28 Oct
2015
28 Oct
'15
9:43 a.m.
From: Ben Sinclair
According to the Microcomputer Interfaces Handbook, my
H9273 backplane
should be all ABCD.
Yup, that was one of the first things I checked - whilst realizing that if
it wasn't, it was too late... :-)
Am I correct in that I shouldn't have any problem
using the RLV11 in
that backplane?
Not that I am aware of.
Noel
28 Oct
28 Oct
9:46 a.m.
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu> wrote:
--
Ben Sinclair
ben at bensinclair.com
From: Ben
Sinclair
According to the Microcomputer Interfaces
Handbook, my H9273 backplane
should be all ABCD.
Yup, that was one of the first things I checked - whilst realizing that if
it wasn't, it was too late... :-)
Am I correct in that I shouldn't have any
problem using the RLV11 in
that backplane?
Not that I am aware of.
Noel
10:25 a.m.
I just tried to read 774400 from ODT, and it just gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com> wrote:
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu> wrote:
--
Ben Sinclair
ben at bensinclair.com
--
Ben Sinclair
ben at bensinclair.com
From: Ben
Sinclair
According to the Microcomputer Interfaces
Handbook, my H9273 backplane
should be all ABCD.
Yup, that was one of the first things I checked - whilst realizing that if
it wasn't, it was too late... :-)
Am I correct in that I shouldn't have any
problem using the RLV11 in
that backplane?
Not that I am aware of.
Noel 
10:29 a.m.
On 2015-10-28 18:25, Ben Sinclair wrote:
I just tried to read 774400 from ODT, and it just
gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu> wrote:
--
Ben Sinclair
ben at bensinclair.com
From: Ben
Sinclair
According to the Microcomputer Interfaces
Handbook, my H9273 backplane
should be all ABCD.
Yup, that was one of the first things I checked - whilst realizing that if
it wasn't, it was too late... :-)
Am I correct in that I shouldn't have any
problem using the RLV11 in
that backplane?
Not that I am aware of.
Noel
12:35 p.m.
For 18 bit addresses, I assume I would look at 74400 instead? I get
177777 there.
I did just try moving the RLV11 boards below everything else (except
the BDV11), with the same results as before.
On Wed, Oct 28, 2015 at 12:29 PM, Johnny Billquist <bqt at update.uu.se> wrote:
On 2015-10-28 18:25, Ben Sinclair wrote:
--
Ben Sinclair
ben at bensinclair.com
I just tried to read 774400 from ODT, and it just gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com>
wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu>
wrote:
> From: Ben Sinclair
> According to the Microcomputer Interfaces Handbook, my H9273
backplane
> should be all ABCD.
Yup, that was one of the first things I checked - whilst realizing that
if
it wasn't, it was too late... :-)
> Am I correct in that I shouldn't have any problem using the RLV11
in
> that backplane?
Not that I am aware of.
Noel
--
Ben Sinclair
ben at bensinclair.com
1:06 p.m.
Okay, I think I need to be looking at 774400, which doesn't return
anything. Also, the PDP11GUI has a handy RL11 dialog to show all of
the registers, which I think shows the same thing... It's not there!
I've double checked my switch settings, and tested the switches
themselves for continuity just to make sure they weren't bad (which
has happened to me before). Beyond something physically wrong with the
boards, is there anything else I might check?
On Wed, Oct 28, 2015 at 2:35 PM, Ben Sinclair <ben at bensinclair.com> wrote:
For 18 bit addresses, I assume I would look at 74400
instead? I get
177777 there.
I did just try moving the RLV11 boards below everything else (except
the BDV11), with the same results as before.
On Wed, Oct 28, 2015 at 12:29 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Ben Sinclair
ben at bensinclair.com
On 2015-10-28 18:25, Ben Sinclair wrote:
--
Ben Sinclair
ben at bensinclair.com
I just tried to read 774400 from ODT, and it just gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com>
wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu>
wrote:
>
> > From: Ben Sinclair
>
> > According to the Microcomputer Interfaces Handbook, my H9273
> backplane
> > should be all ABCD.
