10 Apr
2013
10 Apr
'13
2:25 a.m.
allison <ajp166 at verizon.net> wrote:
Generally yes. TU58 file format is variable in
that the device appears to the system as
a blocked volume like disk (though slower). The transfer between the TU58 and
its host is serial data and the host request block(s) N and not that
much different than IDE or other intelligent disk system that uses a logial block
numbering
system. The tu58 does not know nor care what is in the 512 byte block. I've run CP/M
from TU58.
ok, then TU58 emulation is, as I had hoped, going to float the boat for me, at least until
either more machines or media to be read crop up here.
Thats plenty of space for PDP-11 OSs.
Unfortunately I have no sound way to (mechanically) _mount_ and power those tiny little
5,25" disks in the '11 rack, that's why they're staying in the BA123.
I forgot the 300MB ESDI disk I got from a listmember together with a Webster Qbus
controller, I'll use that one if I ever want to run a large OS on the VSII.
On 04/09/2013
12:08 PM, Arno Kletzander wrote:
I'm not sure what I'd need it for? There is no other system within walking
distance that would understand the floppy format natively.
Any file I/O thus most likely will happen across the Internet (involving a PC anyway), so
I hoped I'd be fine with the TU58 emulation software discussed below.
I use the PC to make floppies (RX50, RX33 and RX23) and my S100 CP/M system will do RX01
transfers.
But, I have multiple Qbus 11s and most all have floppies so portability between them is
handy. (or a 50ft serial cable). If anything the floppy is always a must on my systems
as all my diags and
base RT11 systems are on that media (RX01, RX02, RX50, RX33, RX23).
Are they also
available in TU58 (file) format? A viable uVAX is more than 150mb, (more like 300-500
for V7),
a loaded PDP11 is 30MB. Just difference is OS utilization.
So I guess it all boils
down on what one wants to run on it.
I've got one functional RD53 in it right now, and a second one I hope to revive at
some point in the future. Likely the bad one has the stuck head problem
easy to fix. just upen it up and unstick it. There are more detailed instruction out
there.
I'v edone it many times and forget the whole you need a clean room thing, My first
salvage is near 25years old now.
I'd not want to try my luck, I'll work on that one once I've got an aquarium
type glove box set up.
[large drives]
You can till you have to format or defrag it. I keep a
stack of sub 1gb
(from 50mb and up) SCSI drives for that reasonas well as a major
heap of MFM drives (St412s 10mb, St225s 20mb and Quantum
D540s 31mb). I just insert them and image copy to them as backup.
OK, seems we mixed up logically and physically large drives here.
I have about none of the first, but several of the latter category amongst those in
question for the machines we're discussing here:
1(2)x RD53 5,25" MFM
1x 300MB 5,25" ESDI,
3x 5MB 14" RL01,
1x 10MB 14" RL02,
(2x 80MB 14" SMD) (I just have a lead on those yet)
[TK50 tape subsystem]
I meant save it for backups or other large data needs
as running diags
off it will be painfully slow. If its bad save it for parts, they are scarce,
usually the controller is fine and the drive is borked.
Understood. I'll find out once I have a drive and media for it.
I've also already found out that the memory board
I'll be getting is 512k_Bytes_ (256kW) and does _not_ have BBU support.
?? BBU?? Unless you mean BBS7 (Io tends to use that).
Battery backup. Feeding it from an alternate 5V source to keep RAM content intact when
mains power is off.
One last thing...
Heat! Qbus 11s produce a bit of that. That means fans must all work,
the location must not be dusty enough to load up the boards and local
temperature not excessively hot.
I fried a 11/23 board while working at the DEC Mill when they had an air
conditioner fail for the office area, when the room hit 96 (36c), cpu
went away, I had plenty of spares. So reliability and room temps are
coupled. Qbus 11s are fairly tolerent of conditions but any you find
are going to be old and may not like additional stress.
A hard "CPU dead" damage without any warning? That's bad.
