erichelgeson / bluescsi Goto Github PK

HI,
I have 2 BlueSCSI ver.1.0c
with same firmware (v1.1-20221203)
with same microSD card and File
the log on both the same result

one working normal and the other shown the blueSCSI as 8 times (2 as Disk Drive , 6 as QUANTUM )

What can I do to fix that

Currently since BlueSCSI is reporting as a QUANTUM FIREBALL1 and SCSI Director 4 tries to issue a vendor specific request for sense page 0x39. Since this is not handled we return an illegal sense request - which SCSI Director shows to the user. The user can just hit OK to continue and the format completes fine.

We could either:

• Find the vendor documentation on the format of the 0x39 sense page response and send that back.
• Identify as a different drive that SCSI Director does not send vendor requests for, but is still in the Apple approved drives.
• Send back all 0's (this does work, but we dont know the format or what this means, or how long it should be)
• Nothing, people don't really use SCSI Director to format, and if you hit OK it works anyways.

Mostly logging this issue for someone (myself included) in the future to revisit.

The like to the 3d printable mount files for the 1.1 desktop version is dead: https://github.com/erichelgeson/BlueSCSI/blob/main/3D-Files/1.1/bluescsi11a_kerosmacmodsv2.stl.zip

I have a Quadra 840AV for which I'm trying to prepare a drive on the BlueSCSI to store games. I made it 30GB using dd and formatted it on my Quadra using 'Drive Setup 1.7.3'. I installed many games on it (over 20GB) using Basilisk II, but when trying to use the disk on my Quadra, first of all it's really slow (it takes a couple of seconds to open the drive and scrolling also takes a few seconds), and the files seem to be corrupted. Whenever I try to run a game, I get an error that there's not enough memory, with zero available! Mounting the image again on Basilisk II, everything works fine, meaning that there is no corruption.

The version I'm using is 1.1-20211215, which is quite old, but I'm waiting for my flasher to arrive so that I can try newer versions. Not sure if there are any possible fixes in later versions.

Not sure if the bug can be reproduced easily. I'll try creating a new large image and see if I get the same problems with a single file.

When using BlueSCSI with the internal host adapter of a MegaSTE I could not figure out whether BlueSCSI supports ICD compatible command semantics or not. The results I got were inconclusive.
I'd like to refer to RaSCSI ticket PiSCSI/piscsi#637 which describes these semantics. The implementation was straightforward because essentially just the first (custom) command byte had to be ignored. hal/gpiobus.cpp in the rascsi project has the details.
Without ICD compatible semantics drives with a capacity of more than 1 GB can only be used partially, for instance.

Hello,

I've successfully used the "Desktop v1.1" version of BlueSCSI with firmware v1.1-20220626 as a replacement for the internal SCSI disk in an Atari TT030 (with the stock Motorola 68030 32MHz and TOS3.06).

Maybe you could therefore update the list of Supported Systems to also include the Atari TT030 as "Working".

Thanks a lot for this awesome device btw.!

• just a data point and to keep a record of this on GitHub
• I know there's not much point running this on a Classic 68K but one has to at least try :)

My Macintosh Classic setup shows "disk question mark" when running either 96MHz or 128MHz overlocked firmware. It's fine with the standard firmware.

BlueSCSI was flashed OK with no errors (though the DFU "Resetting USB to switch back to runtime mode" took a lot longer after flashing the overlocked firmwares) but the Mac won't see any drives through it.

• I bought my BlueSCSI from flamelily UK if that helps 2022-04-11.
• Let me know if I can provide any additional logging.

According to https://github.com/erichelgeson/BlueSCSI/wiki/Usage , one can use up to 7 images on one SD, thus reserving one ID for SCSI controller. But current implementation assumes that ID 7 is used as controller ID:

if(id < NUM_SCSIID && lun < NUM_SCSILUN) { dev = &scsi_device_list[id][lun]; ...
Since NUM_SCSIID == MAX_SCSIID == 7, only IDs below 7 (i.e., 0-6) are processed as images' IDs. Some Akai samplers by default use ID 6 as internal one which limits total number of possible images to 6 on these platforms. Although these samplers support changing their ID even on the fly, default ROM image which is always booted first during system startup (actual OS with another configured ID can be booted later from FDD or HDD), has hardcoded value of 6. Having both sampler ID set to 6 and a BlueSCSI device with ID 6 during boot time causes some weird issues on a sampler unless internal ID is manually changed even in case the OS booted later has another ID set as internal.

