7 Feb
2012
7 Feb
'12
12:16 p.m.
I have a license for VMS 5.2 and was considering buying a working system to run it on. My
microvax I finally gave up the ghost, and I would rather spend time learning than going
through repair at the moment.I have many of the parts below from the uvax 1, but who knows
whats working and whats not, I need to test.
I have a good configuration for the MicroVAX II to build on, but I was looking at the
MicroVAX 3900 since the price disparity is not too significant.
Unfortunately, I don't have a good hardware build plan for the 3900. I am hoping I
can get some good guidance here.
My largest concern is that the hardware is supported by the OS more than anything else.I
want to stick with QBUS.I need to have more than 4 async serial ports with modem control
(so I can hook it up to other systems) considering the DHV11I need ethernet, thinking
DEQNAIm considering the TK50 tape (I would go tk70 if I knew it was backward compat,
I'm just not sure)I dont know which disk controllers to get. I'm guessing it
should be in the KDA50 with RA82 disks, but I have no idea, just guessing.The standard
RX50 floppy.Looking at the VCB02 for video, just as a nicety.And then 64Mb ram.
I have never dealt with the 3900 hardware at all, and I have the hobby funds to procure
something that works.Guessing that I can do it for less than $2k even if I had to buy all
"new" parts. Obviously I want to limit cost as much as possible, but I would
also like to be able to run netbsd on the hardware at a later date if it was interesting
to do so.
I understand that the 3900 can be converted to a microvax III+ with some upgrades, but
will it run VMS 5.2?
Any recommendations?Kevin
7 Feb
7 Feb
1:25 p.m.
On Feb 7, 2012 9:38 AM, "Kevin Reynolds" <tpresence at hotmail.com>
wrote:
I have a license for VMS 5.2 and was considering buying a working system
to run it
on. My microvax I finally gave up the ghost, and I would rather
spend time learning than going through repair at the moment.
Do you want to stick with 5.2 or would you be open to running 7.3 with
hobbyist licenses? I've tinkered with 7.3 on Q-bus VAX from KA630 to KA692.
I haven't done any 5.2 installs and don't know where support for that tops
out.
What kind of cabinet / backplane do you have? KA650 and KA655 CPUs should
be pretty easy to find. 16MB M7622 boards shouldn't be too hard to find
either, although if you want 64MB you might be limited by your backplane.
-Glen
2:14 p.m.
Do you want to stick with 5.2 or would you be open to
running 7.3 with
hobbyist licenses? I've tinkered with 7.3 on Q-bus VAX from KA630 to KA692.
I haven't done any 5.2 installs and don't know where support for that tops
out.
What kind of cabinet / backplane do you have? KA650 and KA655 CPUs should
be pretty easy to find. 16MB M7622 boards shouldn't be too hard to find
either, although if you want 64MB you might be limited by your backplane.
-Glen
The cabinet for the microvax I is different that the BA123 and the BA23 that are available
for the later vaxes I believe. ?This model has a chassis which is
6"x28"x22", and a KD23 CPU (probably faulty). ?It has a maximum RAM
capability of 4MB. ?It definitely cannot run netbsd later than 1.61, maybe nothing before
that, but I haven't validated...
I figured that I had gone as far as I could go with the hardware anyway. ?If I can't
find a good priced replacement in the uv-II or uv-III/III+ line maybe I'll look for a
replacement CPU (if thats all that failed).
I do want to stick with the 5.2 license for the time being, perhaps I will play with my
7.3 hobbyist license when I get bored with it.
Kevin
4:40 p.m.
On Tue, Feb 7, 2012 at 11:14 AM, Kevin Reynolds <tpresence at hotmail.com> wrote:
The cabinet for the microvax I is different that the BA123 and the BA23 that are
available for the later vaxes I believe. ?This model has a chassis which is
6"x28"x22", and a KD23 CPU (probably faulty). ?It has a maximum RAM
capability of 4MB. ?It definitely cannot run netbsd later than 1.61, maybe nothing before
that, but I haven't validated...
I figured that I had gone as far as I could go with the hardware anyway. ?If I can't
find a good priced replacement in the uv-II or uv-III/III+ line maybe I'll look for a
replacement CPU (if thats all that failed).
I do want to stick with the 5.2 license for the time being, perhaps I will play with my
7.3 hobbyist license when I get bored with it.
Kevin
I have never seen a MicroVAX I, but according to this manual the
backplane is the same H9278-A that you would find in a BA23:
http://www.bitsavers.org/pdf/dec/vax/610/EK-KD32A-TD-002_MicroVAX_I_CPU_Tec…
Page 1-11, Backplane - H9278-A
Page 1-12, Figure 1-3 MicroVAX I Backplane
With a H9278-A you could install either a KA630 with up to 16MB, or a
KA650 or KA655 with up to 32MB with DEC memory or up to 64MB with 3rd
party memory. The limitation being the (3x) Q22-CD slots in the
H9278-A.
If you're willing to settle for a KA630 and willing to wait for decent
deal you can sometimes find those in the $10-$20 range on eBay. There
are some there now for $40 including shipping. Where are you located?
