13 Dec
2006
13 Dec
'06
2:18 p.m.
I don't know anything about HP minis, so I don't know what to look for,
but for those on the list that have HPs and might be interested...
The main unit: item # 320060663740
The FP unit: item # 320060663048
No bids yet.
J
13 Dec
13 Dec
3:18 p.m.
New subject: HP HP-2117F on ebay
J Blaser wrote....
I don't know anything about HP minis, so I
don't know what to look for, but
for those on the list that have HPs and might be interested...
The main unit: item # 320060663740
The FP unit: item # 320060663048
No bids yet.
And hopefully there won't be... unless you want parts - cause
it's got
issues.
I noticed this system on ebay a night or two ago. I was going to post a
warning here to people about it, but just didn't get a roundtuit.
1) The power supply for the cpu has been pulled (ie. is not present). That
doesn't bode well ;)
2) The DCPC card (that's DMA to you and me) appears to be missing and the
cable is dangling. This was standard on the 2117 (1000F aka 21MX/F) I think.
3) There is NO memory, nor a memory controller in the box, they appear to
have been snagged.
4) Either the FEM board has been pulled, or they just used the wrong cable
to connect the FAB to the cpu board. Given that slot 10 is empty and thus
creates a break in the interrupt chain, I'm guessing the former. That means
you're likely missing microcode. Could just be vendor microcode... or your
base instruction set ;)
5) There appears to be something homebrew on the TBG board in slot 11. At
least it's nothing I've ever seen. Given the RPL markings on the front this
tends to make me think of process control. Maybe the cable was to allow some
other device to generate the time based interrupts.
6) Third party board in slot 12. Who knows.
7) 13037 disk interface in slot 13, that matches the front panel markings as
likely boot device.
8) Note the 8 channel mux card is not compatible with TSB. Note also the
non-standard edge card connection. PSI board?
9) 7970 tape controller set, handy to have.
10) Two microcircuit boards. I happen to REALLY like these. If anyone buys
this sytem, I'd be willing to pay for one or two of these cards :)
11) Missing board in slot 22... there goes your interrupts. Who knows, maybe
this system ran something that was only polled.
12) Don't recognize the board in slot 23.
On the plus side, it IS an F series... the FP unit is auctioned separately
and they are calling it a power supply. Nope, it's a floating point unit.
There IS no power supply - and there sure is supposed to be ;)
I am not familiar enough with the F series to know if that FP unit includes
all the necessary boards and/or connectors.
Random thoughts....
Jay
14 Dec
14 Dec
1:57 a.m.
New subject: HP HP-2117F on ebay
On Wed, 13 Dec 2006, Jay West wrote:
Many issues...
No bids yet.
And hopefully there won't be... unless you want parts - cause it's got
issues. 1) The power supply for the cpu has been pulled (ie.
is not present). That
doesn't bode well ;)
And you can't just fit in a standard PC power supply...
2) The DCPC card (that's DMA to you and me)
appears to be missing and the
cable is dangling. This was standard on the 2117 (1000F aka 21MX/F) I think.
Yes, it was standard and it's missing.
3) There is NO memory, nor a memory controller in the
box, they appear to
have been snagged.
No memory, no memory controller, nothing...
4) Either the FEM board has been pulled, or they just
used the wrong cable to
connect the FAB to the cpu board. Given that slot 10 is empty and thus
creates a break in the interrupt chain, I'm guessing the former. That means
you're likely missing microcode. Could just be vendor microcode... or your
base instruction set ;)
Of the three 1000F I have all have their firmware in slot 10 on a FAB
because the mainboard simply doesn't have enough sockets for all the ROMs.
5) There appears to be something homebrew on the TBG
board in slot 11. At
least it's nothing I've ever seen. Given the RPL markings on the front this
tends to make me think of process control. Maybe the cable was to allow some
other device to generate the time based interrupts.
Doesn't hurt.
8) Note the 8 channel mux card is not compatible with
TSB. Note also the
non-standard edge card connection. PSI board?
All 8 channel muxes have these edge card connections.
9) 7970 tape controller set, handy to have.
Not sure whether it's the controller for the 7970B or 7970E.
10) Two microcircuit boards. I happen to REALLY like
these. If anyone buys
this sytem, I'd be willing to pay for one or two of these cards :)
These are simple digital I/O boards. But thanks to DCPC any board supports
DMA :-)
12) Don't recognize the board in slot 23.
Same as in slot 13.
On the plus side, it IS an F series... the FP unit is
auctioned separately
Well, it isn't an F series. An F series consists of the proper firmware
ROMs (SIS, VIS, FFP, FPP) and the external FP processor, but everything's
missing. You could plug in the firmware from an E series, and then it's a
1000E.