Good to know anyway, makes me want to add a system monitoring function
(temp and air flow) for emergency shutdown. Maybe something to design
a CD slot board for after all ;), at least before I'll run the machine
unattended.
I also forgot something: I myself wouldn't have thought of putting a NIC
(DEQNA or similar) in a pdp-11, probably just for the notion that back
in those days, computers were too few and far between to come up with the
concept of a Local Area Network. Obviously I was mistaken here, too.
Arno
10 Apr
10 Apr
11:42 a.m.
Arno Kletzander wrote:
OK, seems we mixed up logically and physically large
drives here.
I have about none of the first, but several of the latter category amongst those in
question for the machines we're discussing here:
1(2)x RD53 5,25" MFM
Please be aware that most (maybe now almost all) RD53
drives have failed with the head stuck problem. If you have
an RD53 which can still be read, it is strongly suggested that
you copy the contents to a more reliable drive. It is possible
to open and remove the sticky mess without a clean room,
but read the notes over the past 5 or 10 years which describe
how it should be done. I would agree to use an RD53 as a
scratch drive for temporary storage, but not for saving files
after the drive is powered down. If anyone has any MTBF
stats after the sticky mess has been removed in terms of
how many hours of actually using the drive, that would be
helpful.
1x 300MB 5,25" ESDI,
I have a number of Micropolis ESDI and Hitachi ESDI
drives which seem to be quite reliable. Most are the
Hitachi DK-515 ESDI 5 1/4" 600 MB drives. They
are used with a PDP-11/83 system with their own
PC type power supply and fan since I doubt that using
the internal power supply from the BA123 is helpful.
Those Hitachi drives arrived 4 in a box with that PC
type power supply in the first place which emulated
one of the VAX type of drives which DEC supported.
Since the ESDI drives were exactly what I wanted for
the PDP-11/83 and the DEC reseller had no demand
for these drives from VAX users, the price and usage
was perfect for the PDP-11/83 with an MSCP EDSI
controller. Normally, I used 3 * 600 MB drives which
were usually all identical with drive 0 being the production
drive and drives 1 and 2 being backups.
3x 5MB 14" RL01,
1x 10MB 14" RL02,
VERY reasonable for a small RT-11 system, but prone
to hardware problems (especially filters) in the long run.
(2x 80MB 14" SMD) (I just have a lead on those
yet)
The ESDI drives are more reliable and faster. ESDI controllers
were also less expensive and more available.
[TK50 tape subsystem]
From my point of view, one of the worst backups that
DEC ever produced. While making a simple copy of
a large file is reasonable, the lack of a reasonable
streaming capability makes use of small files or a verify
of a large file extremely difficult. On the other hand,
the TK70 is a good, fast, reliable drive which I used
for many years as my primary backup until I found
some Sony SMO S-501 magneto optical drives.
Jerome Fine
6:42 p.m.
On 04/10/2013 11:42 AM, Jerome H. Fine wrote:
I have one I did that on back in the late 80s and it still runs.
I's gotten at least three years solid power on time in a PC
and after retirement there Its in a current PDP11 and still
runs at least 30 hours a year.
I have a bunch I've done that to. A few have failed due to
controller board issues (saved as media spares) and one
the media is physically good but the controller borked
and wrote all over the servo tracks. No reformat possible
for the servo information.
It was a very good drive save for the gummy rubber bumper
in the head assembly.
The Hitachi were among the best.
The Quantum D540s were also near bullet proof.
RL02 was far more reliable and more common. 10MB happens to
be a handy size for PDP11.
;)
I just can't say enough nice things about TK50.
(set user/mode:sarcasm=off)
But they were the common transfer media and big for its time.
that and I have a small bin full of Compact Tape I media with
stuff on it. Tk70 was the one to have.
There is also TLZ04 tape.
Allison
Arno Kletzander
wrote:
OK, seems we mixed up logically and physically
large drives here.