You might find someone on this list that has spare KA630/KA650/KA655
boards they are willing to part with for a reasonable exchange. I
think I probably have a couple KA630 boards myself that I don't have
installed in any systems. I don't know if I have any spare console
bulkhead connectors that haven't been corroded by battery leaks.
-Glen
5:57 p.m.
On 07/02/2012 21:40, Glen Slick wrote:
On Tue, Feb 7, 2012 at 11:14 AM, Kevin
Reynolds<tpresence at hotmail.com> wrote:
>
> The cabinet for the microvax I is different that the BA123 and the
> BA23 that are available for the later vaxes I believe. This model
> has a chassis which is 6"x28"x22", and a KD23 CPU (probably
> faulty).
I have never seen a MicroVAX I, but according to this
manual the
backplane is the same H9278-A that you would find in a BA23:
Is it not just a BA23? I've not seen the machine you're describing, and
the tech manual is light on pictures, but the front panel is the
standard BA23 front panel and the dimensions are about right. Without
the normal skins that make it look like a storage heater, a BA23 is a
chassis which contains all the parts and can be rack-mounted and there's
a tabletop version as well. I have seen MicroVAX Is in BA23s, they were
moderately common, and I have seen one rack-mounted.
See pictures in the BA23 manual at
http://scandocs.trailing-edge.com/micropdp11-volume_2_enclosures-EK-186AA-M…
Pages 1.2, 1.14 show the chassis layout
Page 2.2 shows the various enclosures
Page 2.20 shows the tabletop appearance; it's just the pedestal turned
sideways and with the foot removed.
Figure 8.3 in the MicroPDP-11 Systems Maintenance Guide has a better
picture, but I don't know if that's online anywhere.
--
Pete Peter Turnbull
Network Manager
University of York
6:34 p.m.
On 02/07/2012 04:40 PM, Glen Slick wrote:
On Tue, Feb 7, 2012 at 11:14 AM, Kevin
Reynolds<tpresence at hotmail.com> wrote:
Most MicroVAX-I machines shipped in BA23s, but I believe they were
also available in BA123s. They are not picky; they just require a
standard 22-bit Q/CD backplane. I've run them in BA11-SA boxes before.
-Dave
--
Dave McGuire
New Kensington, PA
The cabinet for the microvax I is different that the BA123 and the BA23 that are
available for the later vaxes I believe. This model has a chassis which is
6"x28"x22", and a KD23 CPU (probably faulty). It has a maximum RAM
capability of 4MB. It definitely cannot run netbsd later than 1.61, maybe nothing before
that, but I haven't validated...
I figured that I had gone as far as I could go with the hardware anyway. If I can't
find a good priced replacement in the uv-II or uv-III/III+ line maybe I'll look for a
replacement CPU (if thats all that failed).
I do want to stick with the 5.2 license for the time being, perhaps I will play with my
7.3 hobbyist license when I get bored with it.
Kevin
I have never seen a MicroVAX I, but according to this manual the
backplane is the same H9278-A that you would find in a BA23: 
8 Feb
8 Feb
10:23 a.m.
On Tue, Feb 7, 2012 at 2:14 PM, Kevin Reynolds <tpresence at hotmail.com> wrote:
The cabinet for the microvax I is different that the
BA123 and the BA23 that are available for the later vaxes I believe.
Nope. Bog-standard BA23 or BA123. There are later cabinets for
MicroVAX IIIs, but they are models like the BA213. You can take a
MicroPDP-11/23 and re-stuff it as a MicroVAX-I or a MicroVAX-II,
though you will probably need different peripherals (I don't think the
DLV11/DLV1J is supported under VMS, for example, though the RLV12 is).
The uVAX-I has a dual-board CPU that must be installed in slots 1-2 of
a backplane with CD slots, and the memory sits _on_ the Qbus, so with
22 address bits, you top out at 4MB. The later MicroVAXen fit on a
single board and use PMI memory (different boards, matched to the type
of CPU). I don't know the MicroVAX III memory max off the top of my
head (after my time), but the uVAX-II has 1MB on-board and can take,
IIRC, 12MB in the first two slots.
Besides memory, the big limitation I remember from those days about
upgrades was disk size. The uVAX-I shipped with the RQDX1 which did
not support disks over 30MB, I think. (RD52 but not RD53, IIRC). The
RQDX3 supports DEC disks up to the RD54 (154MB) and non-DEC disks,
though it can be tricky to get the Field Service formatter which can
label non-standard disks. Even so, MFM disks top out pretty low. I
think I managed to fit VMS 6.0 on an RD54 on a 9MB uVAX-II, but I had
to leave something behind (like the HELP files). Using an MSCP SCSI
controller bypasses this dance and leaves memory as the critical
resource.
11:06 a.m.
Using an MSCP SCSI
controller bypasses this dance and leaves memory as the critical
resource.
Ethan,
Can I put in a SCSI controller on the QBUS of a microvax II or III? I really want to keep
the QBUS as it is entirely different than everything else I have. Do you know if there is
SCSI support inVMS 5.2?
Kevin
1:14 p.m.
On Feb 8, 2012 8:49 AM, "Kevin Reynolds" <tpresence at hotmail.com>
wrote:
Can I put in a SCSI controller on the QBUS of a microvax II or III? I
really want
to keep the QBUS as it is entirely different than everything
else I have. Do you know if there is SCSI support inVMS 5.2?