My machine here also has the RTE6 and SIGNAL/1000 firmware ROMs.
I am not familiar enough with the F series to know if
that FP unit includes
all the necessary boards and/or connectors.
I think it's complete, but the cables are missing (especially the
connection from the FPP to the MPP on the CPU front panel).
Christian
10:40 a.m.
New subject: HP HP-2117F on ebay
Christian wrote....
Doh! I should have noticed that. The led & plugs on the
left are a dead
giveaway. But since that board requires DMA.... and there's nothing in slot
22....
That's not necessarily correct, the machine on ebay may WELL be an F
series,
and likely is. See that shielded ribbon connector on the back left? That's
connecting the cpu to the FAB board underneath (as well as a dangling end
for what was probably a FEM board in slot 10). So I bet ya a lunch that the
F firmware is whats on the FAB underneath.
Again, Im not an expert on the F series. But the 2111F doesn't have an
external FP processor does it? Just the 2117F does? Both are certainly F
series.
Of the three 1000F I have all have their firmware in
slot 10 on a FAB
because the mainboard simply doesn't have enough sockets for all the ROMs.
I
doubt that seriously. I'd like to see how you could put a FAB board in an
I/O slot ;)
On most 21MX/F's the general F-specific microcode was put on a FAB board
which goes under the cpu board (as a daughtercard). The reason was simple -
they didn't want to waste an I/O slot with a FEM. Another reason is because
many systems had hard coded requirements of what boards you put in what I/O
slots, and a FEM board messed things up. When the FAB board ran out of room
then they'd start going to a FEM. When possible, I like to remove the FAB
and only use the FEM as it's easier to get to the roms and change them (and
I change them a lot). Most people would rather use the FAB as firmware was
rarely changed and they valued the I/O slot real-estate.
A wild guess - this system had vendor supplied proprietary microcode on the
FEM that was in slot 10. When the system was decomissioned, the licensed
firmware board (the whole FEM) was yanked. The F microcode is likely on the
FAB underneath.
More to the point... note that a FAB (and FEM) was optional on the E series.
On the F series, the FAB was required but the FEM was optional. Why? See
above.
I wrote...
> 5) There appears to be something homebrew on the
TBG board in slot 11.
Christian replied...
Doesn't hurt.
Well, it sure could if you
need a TBG and don't have one already. What if
the cable was to generate interrupts to the host from an external device,
and the setup was a 'normally closed' configuration? Your processor would
become rather sluggish when interrupts were armed :D
These are simple digital I/O boards. But thanks to
DCPC any board supports
DMA :-)
There is a reason I prefer these "simple boards". My preferred
setup
typically uses six 12566 boards per system. I find myself changing their
configuration a lot. It's been a while, but I seem to recall I had to
generally change the jumper settings any time I wanted to run diags on them
too. The non-microcircuit badged 12566 boards (GRD TRU +/-) are all soldered
jumpers, some of them 3 way posts with multiple jumpers. The microcircuit
badged 12566 boards are generally the later revision that uses jumper plugs.
MUCH easier to work with if you're changing jumpers a lot. I think the later
revisions also offer some additional functionality that the older ones
don't. Something about a "party line" seems to be lurking in foggy
memory...
12) Don't
recognize the board in slot 23.
Same as in slot 13. On the plus
side, it IS an F series... the FP unit is auctioned
separately
Well, it isn't an F series. An F series consists of the proper firmware
ROMs (SIS, VIS, FFP, FPP) and the external FP processor, but everything's
missing. You could plug in the firmware from an E series, and then it's a
1000E. My machine here also has the RTE6 and SIGNAL/1000
firmware ROMs.
I believe I have those... I think I just saw the DS1000 roms a few
days ago
on my workbench. Ya know, there's a problem with just using a FEM board. You
don't have the sockets for the smaller proms there. Only on the FAB. Ah,
that's why I wanted to find blanks for the larger proms - so I could migrate
the IOP firmware to the FEM. WCS would be more fun....
Jay
1:03 p.m.
New subject: HP HP-2117F on ebay
Christian wrote....
> Well, it isn't an F series. An F series
consists of the proper firmware
> ROMs (SIS, VIS, FFP, FPP) and the external FP processor, but everything's
> missing.
Ok, this piqued my curiosity so I went to the manuals and checked.