>
> Yup, that was one of the first things I checked - whilst realizing that
> if
> it wasn't, it was too late... :-)
>
> > Am I correct in that I shouldn't have any problem using the RLV11
> in
> > that backplane?
>
> Not that I am aware of.
>
> Noel
--
Ben Sinclair
ben at bensinclair.com
1:40 p.m.
Well I just figured out something important...The switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
DR>START
CHANGE HW (L) ? Y
# UNITS (D) ? 1
UNIT 0
11/23 PROCESSOR (L) Y ?
BUS ADDRESS (O) 174400 ?
VECTOR (O) 160 ? 330
DRIVE (O) 0 ?
BR LEVEL (O) 5 ?
CHANGE SW (L) ? Y
DROP ON ERROR LIMIT (L) N ?
AUTOSIZE (L) N ?
CVRLA DVC FTL ERR 00113 ON UNIT 00 TST 009 SUB 000 PC: 020052
BUS RESET DID NOT CLEAR RLCS
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00004 ON UNIT 00 TST 012 SUB 000 PC: 020264
RLCS READ/WRITE ERROR (BIT 0 DON'T CARE)
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00007 ON UNIT 00 TST 015 SUB 000 PC: 020600
BIT SET INSTRUCTION ON RLCS YIELDED WRONG RESULT
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00008 ON UNIT 00 TST 016 SUB 000 PC: 020724
BIT CLEAR INSTRUCTION ON RLCS YIELDED WRONG RESULT
CONTROLLER: 174400 DRIVE: 0
EXP'D: 141776 REC'D: 145776
... And many more!
On Wed, Oct 28, 2015 at 3:06 PM, Ben Sinclair <ben at bensinclair.com> wrote:
Okay, I think I need to be looking at 774400, which
doesn't return
anything. Also, the PDP11GUI has a handy RL11 dialog to show all of
the registers, which I think shows the same thing... It's not there!
I've double checked my switch settings, and tested the switches
themselves for continuity just to make sure they weren't bad (which
has happened to me before). Beyond something physically wrong with the
boards, is there anything else I might check?
On Wed, Oct 28, 2015 at 2:35 PM, Ben Sinclair <ben at bensinclair.com> wrote:
--
Ben Sinclair
ben at bensinclair.com
For 18 bit addresses, I assume I would look at
74400 instead? I get
177777 there.
I did just try moving the RLV11 boards below everything else (except
the BDV11), with the same results as before.
On Wed, Oct 28, 2015 at 12:29 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Ben Sinclair
ben at bensinclair.com On 2015-10-28 18:25, Ben Sinclair wrote:
--
Ben Sinclair
ben at bensinclair.com
I just tried to read 774400 from ODT, and it just gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com>
wrote:
>
> No magic smoke so far!
>
> I think you had written something a while back that would test that
> interrupt. Was that correct, or is there another diagnostic I can use
> to test that?
>
> The 17440 wasn't a typo, it's just my misunderstanding of how the
> addressing works. I'll read the actual location later today and see
> what I find!
>
> On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu>
> wrote:
>>
>> > From: Ben Sinclair
>>
>> > According to the Microcomputer Interfaces Handbook, my H9273
>> backplane
>> > should be all ABCD.
>>
>> Yup, that was one of the first things I checked - whilst realizing that
>> if
>> it wasn't, it was too late... :-)
>>
>> > Am I correct in that I shouldn't have any problem using the RLV11
>> in
>> > that backplane?
>>
>> Not that I am aware of.
>>
>> Noel
>
>
>
>
> --
> Ben Sinclair
> ben at bensinclair.com
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol 
1:46 p.m.
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben Sinclair wrote:
Well I just figured out something important...The
switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
1:58 p.m.
Thanks John, I think the insanity is my own! After changing those
switches I accidentally put the M8014 above the M8013. I've swapped
them and now get slightly different results:
One note, it now takes a while for it to start printing anything,
which I think might mean it's running some of the first test
successfully before getting to these. Earlier runs would start
printing errors immediately.