Kevin
You are not really concerned about native SCSI support but rather MSCP
(disk) and TMSCP (tape). If the OS supports an RQDX controller it should
support an MSCP SCSI controller such as the usual CMD CQD Q-bus SCSI
controllers.
-Glen
1:21 p.m.
On Feb 8, 2012, at 1:14 PM, Glen Slick wrote:
On Feb 8, 2012 8:49 AM, "Kevin Reynolds"
<tpresence at hotmail.com> wrote:
To my knowledge, VMS doesn't really even see past the MSCP at all; there's
not much way it could distinguish between a SCSI disk on a CQD-220 and an
MFM disk on an RDQX3.
- Dave
Can I put in a SCSI controller on the QBUS of a microvax II or III? I
really want
to keep the QBUS as it is entirely different than everything
else I have. Do you know if there is SCSI support inVMS 5.2?
Kevin
You are not really concerned about native SCSI support but rather MSCP
(disk) and TMSCP (tape). If the OS supports an RQDX controller it should
support an MSCP SCSI controller such as the usual CMD CQD Q-bus SCSI
controllers.
1:29 p.m.
On 02/08/2012 01:21 PM, David Riley wrote:
On Feb 8, 2012, at 1:14 PM, Glen Slick wrote:
I can confirm that this is indeed the case. Most will look like an
RA81 or similar. (same MSCP interface, same registers etc)
-Dave
--
Dave McGuire
New Kensington, PA
On Feb 8, 2012 8:49 AM, "Kevin
Reynolds"<tpresence at hotmail.com> wrote:
To my knowledge, VMS doesn't really even see past the MSCP at all; there's
not much way it could distinguish between a SCSI disk on a CQD-220 and an
MFM disk on an RDQX3.
Can I put in a SCSI controller on the QBUS of a microvax II or III? I
really want
to keep the QBUS as it is entirely different than everything
else I have. Do you know if there is SCSI support inVMS 5.2?
Kevin
You are not really concerned about native SCSI support but rather MSCP
(disk) and TMSCP (tape). If the OS supports an RQDX controller it should
support an MSCP SCSI controller such as the usual CMD CQD Q-bus SCSI
controllers. 
2:11 p.m.
On 2/8/2012 10:14 AM, Glen Slick wrote:
I have a CMD CQD-220 SCSI card in my mVAX II and it works great. Boots
off CD and hasn't had a problem with any SCSI disk I've thrown at it yet.
Can I put in a
SCSI controller on the QBUS of a microvax II or III? I
really want to keep the
QBUS as it is entirely different than everything
else I have. Do you know if there is SCSI support inVMS 5.2? 
7:31 p.m.
On 02/08/2012 02:11 PM, David Clark wrote:
On 2/8/2012 10:14 AM, Glen Slick wrote:
I have a CMD CQD-220 SCSI card in my mVAX II and it works great. Boots
off CD and hasn't had a problem with any SCSI disk I've thrown at it yet.
There is no SCSI support in VMS at all. However, any SCSI controller
that emulates MSCP protocol is by default supported.
The RQDX1/2/3 controller for floppy and hard disk does that, CMD CQD-220
SCSI card are examples of that.
The may be others but those are the most known.
Allison
Can I put
in a SCSI controller on the QBUS of a microvax II or III? I
really want to keep
the QBUS as it is entirely different than everything
else I have. Do you know if there is SCSI support inVMS 5.2?
7:48 p.m.
On 2/8/2012 4:31 PM, allison wrote:
There is no SCSI support in VMS at all. However, any
SCSI controller
that emulates MSCP protocol is by default supported.
The RQDX1/2/3 controller for floppy and hard disk does that, CMD
CQD-220 SCSI card are examples of that.
The may be others but those are the most known.
Allison
Right. It's transparent to the user though. The SCSI disk just shows up
as DUB<scsID> and the card does all the work. No changes to VMS are
needed. Makes it very simply. Even easier than putting a SCSI controller
in a PC. They have a full set of SCSI utilities for formatting and
setting up drives built into the card. Actually booting them from the
monitor can be a challenge though. You'll need the manual for sure.
8:18 p.m.
On Wed, Feb 8, 2012 at 7:31 PM, allison <ajp166 at verizon.net> wrote:
There is no SCSI support in VMS at all.
What was the KZQSA for? I got one and thought I had scored a qbus scsi
adapter and quickly determined it was pretty much useless. What I
never understood for sure was if it lacked hardware documentation and
drivers and was not supported for general SCSI use or if it was
broken/crippled in some way so that it couldn't be used for general
SCSI.
-chd
8:25 p.m.
On 02/08/2012 08:18 PM, Charles Dickman wrote:
On Wed, Feb 8, 2012 at 7:31 PM, allison<ajp166 at
verizon.net> wrote:
That's a driver issue and was some sort of "business decision".
(means "got screwed up by suits for no good reason")
And it is an MSCP controller, is it not?
-Dave
--
Dave McGuire
New Kensington, PA
There is no SCSI support in VMS at all.
