11 Jun
2016
11 Jun
'16
9 p.m.
My Nova 3/12 system has a 6045 cartridge hard drive, with one fixed platter, one removable
platter, and a capacity of 10M. I haven't managed to boot my computer from it yet, and
after a long pause, my Nova 3 is requesting another time slice of my attention.
I can see in the drive's technical manual where I can specify surface 0-3 in the
commands. I assume that the built-in bootstrap loader in the Nova 3 reads drive 0,
cylinder 0, surface 0, sector 0, and places it at memory location 0, simply by virtue of
all of the registers having been reset after an initial power-up and spin-up. The
bootstrap loader code I've seen looks like it just issues a read command without
initializing the memory address register and disc address/sector count register. I
don't fully grok that code yet, so maybe I'm mistaken.
I have not found mention yet of which surface numbers correspond to the fixed platter and
which correspond to the removable one. Is the removable platter selected as surfaces 0 and
1, such that the system would normally boot from the removable platter? Or would it
normally boot from the fixed platter, with the removable platter being used to get data on
and off the system?
Incidentally, I wonder if anybody has any original printed technical documentation
relevant to my Nova 3 system which they might like to sell. I'm working off PDFs right
now, and I'd like clean original paper copies for easier perusal and reference.
I'm also interested in replacing the filler panels in my Nova's rack with other
interesting peripherals. Maybe a floppy drive? My 6045 is supposed to be able to be
mixable with 6030 drives on the same bus, according to the 6045 manual. If anybody has
some excess Nova hardware that might fit into my system, please let me know.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11 Jun
11 Jun
10 p.m.
DG Model 6045 disk drive heads 0 and 1 are the removable platter, heads
2 and 3 are the 'fixed' platter. As you said, the controller memory
address and sector/sector/count values are set to 0 by the IORESET
pulse, so a read operation (STRT pulse) will start the reading of the
bootstrap into low memory from the removable disk.
Contact me off-list regarding the documentation...
Bruce
On 6/11/2016 10:00 PM, Mark J. Blair wrote:
My Nova 3/12 system has a 6045 cartridge hard drive,
with one fixed platter, one removable platter, and a capacity of 10M. I haven't
managed to boot my computer from it yet, and after a long pause, my Nova 3 is requesting
another time slice of my attention.
I can see in the drive's technical manual where I can specify surface 0-3 in the
commands. I assume that the built-in bootstrap loader in the Nova 3 reads drive 0,
cylinder 0, surface 0, sector 0, and places it at memory location 0, simply by virtue of
all of the registers having been reset after an initial power-up and spin-up. The
bootstrap loader code I've seen looks like it just issues a read command without
initializing the memory address register and disc address/sector count register. I
don't fully grok that code yet, so maybe I'm mistaken.
I have not found mention yet of which surface numbers correspond to the fixed platter and
which correspond to the removable one. Is the removable platter selected as surfaces 0 and
1, such that the system would normally boot from the removable platter? Or would it
normally boot from the fixed platter, with the removable platter being used to get data on
and off the system?
Incidentally, I wonder if anybody has any original printed technical documentation
relevant to my Nova 3 system which they might like to sell. I'm working off PDFs right
now, and I'd like clean original paper copies for easier perusal and reference.
I'm also interested in replacing the filler panels in my Nova's rack with other
interesting peripherals. Maybe a floppy drive? My 6045 is supposed to be able to be
mixable with 6030 drives on the same bus, according to the 6045 manual. If anybody has
some excess Nova hardware that might fit into my system, please let me know.
10:10 p.m.
On Jun 11, 2016, at 22:00, Bruce Ray <Bruce at
Wild-Hare.com> wrote:
DG Model 6045 disk drive heads 0 and 1 are the removable platter, heads 2 and 3 are the
'fixed' platter. As you said, the controller memory address and
sector/sector/count values are set to 0 by the IORESET pulse, so a read operation (STRT
pulse) will start the reading of the bootstrap into low memory from the removable disk.
Excellent! That makes sense.
Ok, studying the documentation scans I have some more, I think I plugged the hard drive
interface cable into the wrong connector when I reassembled the system after moving it.