1. How to use QMK for this?
2. Which firmware, there's several to choose from?
3. Am I plugging in to the Bluepill, or using a ST-Link?
4. Do I need to change the jumpers on the Bluepill?

Is there a way to o/p SCSI cmd traces as per V2 when BlueSCSI.ini "debug = 1" is enabled to the log file on v1 ?

I am trying to debug BlueSCSI on RM Nimbus ( an MSDOS 80186 pc with old SASI / SCSI 1 controller, that does not support parity).

I suspect that part of my problems are to do with the lack of SCSI BUS drivers on V1 board and the inability of the STM32 to sink the required 38 mA per i/o pin to meet the SCSI BUS specs (The STM32 has a limit of 25 mA -> '0' voltage is too high > 0.8 v).

I would like to confirm that the logic is otherwise fine, if it is I'll hack a V1 to add a 74ALS638-1N 48mA SCSI BUS driver.

Passive term on controller <-> passive term on last drive > 34 mA sink (for 0.8V '0')
Passive term <-> active term > 29 mA sink (for 0.8V '0)
Active term <-> Active term > 29 mA sink

All of which is out of spec for STM32 (max 25 mA sink)

Peter

Hi,
In H.W. Ver 1.0c desktop ver.
what is the need for the 2 jumpers ( term ,on)
and When should it be used?
and if used them must use both or can use one of them only

Thanks

This problem occurred in the compilation today~ I tested the previous version, only 1.0 can be compiled normally, and the others are this problem. I used to be able to compile normally

Compiling` .pio\build\STM32F1\libe88\SdFat\common\upcase.cpp.o
src\BlueSCSI.cpp: In function 'void readSDCardInfo()':
src\BlueSCSI.cpp:199:27: error: 'cid_t' {aka 'struct CID'} has no member named 'mdt_month'; did you mean 'mdtMonth'?
  199 |     LOG_FILE.print(sd_cid.mdt_month);
      |                           ^~~~~~~~~
      |                           mdtMonth
src\BlueSCSI.cpp:201:27: error: 'cid_t' {aka 'struct CID'} has no member named 'mdt_year_high'
  201 |     LOG_FILE.print(sd_cid.mdt_year_high);
      |                           ^~~~~~~~~~~~~
src\BlueSCSI.cpp:202:29: error: 'cid_t' {aka 'struct CID'} has no member named 'mdt_year_low'
  202 |     LOG_FILE.println(sd_cid.mdt_year_low);
      |                             ^~~~~~~~~~~~
src\BlueSCSI.cpp:205:29: error: invalid use of non-static member function 'uint32_t CID::psn() const'
  205 |     LOG_FILE.println(sd_cid.psn);
      |                      ~~~~~~~^~~
In file included from .pio\libdeps\STM32F1\SdFat\src/SdCard/SdCardInterface.h:28,
                 from .pio\libdeps\STM32F1\SdFat\src/SdCard/SdioCard.h:28,
                 from .pio\libdeps\STM32F1\SdFat\src/SdCard/SdCard.h:27,
                 from .pio\libdeps\STM32F1\SdFat\src/SdFat.h:32,
                 from src\BlueSCSI.cpp:39:
.pio\libdeps\STM32F1\SdFat\src/SdCard/SdCardInfo.h:298:12: note: declared here
  298 |   uint32_t psn() const {
      |            ^~~
*** [.pio\build\STM32F1\src\BlueSCSI.cpp.o] Error 1
=============================================================================== [FAILED] Took 6.81 seconds ===============================================================================

Environment    Status    Duration
-------------  --------  ------------
STM32F1        FAILED    00:00:06.811

Some systems expect sector sizes that are not 512 byte, e.g. IBM AS/400 expects 520 or 522 byte.
This is a feature request to add support for such sector sizes.
See prior analysis by PoC.