I have about none of the first, but several of the latter category
amongst those in question for the machines we're discussing here:
1(2)x RD53 5,25" MFM
Please be aware that most (maybe now almost all) RD53
drives have failed with the head stuck problem. If you have
an RD53 which can still be read, it is strongly suggested that
you copy the contents to a more reliable drive. It is possible
to open and remove the sticky mess without a clean room,
but read the notes over the past 5 or 10 years which describe
how it should be done. I would agree to use an RD53 as a
scratch drive for temporary storage, but not for saving files
after the drive is powered down. If anyone has any MTBF
stats after the sticky mess has been removed in terms of
how many hours of actually using the drive, that would be
helpful.
1x 300MB
5,25" ESDI,
I have a number of Micropolis ESDI and Hitachi ESDI
drives which seem to be quite reliable. Most are the
Hitachi DK-515 ESDI 5 1/4" 600 MB drives. They
are used with a PDP-11/83 system with their own
PC type power supply and fan since I doubt that using
the internal power supply from the BA123 is helpful.
Those Hitachi drives arrived 4 in a box with that PC
type power supply in the first place which emulated
one of the VAX type of drives which DEC supported.
Since the ESDI drives were exactly what I wanted for
the PDP-11/83 and the DEC reseller had no demand
for these drives from VAX users, the price and usage
was perfect for the PDP-11/83 with an MSCP EDSI
controller. Normally, I used 3 * 600 MB drives which
were usually all identical with drive 0 being the production
drive and drives 1 and 2 being backups.
Most of the VAXEN (qbus) wanted drive larger than 300mb
as VMS V5 barely fits in 150mb and with users and added
software even 300mb is getting thin.
3x 5MB
14" RL01,
1x 10MB 14" RL02,
VERY reasonable for a small RT-11 system, but prone
to hardware problems (especially filters) in the long run.
(2x 80MB
14" SMD) (I just have a lead on those yet)
The ESDI drives are more reliable and faster. ESDI controllers
were also less expensive and more available.
[TK50 tape subsystem]
From my point of view, one of the worst backups that
DEC ever produced. While making a simple copy of
a large file is reasonable, the lack of a reasonable
streaming capability makes use of small files or a verify
of a large file extremely difficult. On the other hand,
the TK70 is a good, fast, reliable drive which I used
for many years as my primary backup until I found
some Sony SMO S-501 magneto optical drives.
6:20 p.m.
On 04/10/2013 02:25 AM, Arno Kletzander wrote:
Generally yes. TU58 file format is variable in
that the device appears to the system as
a blocked volume like disk (though slower). The transfer between the TU58 and
its host is serial data and the host request block(s) N and not that
much different than IDE or other intelligent disk system that uses a logial block
numbering
system. The tu58 does not know nor care what is in the 512 byte block. I've run CP/M
from TU58. ok, then TU58 emulation is, as I had hoped, going to float the boat for
me, at least until either more machines or media to be read crop up here. The
novelty weill wear off fast as a drive is only 256K, and its serial
line speed
and hence sloowwwwwwww...
Thats plenty of space for PDP-11 OSs.
Unfortunately I have no sound way to (mechanically) _mount_ and power those tiny little
5,25" disks in the '11 rack, that's why they're staying in the BA123.
I forgot the 300MB ESDI disk I got from a listmember together with a Webster Qbus
controller, I'll use that one if I ever want to run a large OS on the VSII.
I got a rack shelf and gutted a few PCs for the disk cage.
Suit yourself.
I've done them on the kitchen table. I must ahve done
at least 8 of them.
The only failures were not media related (controller dies on write and
blows the
track servo).
Keep in mind they have a filter and the spin-up is to clear the platters
and the dust that lands into the filter.
[large drives]
Physical large drive: any that want a lift or two people to install in
the rack.
Anything I can lift with one hand is small. Media capacity is a whole
different
matter.
Yep, however in the Mill on weekends the AC
was running lower and who
know how
many thermal cycles that machine had seen even before I got it.
If anything
the floppy is always a must on my systems as all my diags and
base RT11 systems are on that media (RX01, RX02, RX50, RX33, RX23).