Hopefully I didn't damage anything! On the backplane, there's a terminator on the
bottom rear connector (P3 I/O bus, I think), and there's another card edge installed
behind the top rear connector (P4, disk drive interface, I think). I had the disk drive
cable on the inboard add-on connector instead of the top backplane connector (P4). I think
that was wrong!
But studying the documentation, I see that P4 is hard-wired to slot 10. However, the cards
might be installed incorrectly in my system. I have plugged in:
Slot | Part Number | Description
-----+---------------+--------------------------------
12 | 107-000116-08 | DGC NOVA QUAD MULTIPLEXER
11 | 107-000187-16 | DISK CARTRIDGE CONTROL DGC NOVA
10 | |
9 | |
8 | |
7 | |
6 | |
5 | |
4 | 107-000151-19 | DGC NOVA CASSETTE I/O
3 | | 128K MOSTEK MEMORY
2 | 107-000621-01 | NOVA 3 TRIPLE OPTION
1 | 107-000539-05 | NOVA 3 CPU
I think that I need to both move the hard drive cable to P4 at the top rear of the
backplane, and move the disk controller card to slot 10. Does that sound correct?
I don't know what that expansion card edge is for yet, and I can't get to it
easily right now with the chassis in the rack. Maybe it's wired to the Cassette I/O
card? I should take some detailed pictures to post and discuss tomorrow.
That Cassette I/O card looks interesting to me, and I wonder if there's any chance
that I might find the tape transport that goes with it.
Contact me off-list regarding the documentation...
Will do! And thanks for your indispensable help with my system last time I gave it some
brain bandwidth.
Now I'm off to try and figure out what baud rate my console is strapped for, which I
wasn't smart enough to write down last time I worked with the system.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:18 p.m.
More details: I can't be sure because the Nova rack is presently in a crowded corner
which is keeping me from removing the chassis from the rack in order to get its top panel
off. But it appears to me that the mystery edge connector that I previously had the hard
drive interface cable connected to is wire-wrapped over to slot 11, where the disk
controller is plugged in.
Here are some pictures of the backplane:
http://www.nf6x.net/2016/06/data-general-nova-3-backplane/
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:32 p.m.
Aha! Now I remember. That Cassette I/O card is mostly unpopulated, and I think it's
just installed for its serial port. I think Bruce had told me about that long ago, but my
memories are as dusty as the Nova is at the moment.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
12 Jun
12 Jun
12:44 a.m.
-----Oorspronkelijk bericht-----
From: Mark J. Blair
Sent: Sunday, June 12, 2016 8:32 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Data General Nova 3 and 6045 Drive Questions
Aha! Now I remember. That Cassette I/O card is mostly unpopulated, and I think it's
just installed for its serial port. I think
Bruce had told me about that long ago, but my memories are as dusty as the Nova is at the
moment.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
---------
Just like Mark, it's about time to spend some time on my NOVA3.
I will be following this with interest!
I have very little knowledge of it, Bruce is a great guy. Thanks again Bruce!
Even identifying the hardware was only possible with Bruce's help ...
My system has a 6070 disk drive, also a non-removable disk combined
with a removable cartridge. And when I got the system it also had a
model 6031 single 8" floppy drive. Later I was very lucky to pick up a
6025 tape drive (with tension arms, not vacuum columns). I even got the
controller board. I am still searching for the interconnection cable that
goes from the backplane to the circuit board attached to the tape drive.
Were all bulbs on the front panel of your NOVA intact Mark?
Or did/do you also have dead lamps? Know of a source where to buy them?
- Henk, PA8PDP
1:28 a.m.
On Jun 12, 2016, at 00:44, Henk Gooijen
<henk.gooijen at hotmail.com> wrote:
Just like Mark, it's about time to spend some time on my NOVA3.
I will be following this with interest!
Yay!
Were all bulbs on the front panel of your NOVA intact
Mark?
Or did/do you also have dead lamps? Know of a source where to buy them?
I did have some dead bulbs. My memory is foggy, but I found some archived emails from when
I discussed the bulb replacement two years ago. I think that I used number 2185 bulbs that
I ordered from Digi-Key. Here are the old messages in my archive:
On Jun 14, 2014, at 09:21, Mark J. Blair <nf6x at nf6x.net> wrote:
Do any of y'all know the bulb number for the
fragile little wire-leaded bulbs used in the Data General Nova's switch console? I
have at least one burned out, and a couple have snapped off. I measure about 15V across
the bulbs on my machine, but I don't know yet whether that's the correct voltage,
nor do I know the nominal voltage, current or brightness ratings of the correct bulbs.