Hello,
I have a DB25 BlueSCSI from tom.scsi.blue, and while it works wonderfully with my Mac Plus, testing with Transoft SCSI Pro 4 finds an access time of consistently about 13 milliseconds. I did some research, and it seems that most quality SD cards have access times of 1 or 2 ms. I have tested both a 32 GB Sandisk Ultra and a 256 GB Samsung EVO Select, both formatted as exFAT, and gotten the same results.

I have a project that needs low access times, because it needs to read small amounts of data from the disk quickly. The "seek time" given by SCSI Director is about 2 ms (totally acceptable), and the read rate is about 430 kB/s. I'm using the r0w0s6 version of SCSI Accelerator.

I dug into the BlueSCSI code and it looks like all file reading is passed through SdFat, so I don't think the slowdown is caused by an issue there. SdFat is a reputable and high-speed library, even over SPI. The cards are also run at 50 MHz consistently.

The Retro Theory Twitter post shows access times of about 1.5 ms, and access time seems like it should be disk-limited, not computer-limited. Is there a solution? Or is this just a limitation of the early SCSI implementation in the Plus?

On my HP 9000 712 I can install NEXTSTEP 3.3 to BlueSCSI drive at ID6 (via BlueSCSI cdrom ID3), but when it tries to boot from it (ID6). It will fail due to it's default hard drive geometry given by BlueSCSI, if i manually set it to 139 Sectors, and 4 heads (code below) it then will boot fine.

    case SCSI_SENSE_MODE_FORMAT_DEVICE:  //Drive parameters
      m_buf[a + 0] = SCSI_SENSE_MODE_FORMAT_DEVICE; //Page code
      m_buf[a + 1] = 0x16; // Page length
      if(pageControl != 1) {
        m_buf[a + 11] = 139;//Number of sectors / track
        m_buf[a + 12] = (byte)(dev->m_blocksize >> 8);
        m_buf[a + 13] = (byte)dev->m_blocksize;
        m_buf[a + 15] = 0x1; // Interleave
      }
      a += 0x18;
      if(pageCode != SCSI_SENSE_MODE_ALL) break;

    case SCSI_SENSE_MODE_DISK_GEOMETRY:  //Drive parameters
      m_buf[a + 0] = SCSI_SENSE_MODE_DISK_GEOMETRY; //Page code
      m_buf[a + 1] = 0x16; // Page length
      if(pageControl != 1) {
        unsigned cylinders = dev->m_blockcount / (4 * 139);
        m_buf[a + 2] = (byte)(cylinders >> 16); // Cylinders
        m_buf[a + 3] = (byte)(cylinders >> 8);
        m_buf[a + 4] = (byte)cylinders;
        m_buf[a + 5] = 4;   //Number of heads
      } else {
        m_buf[a + 2] = 0xFF; // Cylinder length
        m_buf[a + 3] = 0xFF;
        m_buf[a + 4] = 0xFF;
        m_buf[a + 5] = 4;   //Number of heads
      }
      a += 0x18;
      if(pageCode != SCSI_SENSE_MODE_ALL) break;
    case SCSI_SENSE_MODE_FLEXABLE_GEOMETRY:
      m_buf[a + 0] = SCSI_SENSE_MODE_FLEXABLE_GEOMETRY;
      m_buf[a + 1] = 0x1E;  // Page length
      if(pageControl != 1) {
        m_buf[a + 2] = 0x03; 
        m_buf[a + 3] = 0xE8; // Transfer rate 1 mbit/s
        m_buf[a + 4] = 4; // Number of heads
        m_buf[a + 5] = 139; // Sectors per track
        m_buf[a + 6] = (byte)dev->m_blocksize >> 8;
        m_buf[a + 7] = (byte)dev->m_blocksize & 0xff;  // Data bytes per sector
      }

https://github.com/erichelgeson/BlueSCSI/wiki/Usage

This page says both:

The naming convention is as follows (file name max 64 characters).

