BlueSCSI to SD

[nogginthenog]

This popped up in my Twitter feed. Classic Mac SCSI but might work with an Amiga.
https://gumroad.com/l/bluescsi-1b

1024KB/sec seems a bit slow?

[Matt_H]

I think that level of performance is comparable to the v5 SCSI2SD. It's a bit cheaper, too, but compatibility seems to be low even among the Mac family. I'd pass, personally, but maybe there's someone here who could use it in a vintage Mac it if it doesn't work in an Amiga...

[erichelgeson]

I'm the creator of BlueSCSI (derived from ArdSCSino) - happy to answer any questions. I've never used a Amiga but besides the vendor ID set to something the Mac disk utility will format - there shouldn't be any issues in other SCSI devices of the same era.

Compatibility listed is just my personal machines or users reports, but should work in any 040 or slower machine except the Plus (which predates the scsi standard so does a few 'odd' things, of which I'm attempting to sort out) - It's really meant for those early scsi1 machines.

While it is slower the seek time is 1ms. Large copies will be slower, but it feels faster during "normal" usage. I have a SCSI2SD 5.1 and this is a bit faster. I also have a SCSI2SD v6 - but that is quite a bit faster (and quite a bit more expensive)

Another project I contribute to is RaSCSI - which has been around for a while but has another resurgence due to classic mac enthusiasts. Can do quite a bit more (CD, ethernet, etc.) I see there are a few older threads but might be worth looking into again as well.

It's a fun little kit to assemble and it's all open. If anyone does get this working on a Amega I'd be happy to list it on the site/github.

[dalek]

Cool project - will certainly be a game changer and especially handy for giving SCSI back to older systems on the cheap.

Will the gerbers be available from somewhere eventually? Never fun to pay for one shipped from the U.S. to Australia when I can order 10 directly from China for less in total. Oh - I think I found them?: https://github.com/erichelgeson/blue...e/master/kicad

[erichelgeson]

They are here - https://github.com/erichelgeson/bluepill_scsi - (i need to combine the repos at some point) - I only ship kits to US and CA since international shipping is so costly. I do have someone in the UK selling a few too.

[dalek]

Thanks, I've ordered some and will give it a crack on a few Amiga SCSI controllers

[Matt_H]

Originally Posted by dalek

Thanks, I've ordered some and will give it a crack on a few Amiga SCSI controllers

Please let us know how it goes. It would be good to have an additional modern SCSI option.

[dalek]

I built two, one with a clone bluepill one with a genuine bluepill (both pass greaseweazle's blinky test). Same behavior on both:

* A2091 v6.6 ROM: WD33C93A-PL SCSI Chip: identifies okay, seems to read okay but hangs when trying to write / partition the disk with hdtoolbox

* GVP SCSI on 030 Combo v4.15 ROM (and tried guru rom) : AM33C93A SCSI Chip: identifies okay, seems to read okay but hangs when trying to write / partition the disk with hdtoolbox (same symptoms as A2091)

* ICD AdSCSI 2000 Rom 4.21: L5380PC-2: seems to work okay sometimes - can actually write the RDB with hdtoolbox but I couldn't get autoboot working (more testing needed)

Tried different SD cards but that's not the issue, as the log gets written just fine on all of them.

Here's the last few lines of SCSI Snoop with A2091 6.6 roms. The write seems to hang and the led on bluescsi stays lit:

