6 Feb
2022
6 Feb
'22
9:20 a.m.
I have one of these and would like to use it, however it appears to only
partially work.? Here is what I have found that works and what doesn't:
1. CSR and DBR are present and operational.
2. Jumpers set to 'factory'.
3. D/A portion works, can deposit codes in ODT and see voltages out on
DAC pins that change depending on the octal value deposited in CSR+4 or +6.
4. A/D portion returns full scale code, either 3777 (2's compliment) or
7777 (offset binary) whether in the input is open or shorted to gnd.
I think the problem is that the A/D inputs are not exactly protected and
damage has occurred to this portion in the past.
Does anyone have any info on the A/D module?? Who made it?? Can you open
it up?? Does XXDP have a test for this?
Doug
6 Feb
6 Feb
9:22 a.m.
The board was made by Data Translation.? There is an identical board
sold by them, in the DT2x6x series but I can'r remember the exact number.
Nigel Johnson, MSc., MIEEE, MCSE VE3ID/G4AJQ/VA3MCU
Amateur Radio, the origin of the open-source concept!
Skype: TILBURY2591nw.johnson at ieee.org
On 2022-02-06 12:20, Douglas Taylor via cctalk wrote:
I have one of these and would like to use it, however
it appears to
only partially work.? Here is what I have found that works and what
doesn't:
1. CSR and DBR are present and operational.
2. Jumpers set to 'factory'.
3. D/A portion works, can deposit codes in ODT and see voltages out on
DAC pins that change depending on the octal value deposited in CSR+4
or +6.
4. A/D portion returns full scale code, either 3777 (2's compliment)
or 7777 (offset binary) whether in the input is open or shorted to gnd.
I think the problem is that the A/D inputs are not exactly protected
and damage has occurred to this portion in the past.
Does anyone have any info on the A/D module?? Who made it?? Can you
open it up?? Does XXDP have a test for this?
Doug
9:26 a.m.
Now that I have thought a bit more, I think it was the DT2762
The board was made by Data Translation.? There is an identical board
sold by them, in the DT2x6x series but I can'r remember the exact number.
Nigel Johnson, MSc., MIEEE, MCSE VE3ID/G4AJQ/VA3MCU
Amateur Radio, the origin of the open-source concept!
Skype: TILBURY2591nw.johnson at ieee.org
On 2022-02-06 12:20, Douglas Taylor via cctalk wrote:
I have one of these and would like to use it, however
it appears to
only partially work.? Here is what I have found that works and what
doesn't:
1. CSR and DBR are present and operational.
2. Jumpers set to 'factory'.
3. D/A portion works, can deposit codes in ODT and see voltages out on
DAC pins that change depending on the octal value deposited in CSR+4
or +6.
4. A/D portion returns full scale code, either 3777 (2's compliment)
or 7777 (offset binary) whether in the input is open or shorted to gnd.
I think the problem is that the A/D inputs are not exactly protected
and damage has occurred to this portion in the past.
Does anyone have any info on the A/D module?? Who made it?? Can you
open it up?? Does XXDP have a test for this?
Doug
9:41 a.m.
Yes, I have a couple of those and they are very close to the AXV11-C,
except for the D/A.? They have a similar method of working.?? The A/D
modules look 'identical' but I am suspicious that there are variations
of these modules that aren't readily apparent.
On 2/6/2022 12:26 PM, Nigel Johnson Ham via cctalk wrote:
Now that I have thought a bit more, I think it was the
DT2762
The board was made by Data Translation.? There is an identical board
sold by them, in the DT2x6x series but I can'r remember the exact number.
Nigel Johnson, MSc., MIEEE, MCSE VE3ID/G4AJQ/VA3MCU
Amateur Radio, the origin of the open-source concept!
Skype:? TILBURY2591nw.johnson at ieee.org
On 2022-02-06 12:20, Douglas Taylor via cctalk wrote:
> I have one of these and would like to use it, however it appears to
> only partially work.? Here is what I have found that works and what
> doesn't:
>
> 1. CSR and DBR are present and operational.
>
> 2. Jumpers set to 'factory'.
>
> 3. D/A portion works, can deposit codes in ODT and see voltages out
> on DAC pins that change depending on the octal value deposited in
> CSR+4 or +6.
>
> 4. A/D portion returns full scale code, either 3777 (2's compliment)
> or 7777 (offset binary) whether in the input is open or shorted to gnd.
>
> I think the problem is that the A/D inputs are not exactly protected
> and damage has occurred to this portion in the past.