And:

Bad

HD99_712 foo bar fizz buzz bang.hda Over 32 chars and invalid SCSI/LUN/Block size

What is the actual maximum filename length—32 or 64 characters?

Tested with the current firmware: When a SCSI driver sends REQUEST SENSE for LUN 7 to a BlueSCSI device configured with ID 0 and LUN 0 (i.e. no LUN specified in the image filename) BlueSCSI returns Sense Key $00, ASC $00.
The SCSI-2 specification says this shall be Sense Key $05, ASC $25:

7.5.3 Selection of an invalid logical unit

The target's response to selection of a logical unit that is not valid is
described in the following paragraphs.

The logical unit may not be valid because:
   a)     the target does not support the logical unit (e.g. some targets
          support only one peripheral device).  In response to an INQUIRY
          command, the target shall return the INQUIRY data with the
          peripheral qualifier set to the value required in 8.2.5.1.  In
          response to any other command except REQUEST SENSE, the target shall
          terminate the command with CHECK CONDITION status.  In response to
          a REQUEST SENSE command, the target shall return sense data.  The
          sense key shall be set to ILLEGAL REQUEST and the additional sense
          code shall be set to LOGICAL UNIT NOT SUPPORTED.

Hi everyone, hi Eric,

I'm supporting Luciano, a friend of mine who has these issues, the NextStations are not mine.
I personally have about 25 Macs but I can not help him, so I'm asking the kind support of the community here.

Machine tested: NeXTstation 68040/25 and NeXTstation Color 68040/25, both with 24 Mb RAM
Schede BLUEscsi: one and the Powerbook version (I can not tell the firmware version), pictures are attached.

OS: NeXTstep 3.1, NeXTstep 3.3, OpenStep 4.0.

Description of the two issues:

1. at startup Verbose Mode it identifies 6 LUN SCSI (all internal IDs are identified).
   Once the OS is loaded it asks to fox or format these 6 ghosts disks
   I can press ignore
   In “root” mode the Build Disk app identifies all these bogus disks with “0” bytes, but they are obviously non usable
   As a consequence any SCSI device I attach, either internal or external, is not recognised or mounted

2. If I switch user (logout and login) I have a system crash, it shows the spinning wheel forever, I need to force the reboot, this happens with any NextStation and any bluescsi card I have

can you help me please with this?

thanks so much

Pierfranco

I hereby attach two pictures of the verbose mode trace log

I tried to order some of these with jlcpcb but they sent back the following issue:

"As shown below, there's a small gap at the indicated position, during ptinting the gap might be expanded and caused crack or deformation for the part.

Will this risk be acceptable for you?"

On the Image Sets page,
https://github.com/erichelgeson/BlueSCSI/wiki/Image-Sets
the last paragraph mentions a pin named "BA1" - was this supposed to be PB1 ?

Also, are pull-up resistors needed for the two pins ? ( due to the mention of a 3x2 block header instead of 2x2)

Thank you !

Hi,

I've got an Atari ST with an AdSCSI micro board from ICD.
(Or more accurately, it's a board that's been reverse engineered by t0ri, and as far as I know, is functionally an exact replica)

I've tested and used it with SCSI2SD v5.0 and the board works perfectly, so, I know the ST and the converter work ok.
(Transfer rate is benchmarking at around 960Kb/sec, and is perfectly fine with the SCSI2SD v5, with 3x HDD and one CDRom)

The ST boots up, but, the activity lights come on, and stay on, on the BlueSCSI, and once you get to the desktop, no HDD is accessible, and neither are either floppy (A:\ or B:, connected separately, or together.)
(So, without the HDD booting, and floppy disabled, I can't load any software to detect the drive/adapter and partition etc)

I noticed there is a "MegaSTE" mode added in an earlier build where you can swap the ID and LUN, and I've tried that, but, it made no difference.
(The AdSCSI Micro only supports SCSI ID 0,1,2,3. It's designed for one single device)

The BlueSCSI v1.1 I built myself, but, I've tried it in an Amiga A2000 with GVP Zorro II card (SCSI+8mb fast) and, it's been working perfectly in that machine for months.
I even built a second BlueSCSI, and that too works in the GVP but not with the AdSCSI Micro.