The F series consists of the 2111F and the 2117F. The 2117F has an external
hardware floating point unit mounted in a rackmount chassis underneath the
cpu chassis. The 2111F also has a hardware floating point unit, but it is
not mounted in a external chassis. It is mounted INSIDE the normal 1000
series chassis. I'd be curious to see just how that was done! There is
obviously no room for the same physical parts. Perhaps it was put in the
memory cage and the cable run to the MPP port? In any case, it must have
been a very different card(s) (although functionally equivalent). Aha... I
just found a picture in the manual that may elucidate this. The 2111 chassis
is just as tall as the 2117 chassis, but it has 5 less I/O & memory slots at
the top. Methinks the space at the top is where a 2111F specific hardware
floating point unit must go.
So, an F series doesn't always have the external FP processor. They all do
have hardware floating point processors as separate systems. Oddly enough,
the 2111F doesn't have DMS as standard while the 2117F does. That accounts
for the difference in instruction set counts.
Jay
15 Dec
15 Dec
2:30 a.m.
New subject: HP HP-2117F on ebay
[
Note to those who don't understand all the terms:
FAB: Firmware Accessory Board
FEM: Firmware Expansion Module
DCPC: Dual Channel Port Controller
TBG: Time Base Generator
WCS: Writable Control Store
SIS: Scientific Instruction Set
VIS: Vector Instruction Set
DMI: Dynamic Mapping Instuctions
FFP: Fast Fortran Processor
FPP: Floating Point Processor
]
On Thu, 14 Dec 2006, Jay West wrote:
I doubt that seriously. I'd like to see how you
could put a FAB board in an
I/O slot ;)
I was wrong, yes. I was thinking of the FEM. It was suggested by HP to use
a FEM board in slot 10 instead of the FAB on the CPU board to easily
access the firmware ROMS.
A wild guess - this system had vendor supplied
proprietary microcode on the
FEM that was in slot 10. When the system was decomissioned, the licensed
firmware board (the whole FEM) was yanked. The F microcode is likely on the
FAB underneath.
Not sure. The Base Set microcode can be either on the CPU board or on the
FEM. My machines have the Base Set on the FEM, as it was recommended ([1]
13-3). But maybe slot 10 contained a WCS board.
More to the point... note that a FAB (and FEM) was
optional on the E series.
On the F series, the FAB was required but the FEM was optional. Why? See
above.
No, the FEM was standard on both E and F series ([1], section 2-1). BTW
the FAB has the lowest priority in the firmware chain, lower than the
firmware on the CPU board.
and likely is. See that shielded ribbon connector on
the back left? That's
connecting the cpu to the FAB board underneath (as well as a dangling end for
what was probably a FEM board in slot 10). So I bet ya a lunch that the F
firmware is whats on the FAB underneath.
Maybe, but maybe it's empty and the ROMs were on the FEM. According to [1]
section 8-3, the standard SIS was installed on the FEM, so no FEM, no SIS.
But no FEM, no DMI/FFP (standard on 1000F, [1] section 7), no VIS,
no DMS, no RTE4/RTE-VI, ... But [2] sections 2-23 to 2-34 are contradicting
[1] ...
I only know what I know, that my machines have their ROMs on the FEM.
Again, Im not an expert on the F series. But the 2111F
doesn't have an
external FP processor does it? Just the 2117F does? Both are certainly F
series.
Only if they have the firmware and FPP that converts an E series to an F
series. If not, it's just the frontpanel with an F instead of an E...
The 2111F has the FPP built internally over the power supply and memory/IO
card cages ([2] figure 1-2).
But anyway, a fully loaded 1000F is a *very* nice machine to play with.
It's a nice number cruncher.
Christian
[1] HP 1000 M/E/F-Series Firmware Installation and Reference Manual
[2] HP 1000 F-Series Computer Installation and Service Manual
8:17 a.m.
New subject: HP HP-2117F on ebay
Christian wrote...
Not sure. The Base Set microcode can be either on the
CPU board or on the
FEM.
It is quite possible that the base set is on the cpu board and the rest is
on the fab, making it "complete". But it is also possible that some of the
required F firmware was on a FEM in slot 10. Or it could have been user or
vendor microcode on a FEM in slot 10. Or some I/O card even, as I've seen
systems with no FEM where they used a 3 connector shielded ribbon cable even
though they only had a FAB, just cause it's the only cable they had
apparently. Of course, it is also possible that there is no base set at all
and someone wrote a custom base set for process control. That's the beauty
of the box, everything was changeable including the base set. All I was
saying is that since there is nothing in slot 10, and there IS a dangling
ribbon cable there, it is likely (although not certain) that some firmware
is missing. I think we can agree that is a reasonable postulation. Whether
the FEM being missing means it is not an F is debatable apparently, but I
was under the impression the base set could be on the cpu board and all the
rest on the FAB. No FEM required.
No, the FEM was standard on both E and F series ([1],
section 2-1).
