22 Apr
2013
22 Apr
'13
3:37 p.m.
Hey all,
I recently acquired a KA655-AA (MicroVAX 3800/3900 CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. If I'm reading the docs correctly (and I have to go all
the way back to the KA630 docs to find the proper switch polarity),
if I'm running it without a bulkhead attached at all, the serial
speed should be 300 baud. When my terminal is running at 300,
8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Am I missing something? Do I need to do some magic with the front
panel ports to make it cooperate? I'm using a DB25 serial cable
that I've been using on my KDF11-B CPU board with no trouble; if
I look at the docs, it looks like it should be fine assuming I can
let DTR float (it's unused in my cable, since it's an external
clock pin on the KDF11-B). I can also see the LEDs on the spine
of the CPU counting down, though I didn't watch for long enough to
see if they ever got to zero.
As far as I know, the memory is connected OK and works, though I
assume I don't really need that if I just want to see a console
prompt. I'm using a very short (~3") 50-pin SCSI ribbon cable,
which I assume is OK. And it's an MS650-BA, which the KA655
manual specifically calls out, because the -AA is too slow.
- Dave
23 Apr
23 Apr
7:15 a.m.
I recently acquired a KA655-AA (MicroVAX 3800/3900
CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. [...] if I'm running it without a bulkhead attached at all,
the serial speed should be 300 baud. When my terminal is running at
300, 8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Have you tried the other speeds the KA630 documentation mentions? It
strikes me as possible, at least, that the KA630 interprets those
signals differently from the KA655.
The other possibility that comes to mind is that its serial output
isn't a normal single-ended RS232, but a differential pair (my memory
calls that RS432, but I don't know how accurate that is) and you've got
the wrong output pin - inverting the data line produces weeeeird data
corruption. (I know because I once made that mistake myself and was
thoroughly baffled for quite a while until I realized what was up.)
[...]; if I look at the docs, it looks like it should
be fine
assuming I can let DTR float (it's unused in my cable, since it's an
external clock pin on the KDF11-B).
If you get any output, I would assume, until finding evidence
otherwise, that DTR is not an issue.
If you can quote some of the garbage you get (preferably a hex dump of
the serial-line data, since there may be nonprintable characters in
it), we might be able to make guesses from that. In particular, "wrong
speed" garbage usually has a very different feel from other kinds of
garbage, but in a way I, at least, find difficult to elucidate.
/~\ 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
7:43 a.m.
On Apr 23, 2013, at 10:15 AM, Mouse wrote:
Actually, it turned out that the problem was that I was plugging
the serial adaptor in upside down (it wasn't keyed and the cable
was a rainbow ribbon that did not have brown on pin 1, so oops).
I would have figured that out sooner if I hadn't still been
seeing data on the line; apparently the Tx makes enough noise on
the Rx pins (which are what you plug into if it's upside down)
to trigger valid-looking bits on the receiving UART.
The standard DEC pinout has single-ended Tx and differential Rx,
so I was pretty sure I wasn't receiving backwards. Of course,
now I'm receiving at 300 baud, which isn't a LOT better. :-)
So now I'm able to boot the VMS install CD, but it hangs after
identifying the devices and asking if they're all there. I'm
still trying to figure out what the deal is with that, though
I suspect it may have something to do with the DLV11J (4x UART)
I have in there, because it seems to identify that as 8 TU58s.
I'll try installing without that plugged in.
Just in case it rings a bell with anyone here, though, if I'm
seeing a hang after this, is there anything in particular I
should do? I'm booting a SCSI CD drive off a CQD-220, which
seems to work OK for booting, at least.
----
If any device does not show up, please take action now to make it
available.
Available device DUA0: device type RA81
Available device DUA1: device type RA82
Available device DUA2: device type RRD40
Available device MSA0: device type MW_TSV05
Available device DDA0: device type TU58
Available device DDA1: device type TU58
Available device DDB0: device type TU58
Available device DDB1: device type TU58
Available device DDC0: device type TU58
Available device DDC1: device type TU58
Available device DDD0: device type TU58
Available device DDD1: device type TU58
Enter "YES" when all needed devices are available: y
<hang, no echo from console>
- Dave
I recently
acquired a KA655-AA (MicroVAX 3800/3900 CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. [...] if I'm running it without a bulkhead attached at all,
the serial speed should be 300 baud. When my terminal is running at
300, 8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Have you tried the other speeds the KA630 documentation mentions? It
strikes me as possible, at least, that the KA630 interprets those
signals differently from the KA655.