READ CAPACITY   25 00 00 00 00 00 00 00 00 00 
IOReq 00249c10 returned. Result 00.
Data in:00 10 00 00 00 00 02 00 
IOReq 002425a0 HD_SCSICMD(write,sense on,cmdlen=6,len=00000000)
TEST UNIT READY 00 00 00 00 00 00 
IOReq 002425a0 returned. Result 00.
IOReq 002425a0 HD_SCSICMD(read,sense on,cmdlen=6,len=00000024)
INQUIRY         12 00 00 00 24 00 
IOReq 002425a0 returned. Result 00.
Data in:00 00 01 01 1f 00 00 00 51 55 41 4e 54 55 4d 20 
46 49 52 45 42 41 4c 4c 31 20 20 20 20 20 20 20 
31 2e 30 20 
IOReq 002425a0 HD_SCSICMD(read,sense on,cmdlen=6,len=000000fe)
MODE SENSE(6)   1a 00 3f 00 fe 00 
IOReq 002425a0 returned. Result 00.
Data in:3b 00 00 08 00 10 00 00 00 00 02 00 03 16 00 00 
00 00 00 00 00 00 00 3f 00 00 00 00 00 00 00 00 
00 00 00 00 04 16 00 00 00 01 00 00 00 00 00 00 
00 00 00 00 00 00 00 00 00 00 00 00 
IOReq 002425a0 HD_SCSICMD(read,sense on,cmdlen=10,len=00000008)
READ CAPACITY   25 00 00 00 00 00 00 00 00 00 
IOReq 002425a0 returned. Result 00.
Data in:00 10 00 00 00 00 02 00 
IOReq 002425a0 CMD_WRITElen=00000200 data=002497a4 offset=00000000
Data out:52 44 53 4b 00 00 00 40 fd 16 c6 a6 00 00 00 07 
00 00 02 00 00 00 00 17 ff ff ff ff 00 00 00 01 
00 00 00 03 ff ff ff ff ff ff ff ff ff ff ff ff 
ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ...
00000200 bytes total

So looks promising maybe with some changes to timings.

[TheGouldFish]

I've been messing around with one today as well, but using the HD8+ on an Amiga 500+ and having similar issues.
FaaastPrep crashes when trying to work out the drive, and HDToolBox hangs when writing out the partition info.

[idrougge]

I have used the original ArdSCSIno (for STM32) on my A3000, and it certainly was possible to partition the drive and run DiskSpeed on it. I did at first have a pirate Bluepill on which it ran slower than a floppy drive, though, before I replaced it with a better one.

[TheGouldFish]

Did some poking around the code today and worked out a single line fix that worked for me.
adding a:
delayMicroseconds(35);
just above the:
LOG("Command:");

allowed me to partition, format, copy Workbench and boot off the bluescsi.
This was on a GVP HD8+.

I've been chatting in their discord so hopefully we can work up a real fix that doesn't need magic numbers.
but it might be worth trying to see if it fixes other Amiga SCSI controllers.

[dalek]

Hmmm doesn't fix it for my A2091 - still hangs in the same place. I tried a bunch of variations of delays (eg 25, 30, 40, 45, 50 etc) but no luck.

When turning on serial debugging, timing breaks even harder and it fails to read as well. (I was using the uart the via usb port, even tried 115200bps).

I'm guessing it's hanging in writeHandshake so will do some more playing around with it.

EDIT: yes, hangs in the loop:
while(!m_isBusReset && !SCSI_IN(vACK));

verified by setting the LED on just before and clearing it just after

[TheGouldFish]

So 35 was working for a 52mb drive (with buffers 30), but when I tried a 500mb drive with 60buffers I needed to increase the delay to 100.

I'll forward on your findings to the discord as well, so hopefully they can trackdown where the issue is.

[Henk]

Originally Posted by TheGouldFish

adding a:
delayMicroseconds(35);
just above the:
LOG("Command:");

allowed me to partition, format, copy Workbench and boot off the bluescsi.
This was on a GVP HD8+.

I am trying to use a BlueSCSI in my GVP HD8+

Can you tell me what file you are talking about? When I look at the SD cart there is one file LOG.txt.

CONTENT:

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VERSION: 1.0-20210410
DEBUG:0 SCSI_SELECT:0 SDFAT_FILE_TYPE:3
SdFat version: 2.0.6
SdFat Max FileName Length: 32
Initialized SD Card - lets go!
Not an image: LOG.txt
Imagefile: HD00_512.hda / 33553920bytes / 32767KiB / 31MiB
ID:LUN0:LUN1:
0: 512:----:
1:----:----:
2:----:----:
3:----:----:
4:----:----:
5:----:----:
6:----:----:
Finished initialization of SCSI Devices - Entering main loop.
ng main loop.

Thanks in advance!