The HDD Image I've tried is an 8GB image, one single HDD. Tried with 256MB, 512MB and so on, too.

The same image works fine on Hatari, and written to the SD Card on the SCSI2SD, not that it matters, since even a blank image fails to boot, and leaves the floppy disabled.

So, to that end, is there a debug build of the latest release?

I just tried setting up a SGI Indy on a 20220404 based 1.1b board I just assembled. With just one file on the exfat formatted sd, HD10_512IndyIrix53.hda, at boot I get a bad error :

sc0,1,0: cmd=0x12 Unexpected info phase 47, state 40. Resetting SCSI bus

this message loops for a while and the PROM Diagnostics fails.

A scsi2sd seems to working ok.

Is there anything I can do to provide more log/debug ?

Btw, I have quite a few "exotic" scsi machines, from HP, Sun, SGI, Commodore, Atari, NeXT, Apple or Roland, I may come with some other test results over time, don't know if you'd like me to test some in a certain order, just tell me ! I have also some coding knowledge but could not find any scsi protocol documentation in a digestable format, do you have any pointer ? I may help.

Anyway thanks a lot for your work, I really appreciate it.

I used it in Korg Triton Pro X
Bluescsi is formatting in Korg Triton Pro X that can be stored normally. But when I want to copy the files from the simulator, MAC's simulation prompts formatting。Is there any way to copy the file inside?
.hda file sent to your email

I have an old Korg D16 and want to upgrade the firmware on it. Unfortunately the only way to do that is with a SCSI CDR/CDRW. I can attach the UDF cdrom image to D16 via the BlueSCSI and it is recognized, but the silly machine will not read it unless it is a CDR or CDRW (Calls it an illegal device). Also getting the WAV files off the machine is only possible via burning them to a SCSI CDR/CDRW (The disk drives use some proprietary format on the machine).

Not sure if this is possible, but I am sure this is not the only device that could benefit from this.

On a better note, I was able to format hda scsi drives on the Korg D16 presented by the blueSCSI. The internal drive is dead and until I replace it I cannot test copying/recording files to it (Should be here this weekend).

I cant for the world of me get the extended led lights to work. Got the ports and the resistors on the bluescsi but then, how do I get it to work, please explaine in laymansterms

The file at https://github.com/erichelgeson/BlueSCSI/blob/main/hw/external-db25/gerbers/bluescsi-1.1-a-DB25-FINAL.zip
seems corrupt. Unable to expand it using either Windows or Mac, and uploading to JLCPCB it doesn't recognize it either.

The IBM AS/400 relies on non-standard OpCodes SKIP READ and SKIP WRITE, which are partially documented in SCSI Command Reference - 2105 Models E10, E20, F10, and F20.
This is a feature request to add support for these OpCodes.

I tried to order some of these with jlcpcb but they sent back the following issue:

"As shown below, the wall thickness like those indicated area are too thin which has a high risk of crack, loss, damage and deformation. Will this risk be acceptable for you?

If not, could you please kindly increase it to be at least 1.0mm to proceed? 2.0mm will be better."

@erichelgeson

HI,

I use:

• H.W. Ver. 1.0-c
• Firmware v1.0-20210410 and also v1.1-20211215
• SD Kingstone 16 GB
• File Name (HD00_2048-T14.hda) and size 2 GB and also i try 1 GB , 512 M.B.

Result in PC:

It work fine in my PC (windows XP- and Adaptec AHA-2940UW SCSI controller) and i can read and write on it

Result in Special O.S. and special SCSI Controller:

I try it in special O.S. and special SCSI Controller (Factory Machine)
and the O.S. can see the disk and detect it's size 2 GB but can't Format It

Needs:

I will try the new Firmware v1.1-20220404 but i need Support for the following :

• Create File More than 2 GB EX: (4 GB) and let the special O.S. take what he need from the file to create his Partitions
• How create more details log / debug to can see what is the exact happen when the special O.S. try to format it ,cause the normal log stopped in Finished initialization of SCSI Devices - Entering main loop.