The other possibility that comes to mind is that its serial output
isn't a normal single-ended RS232, but a differential pair (my memory
calls that RS432, but I don't know how accurate that is) and you've got
the wrong output pin - inverting the data line produces weeeeird data
corruption. (I know because I once made that mistake myself and was
thoroughly baffled for quite a while until I realized what was up.) 
2:52 p.m.
On 04/23/2013 10:43 AM, David Riley wrote:
On Apr 23, 2013, at 10:15 AM, Mouse wrote:
Actually, it
turned out that the problem was that I was plugging
the serial adaptor in upside down (it wasn't keyed and the cable
was a rainbow ribbon that did not have brown on pin 1, so oops).
I would have figured that out sooner if I hadn't still been
seeing data on the line; apparently the Tx makes enough noise on
the Rx pins (which are what you plug into if it's upside down)
to trigger valid-looking bits on the receiving UART.
The standard DEC pinout has single-ended Tx and differential Rx,
so I was pretty sure I wasn't receiving backwards. Of course,
now I'm receiving at 300 baud, which isn't a LOT better. :-)
So now I'm able to boot the VMS install CD, but it hangs after
identifying the devices and asking if they're all there. I'm
still trying to figure out what the deal is with that, though
I suspect it may have something to do with the DLV11J (4x UART)
I have in there, because it seems to identify that as 8 TU58s.
I'll try installing without that plugged in.
Just in case it rings a bell with anyone here, though, if I'm
seeing a hang after this, is there anything in particular I
should do? I'm booting a SCSI CD drive off a CQD-220, which
seems to work OK for booting, at least.
Check your bus grant sequence and also the assigned addresses
of the IO cards (all of them).
I 'll bet you have more than a few on the same address or incorrect ones.
There is a specific set that is best to use and also the grant chain will
have the system see devices and hang using them (or try to).
Allison
I
recently acquired a KA655-AA (MicroVAX 3800/3900 CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. [...] if I'm running it without a bulkhead attached at all,
the serial speed should be 300 baud. When my terminal is running at
300, 8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Have you tried the other speeds the KA630 documentation mentions?
It
strikes me as possible, at least, that the KA630 interprets those
signals differently from the KA655.
The other possibility that comes to mind is that its serial output
isn't a normal single-ended RS232, but a differential pair (my memory
calls that RS432, but I don't know how accurate that is) and you've got
the wrong output pin - inverting the data line produces weeeeird data
corruption. (I know because I once made that mistake myself and was
thoroughly baffled for quite a while until I realized what was up.)
----
If any device does not show up, please take action now to make it
available.
Available device DUA0: device type RA81
Available device DUA1: device type RA82
Available device DUA2: device type RRD40
Available device MSA0: device type MW_TSV05
Available device DDA0: device type TU58
Available device DDA1: device type TU58
Available device DDB0: device type TU58
Available device DDB1: device type TU58
Available device DDC0: device type TU58
Available device DDC1: device type TU58
Available device DDD0: device type TU58
Available device DDD1: device type TU58
Enter "YES" when all needed devices are available: y
<hang, no echo from console>
- Dave
3:32 p.m.
On Apr 23, 2013, at 5:52 PM, allison wrote:
Those are fine; this is actually the exact same setup as my
11/23 with the CPU and RAM swapped out (it's an H9276 backplane,
so it should be fine with that arrangement). That machine was
just fine.
For the record, the setup is:
KA655
MS650-BA
DELQA
CQD-220
I used to have a DLV11J in there as well (configured for CSR/INT
1776500/300) in between the DELQA and the CQD-220, but I removed
that for now in case the TU58 driver was interfering (probably
not, since it didn't solve the problem). I also took out the
TSV05, since the tape drive it goes to isn't behaving right now
anyway.
If it were a grant chain continuity problem, I wouldn't expect
the SCSI card to even be able to boot, since it's MSCP-based
(unless the early MSCP boot stuff doesn't require an interrupt,
which would surprise me).
Just in case
it rings a bell with anyone here, though, if I'm
seeing a hang after this, is there anything in particular I
should do? I'm booting a SCSI CD drive off a CQD-220, which
seems to work OK for booting, at least.
Check your bus grant sequence and also the assigned addresses
of the IO cards (all of them). I 'll bet you have more than a few on the same
address or incorrect ones.