Are they also
available in TU58 (file) format? A viable uVAX
is more than 150mb, (more like 300-500 for V7),
a loaded PDP11 is 30MB. Just difference is OS utilization.
So I guess it all boils
down on what one wants to run on it.
I've got one functional RD53 in it right now, and a second one I hope to revive at
some point in the future. Likely the bad
one has the stuck head problem
easy to fix. just upen it up and unstick it. There are more detailed instruction out
there.
I'v edone it many times and forget the whole you need a clean room thing, My first
salvage is near 25years old now.
I'd not want to try my luck, I'll work on
that one once I've got an aquarium type glove box set up. You can till
you have to format or defrag it. I keep a stack of sub 1gb
(from 50mb and up) SCSI drives for that reasonas well as a major
heap of MFM drives (St412s 10mb, St225s 20mb and Quantum
D540s 31mb). I just insert them and image copy to them as backup.
OK, seems we
mixed up logically and physically large drives here.
I have about none of the first, but several of the latter category amongst those in
question for the machines we're discussing here:
1(2)x RD53 5,25" MFM
1x 300MB 5,25" ESDI,
3x 5MB 14" RL01,
1x 10MB 14" RL02,
(2x 80MB 14" SMD) (I just have a lead on those yet) [TK50 tape subsystem]
Oh, It's been a while, I never use it. No need. If I need the data
preserved
I have it on disk and the system on UPS.
I meant save it for backups or other large data
needs as running diags
off it will be painfully slow. If its bad save it for parts, they are scarce,
usually the controller is fine and the drive is borked.
Understood. I'll find
out once I have a drive and media for it.
I've also already found out that the memory
board I'll be getting is 512k_Bytes_ (256kW) and does _not_ have BBU support.
?? BBU?? Unless you mean BBS7 (Io tends to use that).
Battery backup. Feeding it from an alternate 5V source to keep RAM content intact when
mains power is off.
One last
thing...
Heat! Qbus 11s produce a bit of that. That means fans must all work,
the location must not be dusty enough to load up the boards and local
temperature not excessively hot.
I fried a 11/23 board while working at the DEC Mill when they had an air
conditioner fail for the office area, when the room hit 96 (36c), cpu
went away, I had plenty of spares. So reliability and room temps are
coupled. Qbus 11s are fairly tolerent of conditions but any you find
are going to be old and may not like additional stress.
A hard "CPU
dead" damage without any warning? That's bad.
Good to know anyway, makes me want to add a system monitoring function
(temp and air flow) for emergency shutdown. Maybe something to design
a CD slot board for after all ;), at least before I'll run the machine
unattended.
Mine ran usually for months at a time, if the building AC hadn't failed
the cpu baord would still be in there (it is actually and repaired returned
to my spares box).
I also forgot something: I myself wouldn't have
thought of putting a NIC
(DEQNA or similar) in a pdp-11, probably just for the notion that back
in those days, computers were too few and far between to come up with the
concept of a Local Area Network. Obviously I was mistaken here, too.
PDP-11s were networked before NICs using DDCMP and sync lines. After NICS
Ether pipe. RT11 does little with it and its mostly useless for that
RSTS and RSX
in the later versions were full fledged nodes (DECnet Phase III).
Of course Unix had UUCP via serial lines and modems.
Allison
11 Apr
11 Apr
10:49 a.m.
On Apr 10, 2013, at 2:25 AM, Arno Kletzander <Arno_1983 at gmx.de> wrote:
That's a fun idea. When I'm running my 11/23 in an open backplane,
I make sure I have a decently powerful fan blowing through it and
over the top. Those boards tend to dissipate between 10 and 20
watts each, which is nothing to sneeze at, and the CPUs can get
quite locally hot.
ok, then TU58 emulation is, as I had hoped, going to
float the boat for me, at least until either more machines or media to be read crop up
here.
