-----Original Message-----
From: Tony Duell <ard(a)p850ug1.demon.co.uk>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Sunday, March 21, 1999 2:44 PM
Subject: Re: SWTPc 6800, common format (sharable?) repair journal?
>> First issue: Power system is missing the transformer. I do have the
MP-P,
>> the filter cap, and the rectifier bridge.
>
>MP-P?
(Power supply) it's the fused board that I'm guessing puts the power on the
buss.
>>
>> I found a nice ad in Kilobaud issue #2 on the inside cover that gave
values:
>> 10amp, 25amp rectifier bridge, and 91000mfd computer grade filter cap. My
>> nighttime reading this evening shall be "Building Power Supplies",
Archer,
>> cat 62-5025. ;)
>
>>From those components, I think it must be a linear PSU (or at least a
>mains frequency transformer).
>
>What secondary voltages do you need? What currents?
The transformer T1 from specs lists at:
7.25vac @10amp and 24vac @0.5amp secondary 120 vac@1amp/240 vac(a)0.5amp
primary, power transformer.
>One way to work those out is to determine what voltages you need across
>the smoothing caps and then divide by sqrt(2). For the currents, multiply
>the maximum load current by sqrt(2) and add a bit for safety.
>
>Now work out the total power (sum of current*voltage of course). That'll
>give you an idea of the VA rating of the transformer you need.
I will spend some time to more understand those last two paragraphs...
>The transformer might be a standard part, or it might have been custom.
>If the latter, in the UK you can get 'transformer kits' - core + bobbin
>with pre-wound mains primary. You wind the appropriate secondaries (the
>kit instructions tell you how many turns you need for 1V output) and
>assemble the core laminations.
I'd definately buy one if still in production but I definately want to make
a power supply (probably for the sym *if* that one can also not be bought.
;)
>> First question: The baud lines on both busses (110,150,300,600,1200)
caugt
>> my eye and that ad above states: "Crystal controlled oscillator( 1,7971 )
>> provides the clock signal for the processor (before?) and is divided down
by
>> the MC14411 (both on MP-A2 GK) to provide the various baud rate outputs
for
>> the interface circuts. Full buffering on all data and address busses..."
>> Does that mean that all these cards are serial?
>
>I doubt it. More likely they just bussed some useful clock frequencies
>over the backplane. Then all the serial cards (using 6850s?) could use
>them for the baud rate clocks. Saves putting a crystal + divider on each
>serial port card.
>
>> reports and make some kind of open reference. I sure would find it
usefull.
>> (read that as alot of things to repair. :)) I imagine alot of you have
>> encounters in repairland... It would be nice to have alot of this stuff
in
>
>I seem to get ever more things to repair... (and not all of them are
>classic computers...)
i would imagine. ;)
>> a prolog predicate database for pattern searches.
>
>-tony
>
>
- Mike: dogas(a)leading.net
I have started restoring a swtpc 6800 and thought I'd ask for some comments
on restoration journaling. I've completed as detailed an initial inventory
as I can manage. I recorded the two busses on MP-B and some part numbers
that were disappearing from memory chips( 4x9 819 MM5257 ) on SSB M-16-R,
and transister( 18n9 ) from rectifier bridge. The soldering job on MP-A2
has alot of resin around connnections that I'm already pretty leary of. Ill
start off by digging for all the hardware documentationI can find...
First issue: Power system is missing the transformer. I do have the MP-P,
the filter cap, and the rectifier bridge.
I found a nice ad in Kilobaud issue #2 on the inside cover that gave values:
10amp, 25amp rectifier bridge, and 91000mfd computer grade filter cap. My
nighttime reading this evening shall be "Building Power Supplies", Archer,
cat 62-5025. ;)
First question: The baud lines on both busses (110,150,300,600,1200) caugt
my eye and that ad above states: "Crystal controlled oscillator( 1,7971 )
provides the clock signal for the processor (before?) and is divided down by
the MC14411 (both on MP-A2 GK) to provide the various baud rate outputs for
the interface circuts. Full buffering on all data and address busses..."
Does that mean that all these cards are serial?
Any switheads out there? Any thoughts on the forthcomming "This Old
Computer" episode? It would be cool to accumulate these restoration
reports and make some kind of open reference. I sure would find it usefull.
(read that as alot of things to repair. :)) I imagine alot of you have
encounters in repairland... It would be nice to have alot of this stuff in
a prolog predicate database for pattern searches.
- Mike: dogas(a)leading.net
"I am but an egg" -VMS
Oh boy I'm getting in the eBay discussion. I must be really bored.