There is a specific set that is best to use and also the grant chain will
have the system see devices and hang using them (or try to).
There are definitely no address conflicts unless the KA655 is
adding something I'm not aware of (the DLV11J is NOT set to
provide the console SLU address, if that even matters, but it
is also not currently installed). My guess is that it's just
a matter of the CQD-220 not playing nicely with my CD-ROM drive,
though I'm a bit puzzled as to why it boots the initial part
and barfs at trying to start BACKUP. Some extra verbosity
would be handy there; I suppose if I get really desperate, I
could just try halting and disassembling and see where it's
waiting.
- Dave
8:26 a.m.
On 04/23/2013 10:15 AM, Mouse wrote:
you need the bulkhead adaptor to set baud and all.
You will have to
make your own cable, that fakes the bulkhead switches.
Allison
I recently
acquired a KA655-AA (MicroVAX 3800/3900 CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. [...] if I'm running it without a bulkhead attached at all,
the serial speed should be 300 baud. When my terminal is running at
300, 8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Have you tried the other speeds the KA630 documentation mentions?
It
strikes me as possible, at least, that the KA630 interprets those
signals differently from the KA655.
The other possibility that comes to mind is that its serial output
isn't a normal single-ended RS232, but a differential pair (my memory
calls that RS432, but I don't know how accurate that is) and you've got
the wrong output pin - inverting the data line produces weeeeird data
corruption. (I know because I once made that mistake myself and was
thoroughly baffled for quite a while until I realized what was up.)
Its RS423. Its an abbreviate flavor of RS232 for fewer active pins.
[...]; if I
look at the docs, it looks like it should be fine
assuming I can let DTR float (it's unused in my cable, since it's an
external clock pin on the KDF11-B).
If you get any output, I would assume, until
finding evidence
otherwise, that DTR is not an issue.
If you can quote some of the garbage you get (preferably a hex dump of
the serial-line data, since there may be nonprintable characters in
it), we might be able to make guesses from that. In particular, "wrong
speed" garbage usually has a very different feel from other kinds of
garbage, but in a way I, at least, find difficult to elucidate. /~\ 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
9:18 a.m.
Have you tried
the other speeds [...]
you need the bulkhead adaptor to set baud and all. You will have to make your own cable, that fakes the
bulkhead
switches.
Aye, that's something I would definitely do/suggest. 300 works
minimally, but is pretty painful. As I'm sure pretty much everyone
here knows from experience. :)
/~\ 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
9:26 a.m.
On Apr 23, 2013, at 11:26 AM, Allison <ajp166 at verizon.net> wrote:
It's a weird hybrid, in that there is a differential receiver. But
if you ground one leg of the receiver, it's fine. I'm using the same
cables that I use on my -11, and that works fine.
As I mentioned in an update, it turns out that I had plugged the cable
in upside down because it wasn't keyed either mechanically or visually.
On 04/23/2013 10:15 AM, Mouse wrote:
you need the bulkhead adaptor to set baud and all.
Or jumpers, but yes. If the switch pins are all open, that does
correspond to 300 baud, as expected.
I
recently acquired a KA655-AA (MicroVAX 3800/3900 CPU) and some
appropriate RAM, and I'm having trouble grokking the console SLU
output. [...] if I'm running it without a bulkhead attached at all,
the serial speed should be 300 baud. When my terminal is running at
300, 8 bit, no parity, 1 stop bit, I get some garbage on screen but
not much more.
Have you tried the other speeds the KA630 documentation mentions?
It
strikes me as possible, at least, that the KA630 interprets those
signals differently from the KA655. The other
possibility that comes to mind is that its serial output
isn't a normal single-ended RS232, but a differential pair (my memory
calls that RS432, but I don't know how accurate that is) and you've got
the wrong output pin - inverting the data line produces weeeeird data
corruption. (I know because I once made that mistake myself and was
thoroughly baffled for quite a while until I realized what was up.)
Its RS423. Its an abbreviate flavor of RS232 for fewer active pins. You will have to make your own cable, that fakes the
bulkhead switches.
It's working for now, but I definitely have plans to make a small board
to do that; among other things, the bulkhead also supplies the battery
power for the NVRAM and switches for run mode, etc. that I need. I
just haven't had time yet.
- Dave
4263
days inactive
4264
days old
7 comments
3 participants
participants (3)
-
ajp166＠verizon.net -
fraveydank＠gmail.com -
mouse＠Rodents-Montreal.ORG