Apparently HP documentation is at odds then. See the HP 1000 M/E/F
series
computers technical reference handbook, part number 5955-0282, appendix A,
pages A-3 and A-4. It's quite clear there that the FAB was optional on the E
and standard on the F (and ostensibly standard because it was necessary to
house the larger microcode set).
BTW the FAB has the lowest priority in the firmware
chain, lower than the
firmware on the CPU board.
Isn't that priority in the sense of pecking order
for entry points, not in
the sense of any type of execution priority?
Maybe, but maybe it's empty and the ROMs were on
the FEM.
That is certainly possible, hence my comment.
I only know what I know, that my machines have their
ROMs on the FEM.
Ditto for me... that my machines have their roms on the FAB.
But anyway, a fully loaded 1000F is a *very* nice
machine to play with.
It's a nice number cruncher.
I've noticed my E's beat the pants off my
2100A/S machines - the specs say
approximately twice as fast. But I still love the front panel appeal of the
2100 over the 21MX any day :) It's easier to enter programs with and it
certainly has more blinkylights :)
Jay
9:52 a.m.
New subject: HP HP-2117F on ebay
The plot thickens...
A more definitive source than user manuals... is the HP1000 M/E/F-series
computer engineering and reference documentation (92851-90001).
According to the section "HP 1000 E and F-series computer central processing
Unit (CPU) Theory of Operation" on page IA 4-33 "Note that in F-series
computers a customer may elect to have a 12791A FEM board in lieu of the
FAB."
Note the use of the word "customer" and "may". This would certainly
indicate
that a FEM was not necessary in an F. This would suggest that a F was
shipped with FAB and not FEM (which was optional) unless the customer
wanted/needed more space for microcode. This also suggests that a valid F
configuration is FAB only, or FEM only, or FAB & FEM. It also suggests that
it would be fairly common for a user to buy the system with just FAB, and
later (not from factory) add a FEM board when they needed more space to use
in addition to OR in place of, the FAB.
So it would appear our experiences both are correct. It seems the systems
you have acquired were ordered with the FEM instead of FAB *option*. The
systems I have acquired were ordered with the standard part, which got you a
FAB and no FEM. But in any case, the determination from the engineering docs
is, an F most certainly does not have to have a FEM - and a FEM is not
standard/required.
I agree that the user manuals seem to be contradictory on this subject. For
example, the firmware installation manuals do in fact say that SIS can be
installed in a FAB, it says this in the section on the FAB. But in the
section on SIS, it mentions just the FEM I believe.
Jay West
16 Dec
16 Dec
10:15 a.m.
New subject: HP F series
More 21MX F series madness....
I have heard in two places now (one being Christian's post) that the E and F
series mainboards were interchangeable. The HP engineering & reference
documentation indicates that this was not always the case through the entire
product line.
There is clear mention that prior to certain date codes, there were separate
mainboards for the E and F, and they were not interchangeable. After that
date code, the boards were identical and interchangeable between the two.
So, long story short, do not just assume the boards are the same and assume
you can swap them. If they are prior date codes, you'll be disappointed ;)
Now this has me curious... I wonder exactly what the difference between the
E and F mainboards was that made them not compatible prior to the date code
specified? I'll see if I can find further clarification.
Jay
5 Jan
5 Jan
12:45 p.m.
New subject: HP HP-2117F on ebay
On 12/13/06, Jay West <jwest at classiccmp.org> wrote:
J Blaser wrote....
And the winner is... Warren. # 320067174880 # 320067174671
Looks like it was fairly well loaded with a complete set of memory
cards and I/O cards. Hard to tell from the picture but it looks like
it might even have the cable that connects the front panel to the two
cards in the FPP box.
I'm not sure if the FPP box will power up completely without the power
in/out cable between the two boxes. I have never found a currently
available connector shell part that will exactly mate with those 9-pin
power supply connectors on the back. Maybe it doesn't matter if you
power the boxes up in the right order. It's been a while since I
bothered connecting the FPP box to my 2117F.
-Glen
I don't know anything about HP minis, so I
don't know what to look for, but
for those on the list that have HPs and might be interested...
The main unit: item # 320060663740
The FP unit: item # 320060663048
No bids yet.
And hopefully there won't be... unless you want parts - cause
it's got
issues.
2:36 p.m.
New subject: HP HP-2117F on ebay
J Blaser wrote....
> >I don't know anything about HP minis,...
> > The main unit: item # 320060663740
> > The FP unit: item # 320060663048
> > No bids yet.
To which I replied....
> And hopefully there won't be... unless you
want parts - cause it's got
> issues.
And Glen wrote...
And the winner is... Warren. # 320067174880 #
320067174671
Hey, not the same machine I was referencing ;) The one Glen mentions
looked
pretty complete.