> Choose an option:
> 
> [1] Upgrade via USB.
>  - You must have BlueSCSI v1.1-20220404-USB or greater already flashed.
> [2] Prepare a blank BluePill or older BlueSCSI firmware for USB support.
>  - You must have a STLink v2.
> 
> [H] Help
> [P] Page Sysop
> [Q] Hang Up
> 
> Choice?2
> Flashing the generic_boot20_pc13.bin
> * Move BOOT0 to 1 (Jumper furthest from the RESET button on the BluePill)
> * Connect the ST-LinkV2 to your BlueSCSI and connect it to your computer.
> Press any key to continue . . .
> Starting flash...
> NOTE: Some BulePills can take up to 2 minutes to flash.
> xPack OpenOCD x86_64 Open On-Chip Debugger 0.11.0+dev (2022-03-25-17:32)
> Licensed under GNU GPL v2
> For bug reports, read
>         http://openocd.org/doc/doxygen/bugs.html
> 0
> Info : auto-selecting first available session transport "hla_swd". To override use 'transport select <transport>'.
> Info : The selected transport took over low-level target control. The results might differ compared to plain JTAG/SWD
> Info : DEPRECATED target event trace-config; use TPIU events {pre,post}-{enable,disable}
> Warn : Transport "hla_swd" was already selected
> hla_swd
> Info : clock speed 1000 kHz
> Info : STLINK V2J37S7 (API v2) VID:PID 0483:3748
> Info : Target voltage: 3.110657
> Info : [stm32f1x.cpu] Cortex-M3 r2p0 processor detected
> Info : [stm32f1x.cpu] target has 6 breakpoints, 4 watchpoints
> Info : starting gdb server for stm32f1x.cpu on 3333
> Info : Listening on port 3333 for gdb connections
> target halted due to debug-request, current mode: Thread
> xPSR: 0x01000000 pc: 0x080000f0 msp: 0x20002800
> ** Programming Started **
> Info : device id = 0x10010414
> Info : flash size = 256kbytes
> ** Programming Finished **
> ** Verify Started **
> ** Verified OK **
> ** Resetting Target **
> shutdown command invoked
> Check output if successful (Windows...)
> * Move the BOOT0 Jumper back to position 0
> * Disconnect the BlueSCSI from the STLink
> Ready to flash BlueSCSI firmware.
> Press any key to continue . . .
> Flashing BlueSCSI-v1.1-20221203-STM32F1-USB.bin
> * Move BOOT1 to 1 (Jumper nearest the RESET button on the BluePill)
> * Connect via USB to the BluePill
> dfu-util 0.11
> 
> Copyright 2005-2009 Weston Schmidt, Harald Welte and OpenMoko Inc.
> Copyright 2010-2021 Tormod Volden and Stefan Schmidt
> This program is Free Software and has ABSOLUTELY NO WARRANTY
> Please report bugs to http://sourceforge.net/p/dfu-util/tickets/
> 
> Warning: Invalid DFU suffix signature
> A valid DFU suffix will be required in a future dfu-util release
> Waiting for device, exit with ctrl-C

Bootloader loads into Bluescsi but FW does not. I've tried everything and I don't know what causes this problem.

In STM32 ST-LINK Utility v4.6.0 it verifies the bootloader file and everything is correct.

What am I doing wrong?

Not all formats are supported. The DSK I made with Mini vMac is not recognized properly

At least for users of an Atari MegaSTE it would be great if BlueSCSI was supporting other LUNs than only LUN 0. The MegaSTE has a restricted internal SCSI port, which only supports a single SCSI ID/device, i.e. it cannot access more than one BlueSCSI image file. With support for several LUNs MegaSTE users could not just use BlueSCSI as a replacement for exactly one hard disk drive, but also for more hard disk drives or a CD-ROM drive.

Hey there!
I'm trying to get a bluescsi running in an old AT machine.
The installation of DOS5 works fine, partitioning is no problem and setting the primary partition active also works without a hitch.