>
> Does anyone have any info on the A/D module?? Who made it?? Can you
> open it up?? Does XXDP have a test for this?
>
> Doug
>
10:13 a.m.
I just took another look at the AXV11 board.? The A/D module has a
diagram glued on top which shows the block functions of the device.? If
you look close the pin numbers are listed where signals are brought out
to the 36pin connector that plugs the module into the host board.
I may be able to inject a signal on one of the inputs and see how far it
gets.? I think I can trace it through the multiplexer, the programmable
gain amp and to the sample and hold and see what the A/D chip input is.
Doug
On 2/6/2022 12:26 PM, Nigel Johnson Ham via cctalk wrote:
Now that I have thought a bit more, I think it was the
DT2762
The board was made by Data Translation.? There is an identical board
sold by them, in the DT2x6x series but I can'r remember the exact number.
Nigel Johnson, MSc., MIEEE, MCSE VE3ID/G4AJQ/VA3MCU
Amateur Radio, the origin of the open-source concept!
Skype:? TILBURY2591nw.johnson at ieee.org
On 2022-02-06 12:20, Douglas Taylor via cctalk wrote:
> I have one of these and would like to use it, however it appears to
> only partially work.? Here is what I have found that works and what
> doesn't:
>
> 1. CSR and DBR are present and operational.
>
> 2. Jumpers set to 'factory'.
>
> 3. D/A portion works, can deposit codes in ODT and see voltages out
> on DAC pins that change depending on the octal value deposited in
> CSR+4 or +6.
>
> 4. A/D portion returns full scale code, either 3777 (2's compliment)
> or 7777 (offset binary) whether in the input is open or shorted to gnd.
>
> I think the problem is that the A/D inputs are not exactly protected
> and damage has occurred to this portion in the past.
>
> Does anyone have any info on the A/D module?? Who made it?? Can you
> open it up?? Does XXDP have a test for this?
>
> Doug
>
9:33 a.m.
Are you triggering an A/D conversion via the CSR or external signal??
Then check the A/D done bit.
See? EK-AXV11-UG-02 Chapter 4.
?? Jerry
On 2/6/22 11:20 AM, Douglas Taylor via cctalk wrote:
I have one of these and would like to use it, however
it appears to
only partially work.? Here is what I have found that works and what
doesn't:
1. CSR and DBR are present and operational.
2. Jumpers set to 'factory'.
3. D/A portion works, can deposit codes in ODT and see voltages out on
DAC pins that change depending on the octal value deposited in CSR+4
or +6.
4. A/D portion returns full scale code, either 3777 (2's compliment)
or 7777 (offset binary) whether in the input is open or shorted to gnd.
I think the problem is that the A/D inputs are not exactly protected
and damage has occurred to this portion in the past.
Does anyone have any info on the A/D module?? Who made it?? Can you
open it up?? Does XXDP have a test for this?
Doug
9:43 a.m.
Yes, I am putting 1 into the CSR to start the conversion and I do get a
200(8) indicating that the conversion is complete.? I wonder if this
means the A/D chip is OK but something else, like the multiplexer chip
or gain amp is fried.
On 2/6/2022 12:33 PM, Jerry Weiss via cctalk wrote:
Are you triggering an A/D conversion via the CSR or
external signal??
Then check the A/D done bit.
See? EK-AXV11-UG-02 Chapter 4.
?? Jerry
On 2/6/22 11:20 AM, Douglas Taylor via cctalk wrote:
> I have one of these and would like to use it, however it appears to
> only partially work.? Here is what I have found that works and what
> doesn't:
>
> 1. CSR and DBR are present and operational.
>
> 2. Jumpers set to 'factory'.
>
> 3. D/A portion works, can deposit codes in ODT and see voltages out
> on DAC pins that change depending on the octal value deposited in
> CSR+4 or +6.
>
> 4. A/D portion returns full scale code, either 3777 (2's compliment)
> or 7777 (offset binary) whether in the input is open or shorted to gnd.
>
> I think the problem is that the A/D inputs are not exactly protected
> and damage has occurred to this portion in the past.
>
> Does anyone have any info on the A/D module?? Who made it?? Can you
> open it up?? Does XXDP have a test for this?
>
> Doug
>
10:55 a.m.
These A/D systems use methods to isolate the sensitive analog signals
from the electrical noise and ground plane of the
computer.? Typically a
differential input is standard, so you will probably need
wire up two
inputs.