What was the KZQSA for? I got one and thought I had scored a qbus scsi
adapter and quickly determined it was pretty much useless. What I
never understood for sure was if it lacked hardware documentation and
drivers and was not supported for general SCSI use or if it was
broken/crippled in some way so that it couldn't be used for general
SCSI.
8:54 p.m.
On Wed, Feb 8, 2012 at 5:25 PM, Dave McGuire <mcguire at neurotica.com> wrote:
?And it is an MSCP controller, is it not?
An M5976 KZQSA is not an MSCP controller.
An M5977 RQZX1 is supposed to be an MSCP and TMSCP controller. I
don't have any first hand knowledge of a RQZX1, just the KZQSA.
9:05 p.m.
On 02/08/2012 08:54 PM, Glen Slick wrote:
On Wed, Feb 8, 2012 at 5:25 PM, Dave
McGuire<mcguire at neurotica.com> wrote:
Ahh, the RQZX1 is what I was thinking of. So it's still just a
driver issue with the KZQSA, yes?
-Dave
--
Dave McGuire
New Kensington, PA
And it is an MSCP controller, is it not?
An M5976 KZQSA is not an MSCP controller.
An M5977 RQZX1 is supposed to be an MSCP and TMSCP controller. I
don't have any first hand knowledge of a RQZX1, just the KZQSA. 
9 Feb
9 Feb
1:57 a.m.
New subject: KZQSA, was : Re: VAX for running old version of VMS
Zitat von Dave McGuire <mcguire at neurotica.com>:
Ahh, the RQZX1 is what I was thinking of. So
it's still just a
driver issue with the KZQSA, yes?
All from memory, so ...
We had this discussion years ago already. AFAIRC, the KZQSA doesn't
have any boot EPROM, so it has to use the build in monitor program of
the machine it is attached to. And only the 4000 series had the
commands build in, to BOOT from a
KZQSA. After you boot, you can use the KZQSA on other machines too, if
your OS iinstalled the drivers for it.
8 Feb
8 Feb
8:31 p.m.
On Feb 8, 2012 5:22 PM, "Charles Dickman" <chd at chdickman.com> wrote:
What was the KZQSA for? I got one and thought I had scored a qbus scsi
adapter and quickly determined it was pretty much useless. What I
never understood for sure was if it lacked hardware documentation and
drivers and was not supported for general SCSI use or if it was
broken/crippled in some way so that it couldn't be used for general
SCSI.
It's great for booting the VMS install media from a SCSI CD-ROM. That's
all I've ever done with one and maybe all they are practically good for.
-Glen
9:13 p.m.
On 02/08/2012 08:58 PM, Eric Smith wrote:
allison wrote:
Which releases and where can you find a KZQSA. All kidding that was a V6
or later controller correct? I'm certain V5.4 didn't.
I could also add the VS2000 as that had the NCR5380 scsi controller but
the scsi driver
was totally broken (tape only and that was a unique version of the TK50).
Allison
There is no SCSI support in VMS at all.
Some releases of VMS support the KZQSA, which is a native SCSI
controller (no MSCP or other emulation). Doesn't that mean that there
is some SCSI support in VMS?
9:44 p.m.
On 02/08/2012 09:13 PM, allison wrote:
I could also add the VS2000 as that had the NCR5380
scsi controller but
the scsi driver
was totally broken (tape only and that was a unique version of the TK50).
There's an echo in here. ;) What was broken was the driver; one can
run NetBSD just fine from that controller.
-Dave
--
Dave McGuire
New Kensington, PA
9 Feb
9 Feb
5:15 a.m.
allison wrote:
Which releases and where can you find a KZQSA. All
kidding that was a V6
or later controller correct? I'm certain V5.4 didn't.
The VMS 5.5-2 SPD
says that the KZQSA is supported. I don't have any
idea whether that was the earliest release that supported it.
8 Feb
8 Feb
9:39 p.m.
On 02/08/2012 08:58 PM, Eric Smith wrote:
allison wrote:
Ahh, and what about the NCR5380 in the MicroVAX/VAXstation-2000?
There's no other processor on that board to run an MSCP emulation layer,
at least not one that I recall. It's only officially supported for
tapes, but people have rewritten EPROMs for those machines such that
they'll boot from a SCSI disk attached to that interface. I don't
recall what device type they show up as, though. (which is sad, because
I used a VAXstation-2000 every day for years, eons ago!)
-Dave
--
Dave McGuire
New Kensington, PA
There is no SCSI support in VMS at all.
Some releases of VMS support the KZQSA, which is a native SCSI
controller (no MSCP or other emulation). Doesn't that mean that there is
some SCSI support in VMS?
10:12 p.m.
allison wrote:
This is kind of strange. I didn't see that response from allison, and given he had a
conflicting response to my original post...IS there a pending message that didn't make
it to the entire list? I have seen that sometimes I get the response to a message before
the actual message itself. Very strange.
There is no SCSI support in VMS at all.
Some releases of VMS support the KZQSA, which is a native SCSI
controller (no MSCP or other emulation). Doesn't that mean that there
is some SCSI support in VMS?
11:08 a.m.