Picture:
https://twitter.com/nf6x/status/477846674543362048/photo/1
On Jun 15, 2014, at 09:06, Mark J. Blair <nf6x at nf6x.net> wrote:
On Jun 15, 2014, at 08:49 , Toby Thain <toby at telegraphics.com.au> wrote:
On Jun 15, 2014, at 14:16, Mark J. Blair <nf6x at nf6x.net> wrote:
Via email, snarfusmaximus writes, "I
replaced mine with yellow LEDs with a series resistor."
That's a pretty tempting solution. Those original bulbs are quite fragile, and appear
to be quite prone to suffering lead breakage near the glass seal.
Here's what I just posted on VCF, for those (few?) folks who are here but not there:
-------- 8< cut here 8< --------
Regarding the console bulbs, I found that one of mine that snapped off at the base still
has an intact filament. I was able to probe the wires at the glass seals to light it up
and measure current. The bulbs are normally powered from an unregulated 14V rail as I
understand things, and I measured the voltage at around 15V on my machine. The broken-off
bulb draws 39mA at 14V and the filament color is quite orange. It draws 50mA at 28V and
has a normal-looking color temperature. I conclude that the original bulbs were probably
designed for 28V operation or so, and are used at lower voltage in the Nova 3 to increase
service life (?).
McMaster-Carr and Digi-Key both have some bulbs that look like possible substitutes. I
think I'll try some 2185 bulbs from Digi-Key since I'm ordering other stuff from
them anyway, unless anybody has a better confirmed cross-ref. Those are 28V 40mA bulbs.
I also received an email from another collector who used KH 4-280-040A-1 bulbs from KH
Lamp in his Nova. He reports that he got them from Swedish distributor Elfa as part number
33-657-98. That's a 28V 40mA T1-1/4 bulb.
On Jun 15, 2014, at 12:51 , Brent Hilpert <hilpert at cs.ubc.ca> wrote:
Here's another testimonial for a Nova, this
fellow mentions replacing them with 28V/0.04A bulbs, but doesn't say how he knows
those specs:
http://www.foxdata.com/blog/?tag=nova-312
He's emailed me about his project. I can't access his web site at the moment, and
we're planning to look into that tomorrow.
The Hudson 2187D type he mentions maps to Chicago
CM-2187, listed at Mouser:
http://ca.mouser.com/ProductDetail/Chicago-Miniature/2187/?qs=qp111mKzDjgel…
Cool. If the 2185 bulbs that I'm ordering from Digi-Key don't work well, I'll
try some 2187 bulbs.
12:37 p.m.
Disk testing update:
I have the controller in slot 11 still, and the drive interface cable is connected to the
edge paddle that appears to be wire-wrapped to slot 11. I haven't tried it in slot 10
using edge connector P4 yet, because I think I might need to fiddle with the two blue
wires (bus grant? interrupt?) I see patched between the slots to do that, and I
haven't studied what they do yet. Pictures of the backplane can be found on my blog:
http://www.nf6x.net/2016/06/data-general-nova-3-backplane/
Bootstrap from the drive channel still doesn't seem to work. The bootstrap loader does
appear to be loaded into RAM, and the loader remains looping at 000377, which never gets
overwritten with a word from the drive channel.
Reading the drive status register always returns 0 so far. But, I do seem to be talking to
the drive, because I think I can make it seek!
I cobbled together the following program and assembled it with Toby Thain's assembler
found here:
http://www.telegraphics.com.au/sw/info/dpa.html
02 .TITL dsktst1
03 .LOC 0
05 00000 062477 START: IORST ; Reset I/O channels
06 00001 060633 DIAC 0, DKP ; Drive status register to AC0
07 00002 065433 DIB 1, DKP ; Drive mem addr register to AC1
08 00003 072433 DIC 2, DKP ; Drive disk addr register to AC2
09 00004 034010 LDA 3, SEEK ; Load seek command to AC3
10 00005 075133 DOAS 3, DKP ; Do seek
11 00006 074433 POLL: DIA 3, DKP ; Poll drive status to AC3...