If this is your first time with A/D, suggest you toggle in some code to
trigger and report the A/D conversion repeatedly and use a small voltage
battery with a potentiometer divider to drive the inputs.????? Most of
the analog inputs should be high impedance and while not impervious, can
take +- 30v w/o damage.
?Jerry
On 2/6/22 11:43 AM, Douglas Taylor via cctalk wrote:
Yes, I am putting 1 into the CSR to start the
conversion and I do get
a 200(8) indicating that the conversion is complete.? I wonder if this
means the A/D chip is OK but something else, like the multiplexer chip
or gain amp is fried.
On 2/6/2022 12:33 PM, Jerry Weiss via cctalk wrote:
Are you triggering an A/D conversion via the CSR
or external signal??
Then check the A/D done bit.
See? EK-AXV11-UG-02 Chapter 4.
?? Jerry
On 2/6/22 11:20 AM, Douglas Taylor via cctalk wrote:
> I have one of these and would like to use it, however it appears to
> only partially work.? Here is what I have found that works and what
> doesn't:
>
> 1. CSR and DBR are present and operational.
>
> 2. Jumpers set to 'factory'.
>
> 3. D/A portion works, can deposit codes in ODT and see voltages out
> on DAC pins that change depending on the octal value deposited in
> CSR+4 or +6.
>
> 4. A/D portion returns full scale code, either 3777 (2's compliment)
> or 7777 (offset binary) whether in the input is open or shorted to gnd.
>
> I think the problem is that the A/D inputs are not exactly protected
> and damage has occurred to this portion in the past.
>
> Does anyone have any info on the A/D module?? Who made it? Can you
> open it up?? Does XXDP have a test for this?
>
> Doug
>
8 Feb
8 Feb
11:34 a.m.
Update on this:? I did put together a battery and voltage divider to
test the AXV11.? The label on the A/D module says it brings the output
from the multiplexer to one of the external pins.? I
was able to verify
that the voltage applied to a couple of the A/D inputs makes it
through
the multiplexer when selected using the CSR.? The next output available
is from the Sample and Hold, and this is always pegged at +12v.? Am I
wrong to assume that the sample and hold will 'freeze' its output when
the A/D go bit is set?
At least the D/A's work on the AXV11 I have!? Glass half full.
Next tried the Data Translation DT2762 with the voltage divider source
and this board appears to work correctly.? What I was able to observe
was by applying different voltages to the input I see different binary
values in the DBR.
The manual from bitsavers for this board lists a complete Macro
diagnostic program in an appendix. Does anyone have an electronic copy
of this Macro code?? Or the compiled utility?
Doug
On 2/6/2022 1:55 PM, Jerry Weiss via cctalk wrote:
These A/D systems use methods to isolate the sensitive
analog signals
from the electrical noise and ground plane of the computer.? Typically
a differential input is standard, so you will probably need wire up
two inputs.
If this is your first time with A/D, suggest you toggle in some code
to trigger and report the A/D conversion repeatedly and use a small
voltage battery with a potentiometer divider to drive the inputs.?????
Most of the analog inputs should be high impedance and while not
impervious, can take +- 30v w/o damage.
?Jerry
On 2/6/22 11:43 AM, Douglas Taylor via cctalk wrote:
> Yes, I am putting 1 into the CSR to start the conversion and I do get
> a 200(8) indicating that the conversion is complete.? I wonder if
> this means the A/D chip is OK but something else, like the
> multiplexer chip or gain amp is fried.
>
> On 2/6/2022 12:33 PM, Jerry Weiss via cctalk wrote:
>> Are you triggering an A/D conversion via the CSR or external
>> signal?? Then check the A/D done bit.
>>
>> See? EK-AXV11-UG-02 Chapter 4.
>>
>> ?? Jerry
>>
>> On 2/6/22 11:20 AM, Douglas Taylor via cctalk wrote:
>>> I have one of these and would like to use it, however it appears to
>>> only partially work.? Here is what I have found that works and what
>>> doesn't:
>>>
>>> 1. CSR and DBR are present and operational.
>>>
>>> 2. Jumpers set to 'factory'.
>>>
>>> 3. D/A portion works, can deposit codes in ODT and see voltages out
>>> on DAC pins that change depending on the octal value deposited in
>>> CSR+4 or +6.
>>>
>>> 4. A/D portion returns full scale code, either 3777 (2's
>>> compliment) or 7777 (offset binary) whether in the input is open or
>>> shorted to gnd.
>>>
>>> I think the problem is that the A/D inputs are not exactly
>>> protected and damage has occurred to this portion in the past.