On 02/08/2012 10:23 AM, Ethan Dicks wrote:
The uVAX-I has a dual-board CPU that must be installed
in slots 1-2 of
a backplane with CD slots, and the memory sits _on_ the Qbus, so with
22 address bits, you top out at 4MB. The later MicroVAXen fit on a
single board and use PMI memory (different boards, matched to the type
of CPU). I don't know the MicroVAX III memory max off the top of my
head (after my time), but the uVAX-II has 1MB on-board and can take,
IIRC, 12MB in the first two slots.
The MicroVAX-II maxes out at 16MB. The onboard memory adds to
whatever you put in the second and third slots, but if you put two 8MB
boards (or a third-party 16MB board, DEC didn't make one but I think
Chrislin did, and perhaps Dataram) the onboard 1MB is disabled. That's
why we see MicroVAX-IIs with odd amounts of memory like 5MB, 9MB, 13MB,
but maxed-out ones have 16MB.
-Dave
--
Dave McGuire
New Kensington, PA
7:50 p.m.
On 02/08/2012 10:23 AM, Ethan Dicks wrote:
On Tue, Feb 7, 2012 at 2:14 PM, Kevin
Reynolds<tpresence at hotmail.com> wrote:
The DLV11J
is supported line an DL. But not as console as CPU has one.
I believe the backplane was different.for the uVAX1, and different
The cabinet for the microvax I is different that
the BA123 and the BA23 that are available for the later vaxes I believe.
Nope.
Bog-standard BA23 or BA123. There are later cabinets for
MicroVAX IIIs, but they are models like the BA213. You can take a
MicroPDP-11/23 and re-stuff it as a MicroVAX-I or a MicroVAX-II,
though you will probably need different peripherals (I don't think the
DLV11/DLV1J is supported under VMS, for example, though the RLV12 is). from the 11/23 and uVAXII in the CD slot usage. though
the backplane
is swappable.
Generally a uVAXII in a BA23 is cramped, just not enough slots or
places to stick disk drives. If your doing that use the densest memory
(8mb boards or 16mb is possible), IO(DHV11) and a RQDX3. Some
boards like the RQDX2 are quad width and eat more space than RQDX3
a dual width card. But power may be a factor if care is not taken.
The uVAX-I has a dual-board CPU that must be installed
in slots 1-2 of
a backplane with CD slots, and the memory sits _on_ the Qbus, so with
22 address bits, you top out at 4MB. The later MicroVAXen fit on a
single board and use PMI memory (different boards, matched to the type
of CPU). I don't know the MicroVAX III memory max off the top of my
head (after my time), but the uVAX-II has 1MB on-board and can take,
IIRC, 12MB in the first two slots.
the KA630 was 16mb, KA650 and 655 maxed at 64mb .
Besides memory, the big limitation I remember from
those days about
upgrades was disk size. The uVAX-I shipped with the RQDX1 which did
not support disks over 30MB, I think. (RD52 but not RD53, IIRC). The
RQDX3 supports DEC disks up to the RD54 (154MB) and non-DEC disks,
though it can be tricky to get the Field Service formatter which can
label non-standard disks. Even so, MFM disks top out pretty low. I
think I managed to fit VMS 6.0 on an RD54 on a 9MB uVAX-II, but I had
to leave something behind (like the HELP files). Using an MSCP SCSI
controller bypasses this dance and leaves memory as the critical
resource.
Yep, memory is the critical resource to a point, 9mb will run a lot of
the later
versions but make sure you have swap space on the disk.
If your creative, and the box is a BA123 you can use multiple RD54s
(up to 4 drives are supported) and space is less an issue. Other ways
is a RD54
and a mix of smaller drives. RD52 (quantum D540) is a good drive for
swapping
as it's fast despite being only 31MB.
Using a SCSI controller is the way out but watch for dives over 1GB for
the boot drive.
I forget if that affected only the older 3100s or also the uVAX-II boot
as well.
Allison
8:29 p.m.
On 02/08/2012 07:50 PM, allison wrote:
I believe the backplane was different.for the uVAX1,
and different
from the 11/23 and uVAXII in the CD slot usage. though the backplane
is swappable.
I'm not aware of any different backplane for the MicroVAX-I. I've
certainly run several MicroVAX-Is in different, formerly PDP-11 Qbus
chassis, like the BA11-SA for example, and several non-MicroVAX-I BA23s,
and my "woodie boxes" (internal DEC prototypes of the BA123 that were
built with mostly BA23 parts including a standard BA23 backplane).
Using a SCSI controller is the way out but watch for
dives over 1GB for
the boot drive.
I forget if that affected only the older 3100s or also the uVAX-II boot
as well.
It's not an issue on the MicroVAX-II because it'll only end up
speaking MSCP to the controller, and never knows there's SCSI on the
other end. It's an issue for early integrated desktop VAXstations
because the boot ROM only uses so many bits (I forget how many just now)
to address a logical block.
-Dave
--
Dave McGuire
New Kensington, PA
10:04 p.m.
On Feb 8, 2012, at 8:29 PM, Dave McGuire wrote:
I think it's actually a limitation of early SCSI commands; the 6-byte READ command has
an LBA of 21 bits (1GB if you're talking 512-byte blocks). The 10-byte READ commands
have 32-bit LBAs, which will still only do 2TB (which ought to be enough for anyone,
right?).
- Dave
Using a SCSI
controller is the way out but watch for dives over 1GB for
the boot drive.