Looks like it was fairly well loaded with a complete
set of memory
cards and I/O cards.
I'd yank all those 64K cards and just run with one 256kw
board, keeping the
2nd 256kw board for spare. All those "little" 64k boards just suck up power
and don't give you much extra memory. I generally don't run things that
would require more than a pittance of ram anyways.... so others here will
certainly disagree with me on this point.
I just recognized the "unknown" board at the top in slot 25, it's a PSI
(Programmable Serial Interface) board (built on a 5180-1953). I think that's
a 3rd party board. I would like more info on these boards if anyone has it.
Sketchy info follows that may only be half-accurate: There are four Z80's?
on board as I recall, and 3 proms to dump your z80 code in to. Thus, you can
program this board to do all kinds of fun interesting things - IF you have
the specifications as to what your code needs to do in Z80 to talk to the
backplane or other chips on board. In the given configuration, it was used
to control 4 serial ports I think, but may have been just 2 terminals (the
ones I got were hooked up to a very odd very large graphics terminal that
required 2 serial ports each). I still have a couple of these boards, the
correct cables, and the back panel db25 mounting bracket. I have a manual
for them, but it is the manual for the PSI implementation of the board and
thus no details on coding your own Z80 stuff on it. If anyone has THOSE
specs I'd love a copy.
The other "unknown" board in slot 12, I have no clue, but I'd love to see a
picture of it, from the ebay picture of it what I can see is intriguing.
Warren... I'll gladly trade some of my HP boards in exchange for one or two
of the microcircuit boards in there ;)
I'm not sure if the FPP box will power up
completely without the power
in/out cable between the two boxes. I have never found a currently
available connector shell part that will exactly mate with those 9-pin
power supply connectors on the back. Maybe it doesn't matter if you
power the boxes up in the right order. It's been a while since I
bothered connecting the FPP box to my 2117F.
You know, that's a good question.
Can anyone answer that? Bob S. perhaps? My
hunch given past experience is that the power in/out cable isn't required. I
don't know though. I have looked for those 9 pin power supply connectors
too, to no avail :( I'd love to find some of the new crimp pins at least!!
AMP doesn't seem to remember making them. I would hazard a guess that the
FPP box would need to be powered up before the cpu ;)
Jay West
3:34 p.m.
New subject: HP HP-2117F on ebay
On 1/5/07, Jay West <jwest at classiccmp.org> wrote:
I'd yank all those 64K cards and just run with one
256kw board, keeping the
2nd 256kw board for spare. All those "little" 64k boards just suck up power
and don't give you much extra memory. I generally don't run things that
would require more than a pittance of ram anyways.... so others here will
certainly disagree with me on this point.
Extra RAM is always good for making the self-test count higher on the
blinkenlights. :-)
I just recognized the "unknown" board at the
top in slot 25, it's a PSI
(Programmable Serial Interface) board (built on a 5180-1953). I think that's
a 3rd party board. I would like more info on these boards if anyone has it.
Sketchy info follows that may only be half-accurate: There are four Z80's?
on board as I recall, and 3 proms to dump your z80 code in to.
That sounded similar to what I remembered about the 12792 8-Channel
Mux, but I just looked at the manual and it has a single Z80, DMA,
CTC, 4 SIO/2s, and 16K of DRAM. I wonder if there is anything
interesting you could do with those with your own firmware.
Warren... I'll gladly trade some of my HP boards
in exchange for one or two
of the microcircuit boards in there ;)
I have a few 12566 boards and some relay boards, forget their number.
I have looked for those 9 pin power supply connectors
too, to no avail :( I'd love to find some of the new crimp pins at least!!
AMP doesn't seem to remember making them.
I'd like to make a real plug for the battery backup 820ohm thermistor
bypass. I just have a resistor with its legs folded over wedged into
the connector for now.
-Glen
6:38 p.m.
New subject: HP HP-2117F on ebay
Rather than just 'plug' a resistor into the back of the PSU, I like to get
the
regular cable to the battery backup box, and use that.
Its a common cable, I may have spares.
I pull the lead acid cells out of the battery boxes, and leave the stock
thermistor
in place.
The advantage is that with the battery box in place (or even a blank rear
plate)
the airflow through the I/O cage is much better than when the back plate is
missing.
The same thing is true for the blank plate over the memory card cage. Early
memory controllers will run much hotter without this plate in place.
HP put these parts there for a reason it seems. My I/O slots are packed, so
I try to keep them happy.
Having lots of DMS memory can be a real advantage. I've got software that
gives
you up to 32 separate address spaces, so you can jump between totally
incompatible
programs on a single machine. This allows an IPL program to run HP Basic,
and
then return control back to the original IPL code, etc.
