Golly! I wasn't aware that they actually put a SCSI controller on the
board. That must have been several rev's later than mine. My two boards
were the beta and first release. I put a couple of hundred of these into
the field because I liked them and they allowed pretty compact packaging.
Take a look at some comments in-line below, please.
regards,
Dick
-----Original Message-----
From: Allison J Parent <allisonp_at_world.std.com>
To: Discussion re-collecting of classic computers
<classiccmp_at_u.washington.edu>
Date: Friday, April 16, 1999 6:35 PM
Subject: Re: AMPRO LITTLE BOARD, etc.
><I bought mine new as a kit in '83, as part of the first production run.
><They didn't have HDD support then, and since I made and sold an adapter
><daughterboard to interface to Z-80 processor sockets, one of several, I wa
><not concerned about that. The board always worked reliably and, aside fro
><the occasional need to read a standard diskette, the system was pretty
><complete.
>
>Mine is only slightly later, has the 5380 scsi chip.
>
><Well, the drive attached to my two boards, i.e. the hard drive, is an
ST-50
><that was lying about some fifteen years ago when I happened to need a
drive
>
>The sequence of disks over the years for the HDD side:
>
> xybec and st506 bought new in 82(late). Still have both.
> Adaptec, and Quantum D540 (31mb) much faster.
> the adaptec died and the xybec was in with a st251 for a while
> then the fujitsu 3.5" scsi drive.
>
I was propping the door to a room in the basement with a couple of 10MB
RODIME 3.5" drives which might work really well with this arrangement. I
once ran one of these with a 1"-high 3.5"Sony drive which a PC believed was
a 1.2MB drive. Maybe I can fool the FDC drivers into doing a similar thing.
That would be handy. There's no "standard" 3.5" driver for CP/M.
>
><>supported. The host interface was very similar in concept and nearly
><>in execution as IDE.
><
><
><The history included in the IDE spec clearly indicates that it was
patterne
><after the 1003-WAH board (of PC/AT fame), which is the PC-bound eqivalent
o
><the 1002. It uses the same IC's, hence the same command structure and bit
>
>Here we go again... I DIDN'T SAY IT WAS IDE. I said it was similar in
>respect that it was a bus level interface for a controller and predates
>IDE. I do know that the wd1003 was what the IDE base design was patterned
>after.
No, you surely didn't say that! It doesn't use the same cable definition or
any thing like it. It just happens to use a 40-conductor cable. The reason
I said I'd try modifying the firmware to talk to an IDE drive is that there
is reason to believe that the command structure is identical although the
hardware would have to be modified to load the registers as it's done on the
16-bit -WAH controller. The fact is that the 1000-05 and the 1002-05 bridge
controller boards use the same chip (WD1010-05) as the HDD interface.
That's where the command registers are located and they'll require
programming in the same way. Since the IDE drive has, essentially, the same
register set on it to accomplish the controller functions, and the register
set mapped in the same way as the WD1003-WAH, it should be possible to make
it work similarly.
><definitions. I'm inclined to try hooking an IDE drive up in place of the
><1000-05 board and ST-506 drive just to see what it does. I'd imagine they
><emulated it faithfully.
>
>>From the programming I did in the late 80s with them I'd say they did.
>
><The WD hardware required a fair amount of time to do its job, particularly
><the BMAC chip, which is really just an 8042. The access time for which th
><WD controller series was built was a bit slower than the access window fro
><a 4 MHz would consistently allow. Consequently, I put a device specific
><wait-state generator on my adapter daugherboard. It decoded the address
an
><generated a wait only for this one device, since the AMPRO guys, or whoeve
><else generates wait-states when they think they're needed could still do i
><on their own for other devices.
>
>Explains alot. the first hard disk system I had was S100 so the standard
>for performance was already set in my mind by '83.
>
><Yes. I have a bunch of the ADAPTEC 4070 bridge controllers left over from
><something. They record in RLL code, so the "little" drives I used to use
a
><doorstops, etc, (ST506, TM602, SA6-whatever) can be nearly a full CP/M
><"volume," i.e. they hold about 7.2 MB which is nearly the max for a CP/M
><drive size. It works out well.
>
>I have one of those that replaced the MFM one from way back.
>
><I wasn't aware that ZILOG or MOSTEK had 8MHz peripherals, though it doesn'
><surprise me. I ws siphoned off into a bunch of 8748/8751 stuff in
mid-1983
><and stopped following the Z-80 for a while. By the time I was able to com
>
>Finding them can be tough. the 6mhz are more common. 8530s or 8330s are
>easier to find and offer better perfomance.
>
I'm really not that hot for it, but think it would be charming to put a set
of '-H' parts in an early AMPRO Little Board and make it work at 8MHz. The
timing should work with 100ns DRAMs. Aside from the PROM, I doubt anything
other than the peripherals would be affected by the speed change. Of course
the FDC would require a different tap from the clock divider, but everything
else should work as is, save, perhaps the PROM. The PROM might work if the
clock switch were hacked as well. (I believe there was a little sorcery
with switching the clock speed after copying the PROM into RAM, by switching
the preset on a counter. What I liked about this was the really sensible
packaging you could use with these small boards. I have a video
display/keyboard on a similar form factor which I'd really like to package
with these other two boards and the drives. That would be truly minimalist
for the time.
>
>Allison
>
>
Received on Fri Apr 16 1999 - 20:25:42 BST