CVRLA DVC FTL ERR 00037 ON UNIT 00 TST 032 SUB 000 PC: 023576
RLMP: CRC OF DA+3 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 145554 REC'D: 175554
CVRLA DVC FTL ERR 00038 ON UNIT 00 TST 032 SUB 000 PC: 023634
RLMP: CRC OF CRC OF DA+4 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 151554 REC'D: 171554
CVRLA DVC FTL ERR 00039 ON UNIT 00 TST 032 SUB 000 PC: 023730
MAINT. FILL/EMPTY FIFO DMA DATA TRANSFER COMPARE ERROR
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
BA: 005760 DA: 155560 ADDR: 004762 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004764 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004766 EXP'D: 155555 REC'D 175555
On Wed, Oct 28, 2015 at 3:46 PM, John Wilson <wilson at dbit.com> wrote:
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben Sinclair
wrote:
--
Ben Sinclair
ben at bensinclair.com
Well I just figured out something important...The
switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
3:37 p.m.
Even better. Sorry for writing responses without checking what else is
in my inbox. I've caught up now, though.
And you definitely look like you got a stuck bit in there.
All of the errors show expected and received data, and in all cases bit
13 is on while it shouldn't be.
Could be a bad bus driver, a bad connection, or something more funny.
But start with the obvious ones. Do you have any extender cards so that
you can access the card with the power on, and measure signals. Checking
the source of bit 13 on output and trace if the signal is always on
somewhere would be a good exercise.
Do someone remember if there are some register in the RLV11 that can be
written and read back getting the same data out again? That could also
be interested to test things with...
Johnny
On 2015-10-28 21:58, Ben Sinclair wrote:
Thanks John, I think the insanity is my own! After
changing those
switches I accidentally put the M8014 above the M8013. I've swapped
them and now get slightly different results:
One note, it now takes a while for it to start printing anything,
which I think might mean it's running some of the first test
successfully before getting to these. Earlier runs would start
printing errors immediately.
CVRLA DVC FTL ERR 00037 ON UNIT 00 TST 032 SUB 000 PC: 023576
RLMP: CRC OF DA+3 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 145554 REC'D: 175554
CVRLA DVC FTL ERR 00038 ON UNIT 00 TST 032 SUB 000 PC: 023634
RLMP: CRC OF CRC OF DA+4 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 151554 REC'D: 171554
CVRLA DVC FTL ERR 00039 ON UNIT 00 TST 032 SUB 000 PC: 023730
MAINT. FILL/EMPTY FIFO DMA DATA TRANSFER COMPARE ERROR
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
BA: 005760 DA: 155560 ADDR: 004762 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004764 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004766 EXP'D: 155555 REC'D 175555
On Wed, Oct 28, 2015 at 3:46 PM, John Wilson <wilson at dbit.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben
Sinclair wrote:
Well I just figured out something important...The
switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
29 Oct
29 Oct
6:39 a.m.
Moving the boards after before the DLV-11 was an attempt to see if
Noel's idea around the DLV-11 not passing the grant might have been an
issue, so not completely random! However, I've been guilty of plenty
of random testing before!
I unfortunately don't have an extender card... I just searched eBay
for a few things, but didn't see any. Do they have a part number or
some other description that I might look for?
On Wed, Oct 28, 2015 at 5:37 PM, Johnny Billquist <bqt at update.uu.se> wrote:
Even better. Sorry for writing responses without
checking what else is in my
inbox. I've caught up now, though.
And you definitely look like you got a stuck bit in there.
All of the errors show expected and received data, and in all cases bit 13
is on while it shouldn't be.
Could be a bad bus driver, a bad connection, or something more funny. But
start with the obvious ones. Do you have any extender cards so that you can
access the card with the power on, and measure signals. Checking the source
of bit 13 on output and trace if the signal is always on somewhere would be
a good exercise.
Do someone remember if there are some register in the RLV11 that can be
written and read back getting the same data out again? That could also be
interested to test things with...
Johnny
On 2015-10-28 21:58, Ben Sinclair wrote:
--
Ben Sinclair
ben at bensinclair.com
Thanks John, I think the insanity is my own! After changing those
switches I accidentally put the M8014 above the M8013. I've swapped
them and now get slightly different results:
One note, it now takes a while for it to start printing anything,
which I think might mean it's running some of the first test
successfully before getting to these. Earlier runs would start
printing errors immediately.