I forget if that affected only the older 3100s or also the uVAX-II boot
as well.
It's not an issue on the MicroVAX-II because it'll only end up speaking MSCP to
the controller, and never knows there's SCSI on the other end. It's an issue for
early integrated desktop VAXstations because the boot ROM only uses so many bits (I forget
how many just now) to address a logical block.
10:18 p.m.
On 02/08/2012 10:04 PM, David Riley wrote:
On Feb 8, 2012, at 8:29 PM, Dave McGuire wrote:
I think it's actually a limitation of early SCSI commands; the 6-byte
READ command has an LBA of 21 bits (1GB if you're talking 512-byte
blocks). The 10-byte READ commands have 32-bit LBAs, which will
still only do 2TB (which ought to be enough for anyone, right?).
That's what I said...kudos for remembering the number of LBA bits,
which I forgot. This is handled in the boot ROM, though...once the
system is booted, larger-capacity drives aren't a problem.
-Dave
--
Dave McGuire
New Kensington, PA
Using a
SCSI controller is the way out but watch for dives over
1GB for the boot drive. I forget if that affected only the older
3100s or also the uVAX-II boot as well.
It's not an issue on the MicroVAX-II because it'll only end up
speaking MSCP to the controller, and never knows there's SCSI on
the other end. It's an issue for early integrated desktop
VAXstations because the boot ROM only uses so many bits (I forget
how many just now) to address a logical block.
10:31 p.m.
On Feb 8, 2012, at 10:18 PM, Dave McGuire wrote:
Well, I'd forgotten it was 5380, which just passes the commands straight through;
I've been wading through 53C90[A] datasheets long enough that I forgot it existed; the
53C90 handles a lot of the commands autonomously.
- Dave
I think
it's actually a limitation of early SCSI commands; the 6-byte
READ command has an LBA of 21 bits (1GB if you're talking 512-byte
blocks). The 10-byte READ commands have 32-bit LBAs, which will
still only do 2TB (which ought to be enough for anyone, right?).
That's what I said...kudos for remembering the number of LBA bits, which I forgot.
This is handled in the boot ROM, though...once the system is booted, larger-capacity
drives aren't a problem.
9 Feb
9 Feb
12:11 a.m.
On 02/07/2012 02:14 PM, Kevin Reynolds wrote:
The cabinet for the microvax I is different that the
BA123 and the
BA23 that are available for the later vaxes I believe. This model
has a chassis which is 6"x28"x22",
Can you take a picture of this box please? It sounds like a BA23 to
me, and pretty much every MicroVAX-I shipped in bog-standard BA23s.
and a KD23 CPU (probably faulty).
KD32.
It has a maximum RAM capability of 4MB. It definitely
cannot run
netbsd later than 1.61, maybe nothing before that, but I haven't
validated...
Are you certain that it can run NetBSD *at all*? To my knowledge,
the somewhat unusual memory addressing scheme of the KD32 (memory
directly on the Qbus, unique for a VAX) would require quite a bit of
specific kernel support in NetBSD, and I don't think anyone ever did
that. It's possible that it happened and I missed it, but I doubt it;
port-vax is a pretty low-traffic list and I've been on it since like
1993. (when the ONLY VAX it ran on was the 11/750)
-Dave
--
Dave McGuire
New Kensington, PA
8:52 a.m.
Are you certain that it can run NetBSD *at all*? To my
knowledge,
the somewhat unusual memory addressing scheme of the KD32 (memory
directly on the Qbus, unique for a VAX) would require quite a bit of
specific kernel support in NetBSD, and I don't think anyone ever did
that. It's possible that it happened and I missed it, but I doubt it;
port-vax is a pretty low-traffic list and I've been on it since like
1993. (when the ONLY VAX it ran on was the 11/750)
-Dave
I can't say it ran any version of netbsd at all. ?I know that after 1.6.1?I had tried
it and only got kernel panics. ?After a time with no solution?I gave up.
Kevin
7 Feb
7 Feb
7:10 p.m.
On 02/07/2012 12:16 PM, Kevin Reynolds wrote:
I have a license for VMS 5.2 and was considering
buying a working system to run it on. My microvax I finally gave up the ghost, and I
would rather spend time learning than going through repair at the moment.I have many of
the parts below from the uvax 1, but who knows whats working and whats not, I need to
test.
MicroVAX-1 is a oldest Qbus VAX, very limited and slow. You can hope to
use maybe 8mb of ram, if you can find it.
Also the backplane is small so getting every thing is is hard. The box
is basic BA23 but the backplan is unique to
uVAX-1. I think the last supporing version was V5 but not later than 5.3.
I have a good configuration for the MicroVAX II to
build on, but I was looking at the MicroVAX 3900 since the price disparity is not too
significant.
The performance of the 3900 is much greater. The UvaxII with KA630 is
the first chip VAX and the KA650 is faster but uses different memory.
Unfortunately, I don't have a good hardware build
plan for the 3900. I am hoping I can get some good guidance here.