If you manage to find a QBUS SCSI card (I have the CMD CQD-220, which
I like a lot), you can use a Zip drive for removable storage (or even
primary storage, though having done that for a bit I'm not sure I like
the idea of torturing the disks like that). It's not exactly "classic",
but they are plentiful and cheap and fairly reliable if you stick with
the 100 MB versions. They're also BIG (in storage size) compared to
an RL02, and they're much better about being dropped. :-)
Unfortunately I have no sound way to (mechanically)
_mount_ and power those tiny little 5,25" disks in the '11 rack, that's why
they're staying in the BA123.
I forgot the 300MB ESDI disk I got from a listmember together with a Webster Qbus
controller, I'll use that one if I ever want to run a large OS on the VSII.
You could always run your -11 from the BA123. It's a bit of overkill,
but you can fit all the disks you want in there and pretty much never
worry about running out of slots for boards.
Of course, if you're actively using the MVII, don't do that.
One last
thing...
Heat! Qbus 11s produce a bit of that. That means fans must all work,
the location must not be dusty enough to load up the boards and local
temperature not excessively hot.
I fried a 11/23 board while working at the DEC Mill when they had an air
conditioner fail for the office area, when the room hit 96 (36c), cpu
went away, I had plenty of spares. So reliability and room temps are
coupled. Qbus 11s are fairly tolerent of conditions but any you find
are going to be old and may not like additional stress.
A hard "CPU dead" damage without any warning? That's bad.
Good to know anyway, makes me want to add a system monitoring function
(temp and air flow) for emergency shutdown. Maybe something to design
a CD slot board for after all ;), at least before I'll run the machine
unattended. I also forgot something: I myself wouldn't have
thought of putting a NIC
(DEQNA or similar) in a pdp-11, probably just for the notion that back
in those days, computers were too few and far between to come up with the
concept of a Local Area Network. Obviously I was mistaken here, too.
The QBUS LAN cards definitely have more of a "meant for VAX" feel
to their documentation, IMO, but they do work fine in an -11.
If you can find a DECNET distribution, you can usually even get
the drivers for them running on a DEC OS as well! TCP/IP is
somewhat rare, though, because the stack takes up so much room.
I know Johnny Billquist mentioned that he had at least UDP
running on RSX-11M+; I can't find the email now, so I can't
remember about TCP. There are commercial TCP/UDP stacks for
the DEC operating systems as well.
If you can, I'd recommend getting a DELQA instead of a DEQNA.
The DELQA is backwards-compatible with the DEQNA and is a lot
more reliable (mostly due to lower power dissipation). I'm
given to understand it takes a lot of load off the CPU as well.
The DELQA can use the DEQNA cable kit, so no need to worry
about that.
- Dave
11:57 a.m.
On 04/11/2013 10:49 AM, David Riley wrote:
The QBUS LAN cards definitely have more of a
"meant for VAX" feel
to their documentation, IMO, but they do work fine in an -11.
If you can find a DECNET distribution, you can usually even get
the drivers for them running on a DEC OS as well!
And HECnet!! Tremendous fun!
TCP/IP is
somewhat rare, though, because the stack takes up so much room.
I know Johnny Billquist mentioned that he had at least UDP
running on RSX-11M+; I can't find the email now, so I can't
remember about TCP. There are commercial TCP/UDP stacks for
the DEC operating systems as well.
TCP as well!!
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
12:37 p.m.
On 04/11/2013 10:49 AM, David Riley wrote:
I have a complete 11 in a BA123 and
that is the nicest as there are
plenty of
disk bays and the fans are speed controlled and generalaly quieter.
Also its easier to use the RQDX controller with the breakout board for
floppy
and RDxx disks. imagine rx33m and 4 RD52s with room for more (TK tape bay)
Add a CMD disk controller to that and use 1RZ55 and 3 RZ56s (5.4gb
total) for
uVAXII and that is seriously plenty. I used those as I have 10 of them and
backup is a drive swap after running SABackup (image the system disk).
Of course with older style PC 3.5" adaptors for the 5.25" bays 1gb
and 4.3GB Baracudas can be loaded in droves.