CVRLA DVC FTL ERR 00037 ON UNIT 00 TST 032 SUB 000 PC: 023576
RLMP: CRC OF DA+3 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 145554 REC'D: 175554
CVRLA DVC FTL ERR 00038 ON UNIT 00 TST 032 SUB 000 PC: 023634
RLMP: CRC OF CRC OF DA+4 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 151554 REC'D: 171554
CVRLA DVC FTL ERR 00039 ON UNIT 00 TST 032 SUB 000 PC: 023730
MAINT. FILL/EMPTY FIFO DMA DATA TRANSFER COMPARE ERROR
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
BA: 005760 DA: 155560 ADDR: 004762 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004764 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004766 EXP'D: 155555 REC'D 175555
On Wed, Oct 28, 2015 at 3:46 PM, John Wilson <wilson at dbit.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben Sinclair wrote:
Well I just figured out something important...The switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
6:53 a.m.
Assuming I eventually get the controller working, I started looking at
my cabling and I don't think I have what I need to hook up the drive.
A while back I was talking to someone else who suggested using a
ribbon cable from the controller to directly inside the drive. That
technically will fit if I remove the ribbon inside the drive that goes
to the external connectors, but then I won't have a terminator
installed. I do have a terminator though!
I think what I really need, since I don't have a cab kit or anything,
is a bc80m, which I believe is a ribbon with a ground lug on the
controller side, and whatever type of connector the drive uses on the
other end. Is that correct? Not that I can even find a cable like that
anywhere currently...
On Thu, Oct 29, 2015 at 8:39 AM, Ben Sinclair <ben at bensinclair.com> wrote:
Moving the boards after before the DLV-11 was an
attempt to see if
Noel's idea around the DLV-11 not passing the grant might have been an
issue, so not completely random! However, I've been guilty of plenty
of random testing before!
I unfortunately don't have an extender card... I just searched eBay
for a few things, but didn't see any. Do they have a part number or
some other description that I might look for?
On Wed, Oct 28, 2015 at 5:37 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Ben Sinclair
ben at bensinclair.com
Even better. Sorry for writing responses without
checking what else is in my
inbox. I've caught up now, though.
And you definitely look like you got a stuck bit in there.
All of the errors show expected and received data, and in all cases bit 13
is on while it shouldn't be.
Could be a bad bus driver, a bad connection, or something more funny. But
start with the obvious ones. Do you have any extender cards so that you can
access the card with the power on, and measure signals. Checking the source
of bit 13 on output and trace if the signal is always on somewhere would be
a good exercise.
Do someone remember if there are some register in the RLV11 that can be
written and read back getting the same data out again? That could also be
interested to test things with...
Johnny
On 2015-10-28 21:58, Ben Sinclair wrote:
--
Ben Sinclair
ben at bensinclair.com
Thanks John, I think the insanity is my own! After changing those
switches I accidentally put the M8014 above the M8013. I've swapped
them and now get slightly different results:
One note, it now takes a while for it to start printing anything,
which I think might mean it's running some of the first test
successfully before getting to these. Earlier runs would start
printing errors immediately.
CVRLA DVC FTL ERR 00037 ON UNIT 00 TST 032 SUB 000 PC: 023576
RLMP: CRC OF DA+3 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 145554 REC'D: 175554
CVRLA DVC FTL ERR 00038 ON UNIT 00 TST 032 SUB 000 PC: 023634
RLMP: CRC OF CRC OF DA+4 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 151554 REC'D: 171554
CVRLA DVC FTL ERR 00039 ON UNIT 00 TST 032 SUB 000 PC: 023730
MAINT. FILL/EMPTY FIFO DMA DATA TRANSFER COMPARE ERROR
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
BA: 005760 DA: 155560 ADDR: 004762 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004764 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004766 EXP'D: 155555 REC'D 175555
On Wed, Oct 28, 2015 at 3:46 PM, John Wilson <wilson at dbit.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben Sinclair wrote:
>
> Well I just figured out something important...The switch numbers on my
> board backwards! I was going over the docs again, and noticed that MSB
> and LSB in their board diagram were on the wrong sides, compared to
> the switch numbers referenced on the next page. I'm not sure if that's
> just the way it is, or maybe someone replaced my switches at some
> point. After changing my switches I can now read 774400, which gives
> me 144201.
