• Please review our updated Terms and Rules here

ISA 8-bt SCSI project interest?

Chuck(G)

Chuck(G)

25k Member
Joined
Jan 11, 2007
Messages
44,626
Location
Pacific Northwest, USA
This is just something that I'm tossing out for comments.

After going through the forum listings and looking at the trouble that people have trying to get a bootable SCSI disk going, I wonder if it might be of interest to anyone to start an 8-bit SCSI controller project?

It could be done fairly economically by using a modern microcontroller instead of one of the legacy SCSI interface chips. ASPI support for other devices might even be a possibility.

The design would probably be not more than 5 or 6 ICs, tops.

Thoughts? Or is SCSI too much of a dead horse?
 
I am definitely interested in SCSI interfacing - SCSI has been reliable for me across many different machines. However, doing the equivalent of a new SCSI chipset from scratch seems daunting. Are you thinking about patterning the microcontroller code after an existing chipset?

On a related topic, I've also seen your posts about the old NCR chipset on classiccmp - I'd like to get the application notes.
 
8 bit SCSI would be more useful to me then the IDE card (which I did buy), but would probably be more work to design.
 
The electrical part of SCSI is pretty simple; most of the work is in the programming. Electrically, I don't think SCSI is any more complicated than IDE.
 
What kinds of drives would one hope to support with this card? I take it LDV drives with backplanes would be taking it too far? :D :p

I am a bit in the dark with older SCSI tech, so to me it does not make too much sense, what benefit would the SCSI have over the XTIDE project? Especially looking at drive sizes (storage size...) SCSI and IDE both have comparable storage?

Enlighten my ignorant side please ;)
 
Given that there are already a few (fairly) commonly available 8-bit SCSI controllers out there, I don't know if it's really worth the effort to completely reinvent the wheel. The thing with the XTIDE project, is that it brings functionality that didn't previously exist, or was extremely rare: AT IDE on an XT bus. The SCSI guys, me included, already have some semi-workable options... the IDE guys didn't.

It would, however, be nice to have a working BIOS to use with the most common SCSI cards, to fix some of the niggling little bugs. For example, on my T130B, I'd really like not to have to swap the CPU for a V30 to get it to work reliably (especially on my particular machine, since the EMS card won't work with a V30), and also to get support for larger drives. Having a good BIOS floating around would also allow the use of the cards that didn't come with a BIOS installed (the majority of 'em?)... just burn the new one to chip and plug it in. Just a thought...
 
I was thinking single-ended SCSI-I type of interfacing. If you want to do something else, you can buy a converter. :)

There are SCSI peripherals (e.g. tape drives) that simply have no comparable IDE version. There is also a standard API (ASPI) for drivers. It also was the case in the old days (don't know about now) that SCSI drives were generally built better (and cost more) than MFM/IDE drives.

My other thought on this was if there was the SPI lines from the uC uncommitted, an SD card interface could be added to take care of your hard disk needs. It seems such a waste to put just an SD card interface on an ISA card--so one might as well add something useful like SCSI.
 
Old Thrashbarg said:
Given that there are already a few (fairly) commonly available 8-bit SCSI controllers out there, I don't know if it's really worth the effort to completely reinvent the wheel. The thing with the XTIDE project, is that it brings functionality that didn't previously exist, or was extremely rare: AT IDE on an XT bus. The SCSI guys, me included, already have some semi-workable options... the IDE guys didn't.
Click to expand...

In the range of bootable 8 bit controllers that I've seen, there's the Seagate ST01 (slow and very quirky), Trantor T128 and T130 (also quirky) and the Future Domain TMC-850 (not the most available one). The DTC 3250 and a hanful of others could boo t, but I haven't seen one offered for a very long time.
 
The 8 bit SCSI are hard to find and expensive if you do find one. I was lucky enough to get a TMC-850 without a BIOS chip, soldered in a BIOS and support chip to get a bootable card. Normally they are $50+ which is more expensive then the XT IDE card.
 
My T130B was $15, and I bought it a couple weeks after I passed on one for $25, both found on eBay within the last year, and I've run across a couple since then, too.

Yeah, they're not super common, the TMC-850 being one of the less common models, but there's also not a whole lot of people looking for them. Probably even fewer now that the XTIDE project is going.
 
Well, the thought was to make an SD-card interface adapter for the 8 bit ISA bus, but that leaves a bunch of extra processing power doing nothing. USB host interface (UHCI) is a nightmare if you expect it to be absolutely general (i.e. be able to plug hubs, keyboards, scanners, hard disks, etc. into it). While you might succeed in making a host just to support a USB flash drive, I'd much rather have the SD card.

We've already got IDE, so that leaves SCSI.

Just looking for some ideas...
 
Yeah, the SD card interface would be kinda cool, although I'm not really sure what practical advantage it would give over a compact flash card running on IDE. Is there a particular reasoning behind it, which I'm overlooking?
 
Floppy, as in Teac FD-235HS? Those are pretty uncommon (I have one). I don't see why not, given the right drivers (ay, there's the rub!) I'll probably go as far as an ASPI driver and boot ROM for hard drives and leave everything else to the user's imagination ;).

Did anyone ever make CD-ROM burning software for a PC XT?

(P.S. If you'd like to see something ridiculous, check out the price on this FD235HS drive. Supposedly, they're still available from Teac.)
 
Hi! Well, if you want SD on ISA bus that would be essentially an ISA version of the N8VEM PropIO board. An ISA board with a Propeller uC on it and a uSD socket. It *appears* to be a pretty simple conversion although it would require programming on the PC *and* Propeller sides.

Anyway, if the 8 bit SCSI ISA board ever gets traction I'm game for a PCB. Right now is not a good time but eventually when things calm down a bit.

Thanks and have a nice day!

Andrew Lynch
 
I have no use for a SCSI controller (unless it supported the full suite of external SCSI stuff), but the SD card interface... Might I make the comment that it would be best if the SD card was mounted with a slot through the rear of the bracket, so that it can be externally changed out like a card reader. Basically I'd use it like a card reader with reboots between swaps.
 
Hi! Keeping consistent with Chuck's idea, the Propeller uC has 32 possible GPIO pins. The uSD interface requires 4 IIRC and SCSI-1 requires 18, I think. The rest of the pins on the connector are grounds and power. The Propeller is a 3.3V device which is not a problem but the GPIO pins would require level conversion and/or pass through a bus transceiver/buffer to the SCSI interface.

There two main kinds of SCSI interface connectors I think would be most useful; the 50 internal dual row header for ribbon cables and the 25 pin external DB-25. Assuming the Propeller needs 2 GPIO pins for a EEPROM there are still 30 GPIO pins remaining with uSD needing 4 and the SCSI needing 18.

Where things get tough is the bus interface for the Propeller required several (13?) pins itself such as D0-8, /CS, /RD, /WAIT, A0 & A1. I think a Propeller ISA to SCSI solution is at least possible although wedging in a uSD interface may require either a second Propeller, switching logic, or using a different uC with more GPIO pins.

I am sure there are ARM, PIC, or other uC solutions that would work. Probably something in PLCC or DIP that keeps relatively easy hobbyist buildability without resorting to SMT packages.

Just my $0.02. Thanks and have a nice day!

Andrew Lynch
 
All of my external SCSI devices use Centronix 50, I don't have a single DB25 device myself (other than SCSI ZIP drives). Perhaps a poll is in order about this one - or offer an adapter or both connector options if it's a simple wiring matter?
 
You must log in or register to reply here.
Back
Top