will not remain so forever.

Shreddoc wrote on 2020-10-05, 23:03:

As a last resort, could just source a bunch of cheap, compatible 2.5" spinning drives, and churn through those over time. This is an era when such drives are quite cheap, due to the scrap supply from the past decade+'s notebook computers, but that cheap era will not remain so forever.

That is definitely an option, but It may not come to that. Out of curiosity, I tried it with a 500GB Samsung 860 Evo and it appears to work (did not try formatting, but at least it did not freeze during boot and let met get to the formatting prompt)!

Apparently, the VS-880EX "likes" some SSDs but "hates" some others . I would prefer no to use an 860 Evo because of its power requirements (4W) and because it's waste to use a minimum 250GB premium (expensive) SSD in a 4GB application. I may just get lucky with the Kingston A400 . Fingers crossed .

The Adata SU655 that I have (there appear to be variants) is based on the Maxiotek/Maxio (JMicron spin-off) MAS0902A-B2C .

The Kingston A400 apparently uses a Phison S11 controller .

If the A400 does not work, I do have a few SSDs in service at my place that I could half-heartedly test to try to find which controller(s) the VS-880EX likes . At least I know it likes Samsung 860 EVO SSDs, so that's a fall-back plan (I could always re-purpose a 250GB one after upgrading it, but it still feels like a waste).

EDIT: My initial plan was to be able to use commodity, inexpensive flash cards. I have ordered 2 more generic SD to IDE converters on the off chance they might have less buggy behaviour than my microSD to CF units .

Kingston A400 120GB (likely Phison S11 based, did not disassemble it) : has the same behaviour as the Sandisk CF cards, the Verbatim 2GB CF card and the Kingston 16GB CF card (freezes at boot) .
Kingston KC600 256GB (Silicon Motion SM2259) : has the same behaviour as the Sandisk CF cards, the Verbatim 2GB CF card and the Kingston 16GB CF card (freezes at boot) .
HP S700 250GB (Silicon Motion SM2258XT) : has the same behaviour as the Sandisk CF cards, the Verbatim 2GB CF card and the Kingston 16GB CF card (freezes at boot) .

So apparently, the only SSD that seems to work (still needs to be fully tested) so far is a Samsung 860 EVO .

Just to be clear, I did not return any of these test purchases . IMHO, they are not defective and should not be returned . My compatibility issues in a very specific/odd use-case should not be of concern to the vendor, again IMHO .

EDIT: The VS-880EX also seems to like a Kingston HyperX 240GB SHFS37A/240G (SandForce®SF-2281 based). It gets to the formatting screen .

Darry,

Make sure you have storage disk (not removable) bit set in CF cards. That shows the issues I had with CF cards. Find 2GB or 4GB CF and do a test on a slot 1 machine, by using DOS and prepare them for bootable and if they boot in DOS, might work in your VS-880EX.

Ones that works:
Transcend
Industrial grade CF cards.
Swissbit 8GB CF

Cheers,