>
> The VRLAC0 diagnostic now goes further, but it thinks something is
> wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
8:15 a.m.
On 2015-10-29 14:39, Ben Sinclair wrote:
Moving the boards after before the DLV-11 was an
attempt to see if
Noel's idea around the DLV-11 not passing the grant might have been an
issue, so not completely random! However, I've been guilty of plenty
of random testing before!
Well, not important. We're now beyond that point. The boards are working
well enough that we can address them...
I unfortunately don't have an extender card... I
just searched eBay
for a few things, but didn't see any. Do they have a part number or
some other description that I might look for?
I have no idea if they have any designation or part number. I bet they
do, but I don't remember seeing any on my extenders, and I'm not close
to them now, so I can't check.
Looking through ebay, I found one 2-slot extended here:
http://www.ebay.com/itm/Generic-W984A-Extension-Board-USED-/221810115382?ha…
That will give you an idea what they look like. I've seen 2-, 4- and
6-slot versions. Obviously you can use several 2-slot ones if you need 4
or 6 slots extended.
Johnny
On Wed, Oct 28, 2015 at 5:37 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Even better. Sorry for writing responses without
checking what else is in my
inbox. I've caught up now, though.
And you definitely look like you got a stuck bit in there.
All of the errors show expected and received data, and in all cases bit 13
is on while it shouldn't be.
Could be a bad bus driver, a bad connection, or something more funny. But
start with the obvious ones. Do you have any extender cards so that you can
access the card with the power on, and measure signals. Checking the source
of bit 13 on output and trace if the signal is always on somewhere would be
a good exercise.
Do someone remember if there are some register in the RLV11 that can be
written and read back getting the same data out again? That could also be
interested to test things with...
Johnny
On 2015-10-28 21:58, Ben Sinclair wrote:
Thanks John, I think the insanity is my own! After changing those
switches I accidentally put the M8014 above the M8013. I've swapped
them and now get slightly different results:
One note, it now takes a while for it to start printing anything,
which I think might mean it's running some of the first test
successfully before getting to these. Earlier runs would start
printing errors immediately.
CVRLA DVC FTL ERR 00037 ON UNIT 00 TST 032 SUB 000 PC: 023576
RLMP: CRC OF DA+3 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 145554 REC'D: 175554
CVRLA DVC FTL ERR 00038 ON UNIT 00 TST 032 SUB 000 PC: 023634
RLMP: CRC OF CRC OF DA+4 ERROR (SERIAL DATA PATH)
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
EXP'D: 151554 REC'D: 171554
CVRLA DVC FTL ERR 00039 ON UNIT 00 TST 032 SUB 000 PC: 023730
MAINT. FILL/EMPTY FIFO DMA DATA TRANSFER COMPARE ERROR
CONTROLLER: 174400 DRIVE: 0
BEFORE COMMAND: CS: 112200 BA: 003762 DA: 155552 MP: 175555
TIME OF ERROR: CS: 000200 BA: 005760 DA: 155560 MP: 175554 171554
BA: 005760 DA: 155560 ADDR: 004762 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004764 EXP'D: 155555 REC'D 175555
BA: 005760 DA: 155560 ADDR: 004766 EXP'D: 155555 REC'D 175555
On Wed, Oct 28, 2015 at 3:46 PM, John Wilson <wilson at dbit.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On Wed, Oct 28, 2015 at 03:40:46PM -0500, Ben Sinclair wrote:
>
> Well I just figured out something important...The switch numbers on my
> board backwards! I was going over the docs again, and noticed that MSB
> and LSB in their board diagram were on the wrong sides, compared to
> the switch numbers referenced on the next page. I'm not sure if that's
> just the way it is, or maybe someone replaced my switches at some
> point. After changing my switches I can now read 774400, which gives
> me 144201.
>
> The VRLAC0 diagnostic now goes further, but it thinks something is
> wrong. Now of course I have no idea what it thinks is wrong here!
Bit 11 of the CSR seems to be stuck on. That's the "bad CRC" (or write
check error) bit ... definitely shouldn't be on after a bus init.