My largest concern is that the hardware is supported by the OS more than anything else.I
want to stick with QBUS.I need to have more than 4 async serial ports with modem control
(so I can hook it up to other systems) considering the DHV11I need ethernet,
DHV11 is not an issue for version though 7.3 and likely later.
thinking DEQNA
DEQNA is supported through V5.4 but was being phased out then. Due t
isues with DEQNA finding a working one may be problematic.
DELQA is preferred for reliability and generally works backward and
forward.
Im considering the TK50 tape (I would go tk70 if I
knew it was backward compat, I'm just not sure)I dont know which disk controllers to
get.
The turning point for tk70 is V5.2, or older than V5 tk50, newer either
will work.. However TK50 is usable though 7.3.
I'm guessing it should be in the KDA50 with
RA82 disks, but I have no idea, just guessing.The standard RX50 floppy.Looking at the
VCB02 for video, just as a nicety.And then 64Mb ram.
Forget 64MB ram it won't
fit in a Qbus unless a very late version 3rd
part board. Expect 16mb.
RA82 is big as in a 40intall rack minimum, you will need the matching
controller. RX50 is norm though with late model
Nominal Disk For microVAX is RD54 (several if BA123) or SCSI (RZxx
series in 5.25 or 2.5" formats). For that you need
a SCSI controller.
RQDX3 you can also use RX33 (teac fd55gfr) as its quieter and supports
both RX50 format and RX33 (higher density).
The RQDXn series controller support floppy and hard disk (MFM).
Versions up to VMS V5.4 will fit on RD53 (71mb) though very cramped,
RD54 is better at 150mb. Ideal is a 1gb but that
is SCSI or SMD interfaces.
If you go the VCB make sure you get the harness, pannel insert and the
corresponding monitor. I'd go with an
VT1200 or VT2000 as you cna support serial line (console) and Xterm.
I have never dealt with the 3900 hardware at all, and
I have the hobby funds to procure something that works.Guessing that I can do it for less
than $2k even if I had to buy all "new" parts. Obviously I want to limit cost
as much as possible, but I would also like to be able to run netbsd on the hardware at a
later date if it was interesting to do so.
Check on what NetBSD runs on and build that as last I remember they were
still challenged on what mass storage devices
and video were supported.
I understand that the 3900 can be converted to a
microvax III+ with some upgrades, but will it run VMS 5.2?
Any recommendations?Kevin
NO, a MicrovaxII (KA630 and ram) can be upgraded to a MicrovaxIII, the
3900 is a microVAX-III (KA655 cpu and ram)
There are three Qbus uVAX cpus KA630(uV-II), KA650(3600) and
KA655(3900), they must have the corresponding
memory as the memory for the older does not work. The memories have
different PMI (over the top bus) and cables
plus the later cpus can address more and need faster ram.
The uVIII will run V5 or later, a uVAXII will as well.
No matter what you want the BA123 or the S-box as you need space for
memory, disk, IO, Network, and video.
A good box to run V5.2 is any of the 3100 series and the
microVAX3100/M76GPX is a good find as it's faster than
the uVAXII and smaller/quieter and lower heat output than most Qbus.
The GPX had video and serial console.
The 3100 series uses SCSI-1 disks so a 1.07GB Baracuda is a great drive
for it. The only caveat on disks is the
boot drive must be under 1.07GB. The basic BA42 box will hole 2 or
three 3.5" format SCSI drives or 2 and a
tape. They have external SCSI bus so an external StorageWorks box can
add more drives. TZK50F or the
SCSI version of the TK70 also can be external. All 3100s support at
least several serial lines and some 8 or 16 ,
VMS running LAT can have network terminals so serial ports on the box
are not a must have. All 3100s support
10B2 (coax Ethernet) so you will need a media converter to go from coax
to RJ45 (10bt, twisted pair).
I have:
3 VAXstation-2000s (VMS5.4 on rd54), (Ultrix4.2 on RD53) and one as a
general use formatter.
2 uVAX-IIs the BA23 version is base microVAXII with 12mb and VMX5.4 on
it, the other is
uVAXII/GPX in BA123 (16mb, 2RD52, 2RZ56 and a RX50, DHV11, TK50, 3plane
color video,
DELQA, ) I inherited it from DEC when I left during the great bleed.
2 uVAX3100/M76GPX 32MB ram and multiple disks plus serial port option
(like DHV11).
6 uVAX3100s basic all have 24or more MB ram, and SCSI disks in them.
4 Storage works boxes with RZ56s in them (2 per).
The 3100s and the big uVII can boot as LAVC (Ethernet connected
cluster) as VMS V5.4.
(thats about 1500W of VAXen running and heating)
All of those systems with the exception of the uVAX2000s run V4 though
current. The uVAX2000s
are so physically small that getting enough disk for later than V5.4 is
hard but, they can netboot
later versions and use a smaller (31mb) hard disk for swap space.
The real question is 3100 pizza box (4"high by 18wide by 16deep) or
BA123, Sbox or larger.
Anything in a BA23 is way to cramped for boards, power and only two
possible slots for disks.
The BA123 is fairly large (end table size) but has slots and power that
can take up to 5 5.25" sized
hard disk, floppy, tape. The backplane is large enough and has the
cooling needed for cpu, ram
and IO like DHV11. The disks can be RQDX3 (rx50 or 33 floppy and RD5n
MFM) or find a CMD
SCSI controller and use that with RZ5x, or possibly RZ2x 3.5" format
disks. Networking go with
DELQA and a AUI to twisted pair converter.