Swapping the system map around is great fun!
----- Original Message -----
From: "Glen Slick" <glen.slick at gmail.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Friday, January 05, 2007 6:34 PM
Subject: Re: HP HP-2117F on ebay
On 1/5/07, Jay West <jwest at classiccmp.org>
wrote:
I'd yank all those 64K cards and just run
with one 256kw board, keeping
the
2nd 256kw board for spare. All those "little" 64k boards just suck up
power
and don't give you much extra memory. I generally don't run things that
would require more than a pittance of ram anyways.... so others here will
certainly disagree with me on this point.
Extra RAM is always good for making the self-test count higher on the
blinkenlights. :-)
I just recognized the "unknown" board
at the top in slot 25, it's a PSI
(Programmable Serial Interface) board (built on a 5180-1953). I think
that's
a 3rd party board. I would like more info on these boards if anyone has
it.
Sketchy info follows that may only be half-accurate: There are four
Z80's?
on board as I recall, and 3 proms to dump your z80 code in to.
That sounded similar to what I remembered about the 12792 8-Channel
Mux, but I just looked at the manual and it has a single Z80, DMA,
CTC, 4 SIO/2s, and 16K of DRAM. I wonder if there is anything
interesting you could do with those with your own firmware.
Warren... I'll gladly trade some of my HP
boards in exchange for one or
two
of the microcircuit boards in there ;)
I have a few 12566 boards and some relay boards, forget their number.
I have looked for those 9 pin power supply
connectors
too, to no avail :( I'd love to find some of the new crimp pins at
least!!
AMP doesn't seem to remember making them.
I'd like to make a real plug for the battery backup 820ohm thermistor
bypass. I just have a resistor with its legs folded over wedged into
the connector for now.
-Glen
6 Jan
6 Jan
3:51 p.m.
New subject: HP HP-2117F on ebay
I had written....
> Warren... I'll gladly trade some of my HP
boards in exchange for one or
> two
> of the microcircuit boards in there ;)
To which Glen replied...
I have a few 12566 boards and some relay boards,
forget their number.
That may or may not help, as "12566" means lots of
different things and
doesn't describe an exact board - GRD TRU, + TRU, Microcircuit Interface,
Printer interface... all slightly different variations of the boards
generically marked 12566. In other words, 12566 is not specific enough, you
need the rest of the letters & numbers after it to know exactly what it is.
Some of the variations are just different wirings (and in some cases I think
a different chip or two) giving different electrical characteristics but the
boards look almost identical to the naked eye. The 12566 boards that I
desire, don't look anything like the "normal" 12566 boards. The ones I want
are labled 12566-60032 and have boatloads of dip jumpers on them, no solder
jumpers.
I have lots of 12566 boards, but they aren't the 12566 boards I want.
I'd like to make a real plug for the battery
backup 820ohm thermistor
bypass. I just have a resistor with its legs folded over wedged into
the connector for now.
Yeah, I know just what you mean. Got a bunch of those
820o's folded just the
way you describe. There is an AMP office close to me. I was wondering if I
took one of my connectors in to them.... This could be a good project for
home plastic molding maybe... but then the problem is finding the right
pins. And most crimp pins for those type of connectors aren't just round
pins you solder on, they have little latches & such on them. Ah well.
Jay
7 Jan
7 Jan
7:40 a.m.
New subject: HP HP-2117F on ebay
On 1/6/07, Jay West <jwest at classiccmp.org> wrote:
That may or may not help, as "12566" means
lots of different things and
doesn't describe an exact board - GRD TRU, + TRU, Microcircuit Interface,
Printer interface... all slightly different variations of the boards
generically marked 12566. In other words, 12566 is not specific enough, you
need the rest of the letters & numbers after it to know exactly what it is.
Some of the variations are just different wirings (and in some cases I think
a different chip or two) giving different electrical characteristics but the
boards look almost identical to the naked eye. The 12566 boards that I
desire, don't look anything like the "normal" 12566 boards. The ones I
want
are labled 12566-60032 and have boatloads of dip jumpers on them, no solder
jumpers.
I didn't realize there was much difference between the boards other
than the obvious GRD TRUE vs +TRUE versions. I checked and I have
some 12566-80024 +TRUE, some 12566-80024 GRD TRUE, but only one
12566-60032 Microcircuit. What makes that version specially suited to
your applications?
I also have some 12551-8001 Relay Out boards and then some 11629-6001
44 BIT GPO boards for which I cannot find any information.
-Glen
2:27 p.m.
New subject: HP HP-2117F on ebay
The 12551 relay board is a great item.
You get 16 isolated contacts, on opposite sides of the PCB.
Write a word to the I/O slot and you close the relay contacts for each bit
set.