Unfortunatly I don't seem to be able to boot from the bluescsi. Is that normal? Or is there a trick to it =)?

Cheers!

LUN greater than 0 causes the file to be ignored as though it wasn't there
HD11 - test.hda - this is ignored, led flashes 3 times
HD10 - test.hda - this works fine

VER: 1.1-20221203-USB

Am I misunderstanding the naming convention or is there a bug?

I tried to order some of these on jlcpcb, but they sent back the following issue.

"As shown below, the ratio of length, width and height is more than 1:10, and there's a large empty area inside the model, there will be a high risk of deformation.

Will this risk be acceptable for you?"

Checking the schematic there seems to be free pins on the bluepill that one could add a button to cycle trough CD images.
Could you please add the possibility to do that?

Proposed naming scheme:
DC[id][lun]_2048_[increment]_whatever.img

Hi. Please tell me how to create a disk image larger than 4 GB on a Windows computer? I don't have macOS or Linux.
I found the Disk Jockey program, but it creates disk images for a volume of only up to 4 GB.
Link to the program:
(https://diskjockey.onegeekarmy.eu/)
I will be glad if someone describes a way to create disk images in the format ".hda" or will share a ready-made image for 8 GB or 16 GB.
Thank you in advance.

Hi,

It seems that the "from scratch" option for this project is the PCB that was designed for it. Is there a schematic to solder this on a plain protoboard?

Hello and greetings from Poland!

I recently got a BlueSCSI from a buddy who happened to have one lying around. I wanted to use it my NeXTstation Turbo Color, to replace SCSI2SD. Unfortunately, I can't seem to get it to work.

I lifted the image off of SD card I use in SCSI2SD. It's verified correct by running it in the emulator. The image is 4GB, 512 byte sector. I dropped it on an exFAT SD card under name HD00_512_NS33.img. The adapter detects the image just fine and assigns it to SCSI ID 0 (see LOG.txt below).

The machine doesn't want to talk to it at all. I get all sorts of craziness - from SCSI hangs (termination is on), through checksum errors, blk errors, all the way to kernel panics and read errors when booting. There's no pattern, sometimes the thing will detect the disk and try booting, sometimes it'll error out on SCSI detection, other times it'll just hang in the ROM Monitor.

I connected the adapter to a PC using Adaptec AVA-2904 SCSI controller and it just worked. The image got detected. I also played around with other images just to check if data transfers actually work and stuff. All's fine and dandy under both Linux and Windows XP.

Problematic machine: NeXTstation Turbo Color
Problematic host adapter: NCR 53C90A

What I tried:

• Using different SD cards and both FAT32 and exFAT.
• Assigning the image to different SCSI ID (0, 1, 2, 3).
• Downgrading the firmware - tried 20221203, 20220917 and 20220626.

Just wanted to let you know. If you happen to have any ideas what to do, I'll be happy to give them a go.

Anyways, thanks for a cool project. Even if it doesn't work in my NeXTstation, I'll have a reason to grab another SCSI-based machine to maybe try with it. I'm thinking an SGI O2 ;).

Have a good day (or night)! :)

EDIT: Forgot the LOG.txt...

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VER: 1.1-20221203-USB
DEBUG:0 SDFAT_FILE_TYPE:3
SdFat version: 2.2.0
Sd Format: exFAT
SPI speed: 50Mhz
SdFat Max FileName Length: 64
Initialized SD Card - let's go!
Sd MID:2 OID:TM
Sd Name:SA08G
Sd Date:5/2015
Sd Serial:622993614
Not an image: _HD00_512_EMPTY.img
Not an image: .__HD00_512_EMPTY.img
Not an image: ._HD00_512_NS33.img
 - HD00_512_NS33.img HDD / 3984588800bytes / 3891200KiB / 3800MiB
ID:LUN0:
 0: 512:
 1:----:
 2:----:
 3:----:
 4:----:
 5:----:
 6:----:
Finished initialization of SCSI Devices - Entering main loop.

This is a feature request to add support for Vital Product Data tables for the INQUIRY OpCode.
The Linux tool sg_vpd can be used to validate correct function.