That's a fun idea. When I'm running my 11/23 in an open
backplane,
I make sure I have a decently powerful fan blowing through it and
over the top. Those boards tend to dissipate between 10 and 20
watts each, which is nothing to sneeze at, and the CPUs can get
quite locally hot. Another area where the BA123 is better, cooling and airflow.
The DEQNA has the mop boot for PDP11. seems that is a PDP11 thing.
The DELQA deleted that.
However good DEQNAs are scarce and very old, DELQAs were better
and more common. In any case they existed before the Qbus MicroVAX
became.
On Apr 10, 2013, at 2:25 AM, Arno Kletzander
<Arno_1983 at gmx.de> wrote:
The CMD SCSI card is the way to go.
ok, then TU58 emulation is, as I had hoped, going
to float the boat for me, at least until either more machines or media to be read crop up
here.
If you manage to find a QBUS SCSI card (I have the CMD CQD-220, which
I like a lot), you can use a Zip drive for removable storage (or even
primary storage, though having done that for a bit I'm not sure I like
the idea of torturing the disks like that). It's not exactly "classic",
but they are plentiful and cheap and fairly reliable if you stick with
the 100 MB versions. They're also BIG (in storage size) compared to
an RL02, and they're much better about being dropped. :-) Unfortunately
I have no sound way to (mechanically) _mount_ and power those tiny little 5,25" disks
in the '11 rack, that's why they're staying in the BA123.
I forgot the 300MB ESDI disk I got from a listmember together with a Webster Qbus
controller, I'll use that one if I ever want to run a large OS on the VSII.
You
could always run your -11 from the BA123. It's a bit of overkill,
but you can fit all the disks you want in there and pretty much never
worry about running out of slots for boards. Of course, if you're actively using the MVII,
don't do that.
Get another... ;)
One last thing...
Heat! Qbus 11s produce a bit of that. That means fans must all work,
the location must not be dusty enough to load up the boards and local
temperature not excessively hot.
I fried a 11/23 board while working at the DEC Mill when they had an air
conditioner fail for the office area, when the room hit 96 (36c), cpu
went away, I had plenty of spares. So reliability and room temps are
coupled. Qbus 11s are fairly tolerent of conditions but any you find
are going to be old and may not like additional stress.
A hard "CPU
dead" damage without any warning? That's bad.
Good to know anyway, makes me want to add a system monitoring function
(temp and air flow) for emergency shutdown. Maybe something to design
a CD slot board for after all ;), at least before I'll run the machine
unattended. I also forgot
something: I myself wouldn't have thought of putting a NIC
(DEQNA or similar) in a pdp-11, probably just for the notion that back
in those days, computers were too few and far between to come up with the
concept of a Local Area Network. Obviously I was mistaken here, too.
The QBUS LAN
cards definitely have more of a "meant for VAX" feel
to their documentation, IMO, but they do work fine in an -11. If you can find a DECNET distribution, you can usually
even get
the drivers for them running on a DEC OS as well! TCP/IP is
somewhat rare, though, because the stack takes up so much room.
I know Johnny Billquist mentioned that he had at least UDP
running on RSX-11M+; I can't find the email now, so I can't
remember about TCP. There are commercial TCP/UDP stacks for
the DEC operating systems as well.
If you can, I'd recommend getting a DELQA instead of a DEQNA.
The DELQA is backwards-compatible with the DEQNA and is a lot
more reliable (mostly due to lower power dissipation). I'm
given to understand it takes a lot of load off the CPU as well.
The DELQA can use the DEQNA cable kit, so no need to worry
about that.
The CPU load this was added around V5.* VMS as they discovered
a DEQNA bug that could corrupt packets in a uVAX system so the
packet has a secondary CRC16 added and does the reverse on
the other side to check the data and thats all in software (real pita).
Allison
- Dave
4154
days inactive
4155
days old
6 comments
5 participants
participants (5)
-
ajp166＠verizon.net -
Arno_1983＠gmx.de -
fraveydank＠gmail.com -
jhfinedp3k＠compsys.to -
mcguire＠neurotica.com