Maybe look at some other registers to see if that bit's on everywhere
(possibly bad bus xcvr if so), otherwise some other insanity?
John Wilson
D Bit
8:20 a.m.
On Thu, Oct 29, 2015 at 6:39 AM, Ben Sinclair <ben at bensinclair.com> wrote:
I unfortunately don't have an extender card... I just searched eBay
for a few things, but didn't see any. Do they have a part number or
some other description that I might look for?
If you need an extender card just buy a new one from Douglas Electronics.
I recently bought a 6-DE-8-C ($60) and a 6-DE-11-C ($40). I rarely see
them listed for less than that on eBay. (Also bought a 6-DE-2 S-100
extender from them for $12 while I was at it.)
http://www.douglas.com/index.php/6-de-8.html
The 6-DE-8 is a quad-width, dual-height DEC compatible extender board
with 144 contacts (72/side) on .125" centers on 4 tabs. It is
available as a bare board or with 72-125C-EL connectors installed
(6-DE-8-C option).
Height 8.43" Width 10.436"
http://www.douglas.com/index.php/6-de-11.html
The 6-DE-11 is a dual-width, dual-height DEC compatible extender board
with 72 contacts (36/side) on .125" centers on 2 tabs. It is available
as a bare board or with 72-125C-EL connectors installed (6-DE-11-C
option).
Height 8.43" Width 5.2"
28 Oct
28 Oct
3:31 p.m.
On 2015-10-28 21:40, Ben Sinclair wrote:
Well I just figured out something important...The
switch numbers on my
board backwards! I was going over the docs again, and noticed that MSB
and LSB in their board diagram were on the wrong sides, compared to
the switch numbers referenced on the next page. I'm not sure if that's
just the way it is, or maybe someone replaced my switches at some
point. After changing my switches I can now read 774400, which gives
me 144201.
The VRLAC0 diagnostic now goes further, but it thinks something is
wrong. Now of course I have no idea what it thinks is wrong here!
[...]
Ah! Progress. That explains your initial problems then.
Well, now that we can communicate with the controller, let's move on. Do
you have any drives connected? Do you have the terminator for the drives
in place? Is all the cards in the Q-bus continuous without any gaps in
the bus?
Johnny
DR>START
CHANGE HW (L) ? Y
# UNITS (D) ? 1
UNIT 0
11/23 PROCESSOR (L) Y ?
BUS ADDRESS (O) 174400 ?
VECTOR (O) 160 ? 330
DRIVE (O) 0 ?
BR LEVEL (O) 5 ?
CHANGE SW (L) ? Y
DROP ON ERROR LIMIT (L) N ?
AUTOSIZE (L) N ?
CVRLA DVC FTL ERR 00113 ON UNIT 00 TST 009 SUB 000 PC: 020052
BUS RESET DID NOT CLEAR RLCS
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00004 ON UNIT 00 TST 012 SUB 000 PC: 020264
RLCS READ/WRITE ERROR (BIT 0 DON'T CARE)
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00007 ON UNIT 00 TST 015 SUB 000 PC: 020600
BIT SET INSTRUCTION ON RLCS YIELDED WRONG RESULT
CONTROLLER: 174400 DRIVE: 0
EXP'D: 140200 REC'D: 144200
CVRLA DVC FTL ERR 00008 ON UNIT 00 TST 016 SUB 000 PC: 020724
BIT CLEAR INSTRUCTION ON RLCS YIELDED WRONG RESULT
CONTROLLER: 174400 DRIVE: 0
EXP'D: 141776 REC'D: 145776
... And many more!
On Wed, Oct 28, 2015 at 3:06 PM, Ben Sinclair <ben at bensinclair.com> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
Okay, I think I need to be looking at 774400,
which doesn't return
anything. Also, the PDP11GUI has a handy RL11 dialog to show all of
the registers, which I think shows the same thing... It's not there!
I've double checked my switch settings, and tested the switches
themselves for continuity just to make sure they weren't bad (which
has happened to me before). Beyond something physically wrong with the
boards, is there anything else I might check?
On Wed, Oct 28, 2015 at 2:35 PM, Ben Sinclair <ben at bensinclair.com> wrote:
For 18 bit addresses, I assume I would look at
74400 instead? I get
177777 there.