I've been using eBay for a few years now and here's how I deal with it:
When I find an "object" of interest, I set a limit as to what I'm willing to
pay for it. If the item goes beyond that, I STOP. There is no overinflating
prices because I got pissed for being outbid, it's just that someone else is
willing to pay more and so be it (I'm cheap). But so far I got some items
that I have not seen anywhere else and believe me I search...
On another side I once found an item that I had no (read NO) interest in and
I had to pay $1 for it to get it because it came with a bucnh of stuff I
wanted. I placed it on eBay and got $150 fo it (double of what I paid fo the
whole deal of stuff I wanted: about five computers and accessories) I felt
sorry for the guy who bought it but I'm sure that he was happy with his
purchase (I actually send an extra something with it out of guilt).
Now where am I doing something wrong?
What would be the ideal system?
How can you at the same time make peolple happy and optimize your return?
I'm a hobbyist, I don't speculate on value of computers but I want to be
able to:
1) Find computers that were not popular in my area and are almost impossible
to find.
2) Be able to get at least what I paid for when I sell an item (I prefer
trade though).
3) When I find something on the net, have at least a chance to get it if I
really really want it.
Francois
-----Original Message-----
From: Allison J Parent <allisonp(a)world.std.com>
>Actually while that is true you far from limited to that. I'm building a
>Z280 system, the parts of which were found from old PC motherboards and
disk
>controllers. A little torch work to remove them or their sockets netted
>Simm sockets, PLCC and other chip sockets and the z280s were socketed to
>start with.
Cool... Can I have a copy of your assembly documentation?
Thanks
- Mike
<> Homebrewing is a lost pasttime for the most part. I want to bring it
<> back, at least for myself.
<
<It's different now, you use different chips & environment. But now you can
<define your own cpu & peripherials, what is somekind of fun.
Actually while that is true you far from limited to that. I'm building a
Z280 system, the parts of which were found from old PC motherboards and disk
controllers. A little torch work to remove them or their sockets netted
Simm sockets, PLCC and other chip sockets and the z280s were socketed to
start with.
Also parts like FPGAs depending on type are easily programmable so that
custom parts are doable on modest budgets.
If anything it's better than ever... Now to get that PDP-8 clone project
going. ;)
Allison
What the BIOS does, in this case, isn't relevant, though that should work, I
guess. It's what M$ did in MSD and Windows that matters, since that's
what's getting confused.
Dick
-----Original Message-----
From: Tony Duell <ard(a)p850ug1.demon.co.uk>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Sunday, March 21, 1999 12:53 PM
Subject: Re: OT: Re: Security question (sort of)/goodbye
>> As for the serial ports, the UART or maybe ART (also a valid acronymn, I
>
>You know, I've heard of (and used) UARTs, USRTs, USARTs, ACIAs, ACEs
>(Async Communcation Element, I think, which it what Intel called the
>8250). Never ARTs...
>
>> guess, since they're hardly universal any more) these are normally
>> recognized by means of the interrupt they return. If, let's assume, the
>
>Well, the PC/AT simply looks at the Interrupt ID registers of the UARTs at
>$3FA and $2FA. If it gets the right value back, it assumes the UART
>exists. I've just checked the BIOS source listing (section TEST2)
>
>Of course modern BIOSs probably do something totally different, but without
>the source listing it's impossible to be sure.
>
>-tony
>
Would someone with a PDP-11 configuration guide describe for me what I need
to do to populate my PDP-11/23 chassis?
Here are the questions, the Chassis has a 9 x 4 Q-bus and is about 6" tall,
says PDP-11/23 on the front and has a few switches (3) and a display.
1) What is the designation for this chassis? (BA11?)
2) Looking at it from the back, what boards should be where? (I will have a
CPU,
memory, serial interface, BDV11, RX02 interface and RL01 interface.
3) Are "bus grants" required on this system as well?
4) Would an RQDX1 work in this system?
--Chuck
Hi,
Yesterday I bought an old Sony SMO-S501-11 external SCSI 650MB magneto-optical
drive. This takes 5.25" double-sided 600MB or 650MB disks; it came with 15
600MB disks.
There are some things which hopefully someone on classiccmp will know about, or
at least know where to point me.
The drive works, at least partly. I think the lens may need cleaning, since
sometimes it does not recognise disks; this is using freshly-cleaned disks.
Does anyone have a spare 5.25" MO head cleaning disk? Or a cleaning kit for
5.25" MO disks themselves? (But see below, the real fault may be
temperature-related.)