I find that one VAX with tape is enough as tape is slow and its
accessible through the network
if needed. Same for other resources. I rely on a VT320 for serial
console or VT1200 for both serial
or X-term (DECTerm).
As to IO, find a DEC terminal server as LAT or RLAT allows terminals to
be remote and still be
like they were on a DHV11 (in some cases better), same for serial
interface printers.
It's possible to use a PC as an Xterm though I've not done that. I have
use a PC running PathWorks
to connect via network to the VAXen as a terminal.
Allison
7:36 p.m.
On 02/07/2012 07:10 PM, allison wrote:
4MB max on MicroVAX-I, the memory is straight on the Qbus, no PMI.
-Dave
--
Dave McGuire
New Kensington, PA
I have a
license for VMS 5.2 and was considering buying a working
system to run it on. My microvax I finally gave up the ghost, and I
would rather spend time learning than going through repair at the
moment.I have many of the parts below from the uvax 1, but who knows
whats working and whats not, I need to test.
MicroVAX-1 is a oldest Qbus VAX, very limited and slow. You can hope to
use maybe 8mb of ram, if you can find it. 
7:42 p.m.
There are three Qbus uVAX cpus KA630(uV-II),
KA650(3600) and
KA655(3900), they must have the corresponding memory [...]
There's also the KA620, which is the same as the KA630 as far as memory
interfaces go. However, it won't run VMS (indeed, that's sort of the
point of it).
Back in the day, the KA620 was hard to find. I daresay these days it's
even harder to find, though whether that's just because all MicroVAX
hardware has gotten rarer or not I'm not competent to say.
Fortunately, there's little reason to want one unless you're a
completist...or have some OS software that expects a KA620 but can't
tolerate a KA630 (I don't know of any such software, but it may exist.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
8:11 p.m.
On Tue, Feb 7, 2012 at 4:42 PM, Mouse <mouse at rodents-montreal.org> wrote:
There's also the KA620, which is the same as the KA630 as far as memory
interfaces go. ?However, it won't run VMS (indeed, that's sort of the
point of it).
Back when Boeing Surplus was still open in the Seattle area there were
a couple of rtVAX BA213 boxes (I believe they were KA620s) in there
one day for something like $50. I didn't know much about VAX systems
at the time and didn't grab them when I had the chance but wish I did
now. If you want to do anything with Q-bus systems the BA213 boxes
are nice to work with, but not so great if you have to deal with
shipping them.
-Glen
7:51 p.m.
Kevin;
I have a MicroVax 3400 and MicroVax 4000-300 that I want to get rid
of. I'm in the Washington DC area. Any interest?
Doug
At 12:16 PM 2/7/2012, you wrote:
I have a license for VMS 5.2 and was considering buying
a working
system to run it on. My microvax I finally gave up the ghost, and I
would rather spend time learning than going through repair at the
moment.I have many of the parts below from the uvax 1, but who knows
whats working and whats not, I need to test.
I have a good configuration for the MicroVAX II to build on, but I
was looking at the MicroVAX 3900 since the price disparity is not
too significant.
Unfortunately, I don't have a good hardware build plan for the
3900. I am hoping I can get some good guidance here.
My largest concern is that the hardware is supported by the OS more
than anything else.I want to stick with QBUS.I need to have more
than 4 async serial ports with modem control (so I can hook it up to
other systems) considering the DHV11I need ethernet, thinking
DEQNAIm considering the TK50 tape (I would go tk70 if I knew it was
backward compat, I'm just not sure)I dont know which disk
controllers to get. I'm guessing it should be in the KDA50 with
RA82 disks, but I have no idea, just guessing.The standard RX50
floppy.Looking at the VCB02 for video, just as a nicety.And then 64Mb ram.
I have never dealt with the 3900 hardware at all, and I have the
hobby funds to procure something that works.Guessing that I can do
it for less than $2k even if I had to buy all "new"
parts. Obviously I want to limit cost as much as possible, but I
would also like to be able to run netbsd on the hardware at a later
date if it was interesting to do so.
I understand that the 3900 can be converted to a microvax III+ with
some upgrades, but will it run VMS 5.2?
Any recommendations?Kevin
10:39 p.m.
On 2/7/12 4:51 PM, "Douglas Taylor" <dj.taylor4 at verizon.net> wrote:
Kevin;
I have a MicroVax 3400 and MicroVax 4000-300 that I want to get rid
of. I'm in the Washington DC area. Any interest?
Doug
The 4000/300 is a really nice machine - I have one, running 7.3 - but I'll
warn you, it's heavy. -- Ian
4581
days inactive
4583
days old
39 comments
14 participants
participants (14)
-
ajp166＠verizon.net -
chd＠chdickman.com -
dj.taylor4＠verizon.net -
emu＠e-bbes.com -
eric＠brouhaha.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
glen.slick＠gmail.com -
IanK＠vulcan.com -
mc68010＠gmail.com -
mcguire＠neurotica.com -
mouse＠Rodents-Montreal.ORG -
pete＠dunnington.plus.com -
tpresence＠hotmail.com