I did just try moving the RLV11 boards below everything else (except
the BDV11), with the same results as before.
On Wed, Oct 28, 2015 at 12:29 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Ben Sinclair
ben at bensinclair.com On 2015-10-28 18:25, Ben Sinclair wrote:
>
> I just tried to read 774400 from ODT, and it just gives the question
> mark. So, I don't think it's seeing it at all right now! I'll try
> moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
>
> On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com>
> wrote:
>>
>> No magic smoke so far!
>>
>> I think you had written something a while back that would test that
>> interrupt. Was that correct, or is there another diagnostic I can use
>> to test that?
>>
>> The 17440 wasn't a typo, it's just my misunderstanding of how the
>> addressing works. I'll read the actual location later today and see
>> what I find!
>>
>> On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at
mercury.lcs.mit.edu>
>> wrote:
>>>
>>> > From: Ben Sinclair
>>>
>>> > According to the Microcomputer Interfaces Handbook, my H9273
>>> backplane
>>> > should be all ABCD.
>>>
>>> Yup, that was one of the first things I checked - whilst realizing that
>>> if
>>> it wasn't, it was too late... :-)
>>>
>>> > Am I correct in that I shouldn't have any problem using the
RLV11
>>> in
>>> > that backplane?
>>>
>>> Not that I am aware of.
>>>
>>> Noel
>>
>>
>>
>>
>> --
>> Ben Sinclair
>> ben at bensinclair.com
>
>
>
>
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
--
Ben Sinclair
ben at bensinclair.com
3:28 p.m.
On 2015-10-28 20:35, Ben Sinclair wrote:
For 18 bit addresses, I assume I would look at 74400
instead? I get
177777 there.
Um? 74400 is 15 bits... :-)
18 bits would be 774400. But 22 bits would be 17774400. And 16 bits
would be 174400.
I would primarily expect 16 bits addresses to be used on a KDF-11, but I
could be wrong. Early revisions only supported 18 bit addressing, but
later models supported 22 bit addressing. But how that is reflected in
the micro-ODT on a KDF-11 I don't know. I don't even know if it's
actually micro-ODT, or some more primitivt console mode on that CPU.
I haven't actually worked on a KDF-11 in about 20 years. I know the
KDJ-11 better at this point. (Or the models with a physical front panel...)
I did just try moving the RLV11 boards below
everything else (except
the BDV11), with the same results as before.
I feel that this is just random testing and moving stuff around, which I
doubt is very helpful.
Johnny
On Wed, Oct 28, 2015 at 12:29 PM, Johnny Billquist <bqt at update.uu.se> wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
On 2015-10-28 18:25, Ben Sinclair wrote:
I just tried to read 774400 from ODT, and it just gives the question
mark. So, I don't think it's seeing it at all right now! I'll try
moving it next.
I wonder if it really uses 18 bit addresses in the console.
Johnny
On Wed, Oct 28, 2015 at 11:46 AM, Ben Sinclair <ben at bensinclair.com>
wrote:
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
No magic smoke so far!
I think you had written something a while back that would test that
interrupt. Was that correct, or is there another diagnostic I can use
to test that?
The 17440 wasn't a typo, it's just my misunderstanding of how the
addressing works. I'll read the actual location later today and see
what I find!
On Wed, Oct 28, 2015 at 11:43 AM, Noel Chiappa <jnc at mercury.lcs.mit.edu>
wrote:
>
> > From: Ben Sinclair
>
> > According to the Microcomputer Interfaces Handbook, my H9273
> backplane
> > should be all ABCD.
>
> Yup, that was one of the first things I checked - whilst realizing that
> if
> it wasn't, it was too late... :-)
>
> > Am I correct in that I shouldn't have any problem using the RLV11
> in
> > that backplane?
>
> Not that I am aware of.
>
> Noel
--
Ben Sinclair
ben at bensinclair.com
3116
days inactive
3117
days old
15 comments
5 participants
participants (5)
-
ben＠bensinclair.com -
bqt＠update.uu.se -
glen.slick＠gmail.com -
jnc＠mercury.lcs.mit.edu -
wilson＠dbit.com