I got the drive working on my Amiga, and also an (emulated) Macintosh. Does
anyone have the driver software for MS-DOS PCs that would have come with this
drive? This is necessary to allow 650MB disks to be used under MS-DOS (650MB
disks use 1024-byte sectors).
When the drive refuses to work, inserting a disk results in:
- Drive spins up
- A click sound about every two seconds for about ten times
- Drive spins down.
This repeats twice, after which the drive reports:
Sense Key : Not Ready (0x02)
Add. SenseCode : Logical Unit Not Ready, Cause not Reportable (0x04)
I'll describe a little about the drive itself. It was made in March 1990. It is
SCSI-1, and old MO drives were much slower than current ones; according to the
manual, the drive has a read speed of 680K/sec with an average seek time of
95ms. Spindle speed is 2400rpm.
Interestingly, the actual mechanism does not seem to be SCSI. It is attached to
a SCSI controller inside the case. The drive mechanism part number is
SMO-D501-99. It uses two card-edge connectors, one 34 way the other 20 way. Is
this EDSI or something???
On the SCSI controller PCB is one firmware EPROM, whose label reads:
C501-00
CCP-2.11
? SONY '88 '89
On (one of) the mechanism's PCBs are three EPROMs, whose labels read:
D501 D501 D501-00
MMP-32 MDP-21 DCP-83
? SONY '88 '89 ? SONY '88 '89 ? SONY '88 '89
I tried to dump the data from all four EPROMs. Three seem okay, but the DCP-83
one gave errors. If this EPROM is marginal/bad, that may explain the problems
I have been having getting disks to be recognised -- it seems that immediately
after powering up the drive from cold, disks can be recognised, but after it
has warmed up the drive is much more picky. Do bad EPROMs usually behave like
this, or is it more likely to be some other component?
If anyone has one of these drives, and has dumped the DCP-83 EPROM, or knows
where I might get hold of an image of this EPROM, please let me know. I would
also like to get hold of the OEM manual(s) for this drive.
On a related subject, I want to get hold of disk geometry information for
various types of MO disk for a FAQ that I'm writing. For example, 600MB 5.25"
disks have 576999 sectors of 512 bytes. 640MB 3.5" disks have 310352 sectors
of 2048 bytes. I need this info for all other types of 5.25" MO disks.
-- Mark
I have a User's Manual and Quick Reference Guide for the 2079 terminal
that needs a home. I'm going to pitch it if no one wants it. $5 plus
shipping.
Joe
I am looking for a very simple explanation as to how the date and time
are handled in RSX-11? Just a short description which is parallel
to what is here about RT-11.
In RT-11, the date and time are two very separate quantities. There
is no interaction between the two except:
(a) When the user requests the time, the size of the time value is
checked to see if it exceeds "24 hours"
(b) If the time exceeds "24 hours", the time is reduced by "24 hours"
and the date is incremented by ONE day, the algorithm dependent
on if the SYSGEN has rollover for end of month (and after 31-Dec-99)
included or not.
The DATE command accepts dates from 01-Jan-73, although the date
of 01-Jan-72 is displayed correctly in the DIR utility. For non-Y2K
versions of RT-11 (V5.6 and prior), the DATE command accepts
and displays dates up to 31-Dec-99. For Y2K versions of RT-11
(V5.7), the DATE command accepts and displays dates up to
31-Dec-2099.
The DATE value is kept in a location within the resident monitor and
may be requested by any program via a .DATE request. The USR
tags a newly created file with the current DATE value (whatever
quantity is in the DATE value - no checking is done), otherwise,
the DATE is not used or manipulated by the operating system
(except as noted in b above) and the DATE command. Other
than the utilities (DIR, PIP, etc.) which look at the DATE and
especially look at the creation date of a file (which is not actually
looking at the actual date - but the DATE value at the time when
the file was created), RT-11 does not care about or use the DATE
in any manner and the operating system is totally independent of
the quantity in the DATE value and unaffected by that quantity.
I guess that changing the creation date of a file via PIP to be
the current DATE value via the SET option in PIP is an exception,
but no different than what happens when the USR creates a files
and uses the current DATE value as the creation date.
There could be application programs set up by the user which
do more than simply display or set the DATE value. DATIME.SAV
only does what the DATE command does, but a program executing
as a system job could do much more. In general, I want to ignore
such possibilities in both RT-11 and RSX-11.
One other very minor aspect. In RT-11, the LTC is used to increment
the time at either 50 or 60 Hertz. The time value is available to the
user in ticks. An RT-11 program can request certain time based
operations in terms of the number of ticks. Does RSX-11 have
similar types of things available?
Sincerely yours,
Jerome Fine
RT-11/TSX-PLUS User/Addict
