SCSI2SD Project - anyone interested ?

[Bunsen]

CF cards aren't very common any more

They're still being mass-produced and are widely available. I don't see them disappearing any time soon.

and mounting one on a modern PC for file copying may require extra hardware.

Also the case with SD if you don't have a built-in slot.

Is there a technical reason why CF cards are easier to adapt for SCSI than SD cards?

They're way faster in practice if not in theory? They're both parallel 8/16 bit interfaces? They're not tied up in NDAs and patents?

The SPI protocol will certainly be the limiting factor here.

But yeah... pin count. :/

how many PSoC pins are tied up just in SCSI handling

36 pins.

Leaving 36 unused? Okay then.

CF pinout and technical data.

I count 21 pins marked as "Essential for minimal 8-bit interface", and a further eight marked as "Essential for 16-bit interface", in True IDE Mode.

Total: 29 pins

Sounds doable, not that I have any idea what I'm on about. Opinions from the clueful welcome.

The use of mux/demux ICs on the SCSI and CF side to serialize parallel data might help, at a slight increase in cost and board complexity.

 

[mmcmaster]

It looks like smaller capacity cards really are faster, at least for these purposes.

I just tested a Kingston 2Gb card (unknown class), and it's about 50% faster than the class 10 16Gb I started testing with. Lido7 reports 600kb/sec write.

 

[mmcmaster]

The first batch of devices has been shipped :b&w: I have one assembled device left.

I've uploaded utilities for updating the firmware over USB, and for setting configuration parameters over USB here. The configuration tool includes the magical "--apple" option, which sets the parameters to fool Apple Drive Setup.

For those who ordered devices, you will need to update the firmware before using the config tool. Run "bootloaderhost /path/to/SCSI2SD.cyacd", wait for the prompt, and then connect the USB cable. Run as root on Linux. Seems to work fine on Windows without running an Administrative Console. I don't have a Mac OSX machine to compile or test on. Al of thel required files are available via the download link above. I'll start working on a basic user manual tomorrow to explain how this works.

 

[cb88]

Why not do an SD card raid array.. since you have so many extra pins?

 

[Trash80toHP_Mini]

OOOOOOH!!!!!!!!!!! Striping across a few cards ought to increase throughput tremendously. :approve:

 

[olePigeon]

That was my plan with my 2 Slot Acard if I can get it working. :-/

 

[Bunsen]

That's not going to happen, IMO. Dual slot devices on the Mac SCSI chain appear as a single volume. Mac SCSI doesn't do LUNs; if you want RAID, you need two separate devices.

This is not necessarily a barrier to using two or more SDs on a device for which one is developing new firmware. Either the device stripes the SDs itself, and presents a single volume to the Mac (faster, simpler) or it could present two slots as separate SCSI IDs and let the Mac do software RAID (slower and more complex).

 

[Trash80toHP_Mini]

I was thinking along the lines of the former, but using the extra I/O lines and consecutive reads of multiple cards to fill the que to the point that the converter is waiting for data to funnel through the Mac's SCSI bottleneck vs. vice-versa.

Guessing a second proc a/o some cache might be necessary, but it's hard for me to describe, imanoob. :-/

 

[cb88]

Indeed that is what I meant as well.. "hardware raid" the mac should still only see one drive.

 

[Trash80toHP_Mini]

FastSCSI2SCSI2SDFew? ;D

 

[Paralel]

Hard Disk Speedtool from Intech 3.6 support SCSI RAID on Mac's, but I think it has to be a PowerMac. Not sure. You would have to check their site.

 

[bbraun]

Mine arrived last week and it's been great. I've been using it on a Quadra 800 and a Mac II. I know nothing is ever fast enough, but I haven't had any complaints with the speed. I installed 7.1 in a couple minutes, no problems. This has been pretty neat, thanks a bunch mmcmaster!

I got the PCBs I ordered and the parts to build one should be coming from digikey in the next couple days. Hopefully I'll be able to assemble one. The fine pitch of the processor is a bit intimidating, but I figure I'll give it shot.

 

[techknight]

I might route out some 2.5" versions of these. What do you need to program these processors? Can i build a programmer?

On another note. I think the bases have been covered. Diy scsi. Diy floppy. Diy keyboard/mouse interface. Only thing left now is diy networking. Like localtalk to ethernet. Or a pds ethernet

 

[paws]

DIY Mac :lol:

 

[mmcmaster]

What do you need to program these processors? Can i build a programmer?

Any JTAG programmer compatible with the ARM 10-pin 0.05" connector should work.

 

[Trash80toHP_Mini]

. . . Or a pds ethernet

You've already started working on an'030 PDS ethernet solution.

The Duo modem/power card is the perfect form factor to shoot for and application for prototyping. You can patch wire the necessary signals from a Floppy MicroDock docking connector's solder cones to any convenient interboard connector lashup. Just bolt the connector assembly up to the bottom plastics of your mangled MicroDock and you can hook conveniently sized protoboard up to the connector for testing both direct and bridged '030, '040 and PPC PDS configurations using 230/270c, a 280 or a 2300c Duos. That will cover just about every NuBus architecture Mac and PowerBook. After that's up and running, it will be time to condense your functional ethernet electronics to a size/shape to cover all other PB modem form factors.

PDS slot cards for most desktop will be a piece of cake after that.

Finally, development for Killy Clipped Compact Macs and Luggable PDS 68000 versions can commence . . .

. . . or something to that effect. }

 

[techknight]

yeah I totally forgot about that thread. and project for that matter

 

[bigmessowires]

DIY Mac :lol:

http://www.bigmessowires.com/plus-too/

 

[Trash80toHP_Mini]

. . . add ROMs and stir. [] ]'>

 

[bbraun]

Well, I assembled a board by hand (no baking, just soldering iron) and aside from possibly inhaling some of those super small capacitors, it looks fine, no observable shorts, and continuity tests seem fine.

I'm trying to use a Segger jlink2 with a 20pin to 10pin adapter to program the USB_bootloader. I've got a jlink2 compatible gdbserver that sees the cypress processor and I can load the USB_Bootloader.elf, but doing so doesn't show up as a USB device on the computer, bootloaderhost and scsi2sd-config don't seem to recognize it.

I'm not sure if 'load' from gdbserver is actually programming the device (it appears to be...), or what exactly is going on. It could be my flashing process or there could be a problem with how I assembled the board, I'm not sure yet.