There are also slightly complicated signals for the command and flag flip
flops.
I can get you the pin-outs and specifics on these if needed.
----- Original Message -----
From: "Glen Slick" <glen.slick at gmail.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Sunday, January 07, 2007 10:40 AM
Subject: Re: HP HP-2117F on ebay
On 1/6/07, Jay West <jwest at classiccmp.org>
wrote:
That may or may not help, as "12566"
means lots of different things and
doesn't describe an exact board - GRD TRU, + TRU, Microcircuit Interface,
Printer interface... all slightly different variations of the boards
generically marked 12566. In other words, 12566 is not specific enough,
you
need the rest of the letters & numbers after it to know exactly what it
is.
Some of the variations are just different wirings (and in some cases I
think
a different chip or two) giving different electrical characteristics but
the
boards look almost identical to the naked eye. The 12566 boards that I
desire, don't look anything like the "normal" 12566 boards. The ones I
want
are labled 12566-60032 and have boatloads of dip jumpers on them, no
solder
jumpers.
I didn't realize there was much difference between the boards other
than the obvious GRD TRUE vs +TRUE versions. I checked and I have
some 12566-80024 +TRUE, some 12566-80024 GRD TRUE, but only one
12566-60032 Microcircuit. What makes that version specially suited to
your applications?
I also have some 12551-8001 Relay Out boards and then some 11629-6001
44 BIT GPO boards for which I cannot find any information.
-Glen
8 Jan
8 Jan
6:26 a.m.
New subject: HP HP-2117F on ebay
Glen wrote....
I didn't realize there was much difference between
the boards other
than the obvious GRD TRUE vs +TRUE versions. I checked and I have
some 12566-80024 +TRUE, some 12566-80024 GRD TRUE, but only one
12566-60032 Microcircuit.
It gets even more convoluted because some of the complete
part numbers
(12566-XXXXX) refer to the "same" board but later field replacement vs.
original issue. So two boards with different part #'s may be the "same"
board. My CE handbook has some of these new vs. field replacement board
translations, but certainly not all of them plus it changed over time.
You mention GRD true & + true both being 12566-80024... I suspect you're
looking at the pca assy # rather than the part number or vice versa. They
are both on the same spot on the board. These two boards would not have the
same part number I don't think. I THINK from foggy memory that the original
issue boards for grd tru were 12566-6001 and + true were 12566-6002. I'd
have to dig to be sure... maybe the + true version was ordered as 12566
option 001? I don't recall.
What makes that version specially suited to your
applications?
As I said in a previous post... there's a fair number of jumpers
on the
older style 12566 boards, and all those jumpers are solder-type. Several of
those jumpers are four post jumpers (central jumper with 3 option
selections) and the options aren't mutually exclusive. I imagine many people
use a 12566 for a single specific thing, so they set (solder) the jumpers a
certain way and they never need to change the jumpers again. I however tend
to bounce 12566 boards around in lots of different configurations for
different things in different machines on a regular basis. Back when I was
messing with my HP boxes a lot it wasn't uncommon for me to want to change
the jumpers on a 12566 several times a day. It's rather irritating to have
to get out a soldering iron just to try a 3 second "what if I set it this
way" scenario. There's another reason - I don't remember the exact
specifics... but the diagnostics for the 12566 board require certain of the
jumpers to be set a certain way I think. One of my applications (TSB
interconnect kit I think) required those same jumpers a different way. So to
run diags, I had to break out a soldering iron and then put them back to use
the board. Not a huge deal, but irritating. And no, I'd rather not mount a
custom switch to my board unless I absolutely had to. Well, the later model
12566 boards that I prefer (Microcircuit/12566-60032) have plug jumpers
instead of solder jumpers. So I can change jumper settings in a second or
two and put them back quickly a moment later. In addition, I recall there
being some extra features on the -60032 boards that looked like they would
be handy if one was building a custom device...
Jay West
8:59 a.m.
New subject: HP HP-2117F on ebay
In article <006001c73331$02790c50$6500a8c0 at BILLING>,
"Jay West" <jwest at classiccmp.org> writes:
[...] My CE handbook has some of these new vs. field
replacement board
translations, but certainly not all of them plus it changed over time.
Fortunately for you, the set of translations is now complete and fixed
:-).
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://www.xmission.com/~legalize/book/download/index.html>
Legalize Adulthood! <http://blogs.xmission.com/legalize/>
9:17 a.m.
New subject: HP HP-2117F on ebay
I had written....
> [...] My CE handbook has some of these new vs.
field replacement board
> translations, but certainly not all of them plus it changed over time.
To which Richard replied...