12 00007 000006 JMP POLL ; ...forever
14 00010 003220 SEEK: 003220 ; Seek cyl 620(8):
15 ; 0 0000 1 10 10010000
16 ; -> 0 000 011 010 010 000
17 .END START
This resets the I/O channel, reads the three drive registers into AC0-AC2, issues a near
full seek command, then repeatedly polls the drive status register into AC0. I think.
Running the program after a fresh drive load->ready sequence causes the drive to emit
an audible clunk, so I believe that it is performing the seek! However, stopping the
program and examining the accumulators shows that they all contain 0.
Incidentally, I see that the assembler I used assembles IORST as 062477, thus clearing the
start and done flags. However, other sample programs I have seen show IORST as 062677,
which issues a pulse instead of clearing start/done. I did not observe any difference in
behavior in my test program when I tried both IORST varieties. Can any Nova experts chime
in on any hidden gotchas that might lurk around this?
So, back to the task at hand: It seems that the CPU can talk to the drive, but maybe the
drive can't talk back to the CPU? I think I'll work on bringing up a memory test
next before I look at the hard drive some more.
BTW, the console is running at 4800 baud, while I expected it to be at 9600 baud based on
the strapping. That needs some more investigation, too.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
8:55 p.m.
G'day Mark -
Brief notes...
1) DG part numbers are 005-xxxxxx-yy; 107-xxxxxx-yy numbers are circuit
board artwork numbers. Unfortunately, there is no standard
cross-reference between the two. Briefly, the part number is the
primary reference to be used because a single 107-xxxxxx-yy circuit
board may actually contain multiple 005-xxxxxx-yy parts.
The 005-xxxxxx-yy part number usually exists on a label attached to the
15"x15" board stiffener (the side opposite the finger edge connectors).
Later DG boards also had an "A" number in addition to the product number
("T"), so if both exist the "T 005-xxxxxx-yy" number determines the
exact board function(s).
2) Slot 5 looks like it was also wired for a serial port board (4010 or
4075) at one time since backplane pins A85 and B69 are wired to
someplace in the backplane's paddleboard area. But at this point the
Interrupt priority and Data Channel priority daisy chains have been
jumpered over the slot position to bypass the empty slot. (pins A93/A94,
A95/A96)
3) One integral backplane edge connector may have a black "bus
terminator" plug to reduce noise on the bus. All DG terminators had a
005- number on 'em for identification.
4) 4075 TTY baudrate pretty easy to check; usually 4800 or 9600 baud, 7
data, even parity, 1 stop bit.
5) The 4060 board probably goes in slot 12 - wire-wrap wires go from
slot 12 backplane pins to standard DG 4083 MUX connector distribution
panel (the 16 row by 9 (+ 3) inline pins for each of the 4 serial ports
plus 12 unused ports).
6) 4046 disc controller board in slot 11 - wire-wrap wires seem to go
from slot 11 to edge connector.
Bruce
9:07 p.m.
On Jun 12, 2016, at 20:55, Bruce Ray <Bruce at
Wild-Hare.com> wrote:
G'day Mark -
Brief notes...
Thanks!
4) 4075 TTY baudrate pretty easy to check; usually
4800 or 9600 baud, 7 data, even parity, 1 stop bit.
I determined earlier today that
I had merely had a brain fart regarding the strapping. The card was strapped for 4800, and
I successfully reconfigured it for 9600.
6) 4046 disc controller board in slot 11 - wire-wrap
wires seem to go from slot 11 to edge connector.
Is there any specific reason you know of for the card to have been installed in slot 11
and requiring a laborious paddle board wiring job, rather than plugging it into slot 10
and using the P4 connector?
I am able to seek the drive heads, but can't seem to get any data or status back from
it yet. I'm working on that assembler side project to make it easier to write up test
programs to try out drive functions and debug things. I ran one of the memory checkerboard
tests for quite a while without any apparent failures.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
3114
days inactive
3115
days old
9 comments
3 participants
participants (3)
-
Bruce＠Wild-Hare.com -
henk.gooijen＠hotmail.com -
nf6x＠nf6x.net