Fortunately for you, the set of translations is now
complete and fixed
Yes, it is. But unfortunately for me, my collection of different
vintage CE
handbooks is not complete ;)
Jay West
5 Jan
5 Jan
6:28 p.m.
New subject: HP HP-2117F on ebay
I've got a working 2117F, runs weekly, more or less. Great machine.
Ok, yes, it will power up without the FP box attached, but it will fail the
FPU self-test when the special op-code is run via 'instruction step'.
All F-series machines were ~supposed~ to use fault control memory, I think
this machine is setup that way. Don't pull the small boards until this is
understood.
Memory misconfigurations and missing backup battery bypass plugs are 99%
of the problems people have with these machines. It will fail to pass power
on self
test with misconfigured memory.
----- Original Message -----
From: "Jay West" <jwest at classiccmp.org>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Friday, January 05, 2007 5:36 PM
Subject: Re: HP HP-2117F on ebay
J Blaser wrote....
> >I don't know anything about HP
minis,...
> > The main unit: item # 320060663740
> > The FP unit: item # 320060663048
> > No bids yet.
To which I replied....
> And hopefully there won't be... unless
you want parts - cause it's got
> issues.
And Glen wrote...
And the winner is... Warren. # 320067174880 #
320067174671
Hey, not the same machine I was referencing ;) The one Glen mentions
looked pretty complete.
Looks like it was fairly well loaded with a
complete set of memory
cards and I/O cards.
I'd yank all those 64K cards and just run with one 256kw
board, keeping
the 2nd 256kw board for spare. All those "little" 64k boards just suck up
power and don't give you much extra memory. I generally don't run things
that would require more than a pittance of ram anyways.... so others here
will certainly disagree with me on this point.
I just recognized the "unknown" board at the top in slot 25, it's a PSI
(Programmable Serial Interface) board (built on a 5180-1953). I think
that's a 3rd party board. I would like more info on these boards if anyone
has it. Sketchy info follows that may only be half-accurate: There are
four Z80's? on board as I recall, and 3 proms to dump your z80 code in to.
Thus, you can program this board to do all kinds of fun interesting
things - IF you have the specifications as to what your code needs to do
in Z80 to talk to the backplane or other chips on board. In the given
configuration, it was used to control 4 serial ports I think, but may have
been just 2 terminals (the ones I got were hooked up to a very odd very
large graphics terminal that required 2 serial ports each). I still have a
couple of these boards, the correct cables, and the back panel db25
mounting bracket. I have a manual for them, but it is the manual for the
PSI implementation of the board and thus no details on coding your own Z80
stuff on it. If anyone has THOSE specs I'd love a copy.
The other "unknown" board in slot 12, I have no clue, but I'd love to see
a picture of it, from the ebay picture of it what I can see is intriguing.
Warren... I'll gladly trade some of my HP boards in exchange for one or
two of the microcircuit boards in there ;)
I'm not sure if the FPP box will power up
completely without the power
in/out cable between the two boxes. I have never found a currently
available connector shell part that will exactly mate with those 9-pin
power supply connectors on the back. Maybe it doesn't matter if you
power the boxes up in the right order. It's been a while since I
bothered connecting the FPP box to my 2117F.
You know, that's a good question.
Can anyone answer that? Bob S. perhaps?
My hunch given past experience is that the power in/out cable isn't
required. I don't know though. I have looked for those 9 pin power supply
connectors too, to no avail :( I'd love to find some of the new crimp pins
at least!! AMP doesn't seem to remember making them. I would hazard a
guess that the FPP box would need to be powered up before the cpu ;)
Jay West
8 p.m.
New subject: HP HP-2117F on ebay
Bob wrote....
All F-series machines were ~supposed~ to use fault
control memory, I think
this machine is setup that way. Don't pull the small boards until this is
understood.
Hummm doesn't appear to be any fault control array stuff in there to me.
Jay
8:04 p.m.
New subject: HP HP-2117F on ebay
Bob wrote...
Ok, yes, it will power up without the FP box attached,
but it will fail
the
FPU self-test when the special op-code is run via 'instruction step'.
But that's not quite what I was wondering...
What if the FP box *IS* attached, but the power interconnect (the power
cable between cpu and ffp) cable isn't? This means the ribbon cable is
connected, and the fpu is powered on, and then the cpu is powered on. Will
this work... or is there some mechanism on the power side of the FPP unit
that it will not come on unless the power interconnect cable is on too? I
don't know how this works.
Jay
6328
days inactive
6354
days old
21 comments
6 participants
participants (6)
-
bshannon＠tiac.net -
cc＠corti-net.de -
glen.slick＠gmail.com -
jwest＠classiccmp.org -
legalize＠xmission.com -
oldcpu＠rogerwilco.org