>>>
>>> Does anyone have any info on the A/D module?? Who made it? Can you
>>> open it up?? Does XXDP have a test for this?
>>>
>>> Doug
>>>
>
11:49 a.m.
On Feb 8, 2022, at 2:34 PM, Douglas Taylor via cctalk
<cctalk at classiccmp.org> wrote:
Update on this: I did put together a battery and voltage divider to test the AXV11. The
label on the A/D module says it brings the output from the multiplexer to one of the
external pins. I was able to verify that the voltage applied to a couple of the A/D
inputs makes it through the multiplexer when selected using the CSR. The next output
available is from the Sample and Hold, and this is always pegged at +12v. Am I wrong to
assume that the sample and hold will 'freeze' its output when the A/D go bit is
set?
You're correct, a S/H circuit is supposed to hold the value that was on its input at
the time it was told to take the sample. Typically it won't hold it
"forever"; S/H circuits have a hold time spec chosen so it is substantially
longer than the time it takes the A/D behind it to complete its measurement.
paul
2:22 p.m.
On 2/8/22 13:34, Douglas Taylor via cctalk wrote:
Update on this:? I did put together a battery and
voltage
divider to test the AXV11.? The label on the A/D module
says it brings the output from the multiplexer to one of
the external pins.? I was able to verify that the voltage
applied to a couple of the A/D inputs makes it through the
multiplexer when selected using the CSR.? The next output
available is from the Sample and Hold, and this is always
pegged at +12v.? Am I wrong to assume that the sample and
hold will 'freeze' its output when the A/D go bit is set?
Well, it will probably only hold the approximate voltage for
a few seconds, but should be long enough to see on a voltmeter.
Jon
2:35 p.m.
On 2/8/2022 5:22 PM, Jon Elson via cctalk wrote:
On 2/8/22 13:34, Douglas Taylor via cctalk wrote:
I was only using a voltmeter to look at it.
I'm going to try 2 different measurements; (1) use a scope to look at
the S&H signal on the AXV11 and see what it does over time, (2) look at
the same S&H signal on the Data Translation board which seems to be
working.
Update on this:? I did put together a battery and
voltage divider to
test the AXV11.? The label on the A/D module says it brings the
output from the multiplexer to one of the external pins.? I was able
to verify that the voltage applied to a couple of the A/D inputs
makes it through the multiplexer when selected using the CSR.? The
next output available is from the Sample and Hold, and this is always
pegged at +12v.? Am I wrong to assume that the sample and hold will
'freeze' its output when the A/D go bit is set?
Well, it will probably only hold the approximate voltage for a few
seconds, but should be long enough to see on a voltmeter.
Jon
4:36 p.m.
On 2/8/22 16:35, Douglas Taylor via cctalk wrote:
On 2/8/2022 5:22 PM, Jon Elson via cctalk wrote:
The S&H will likely have a FET as the switch element, and
those might be fairly easily damaged by external pulses.? I
opened up some comparable Xincom modules and made minor repairs.
Jon
On 2/8/22 13:34, Douglas Taylor via cctalk
wrote:
I was only using a voltmeter to look at it.
I'm going to try 2 different measurements; (1) use a scope
to look at the S&H signal on the AXV11 and see what it
does over time, (2) look at the same S&H signal on the
Data Translation board which seems to be working.
Update on this:? I did put together a battery and
voltage divider to test the AXV11.? The label on the A/D
module says it brings the output from the multiplexer to
one of the external pins.? I was able to verify that the
voltage applied to a couple of the A/D inputs makes it
through the multiplexer when selected using the CSR.?
The next output available is from the Sample and Hold,
and this is always pegged at +12v.? Am I wrong to assume
that the sample and hold will 'freeze' its output when
the A/D go bit is set?
Well, it will probably only hold the approximate voltage
for a few seconds, but should be long enough to see on a
voltmeter.
Jon
13 Feb
13 Feb
9:04 a.m.
On 2/8/2022 7:36 PM, Jon Elson via cctalk wrote:
On 2/8/22 16:35, Douglas Taylor via cctalk wrote:
Stopped fooling with the AXV11 for now.? Applied various voltages to the
Data Translation input and recorded the A/D octal values to get an idea
of what the calibration of the board is.? It looks very linear, +/-10v
range.
Using a single battery and voltage divider I was able to generate
voltages on the input of the DT2762 board, however, I had to swap wires
to get negative voltages.? Is it possible to construct a battery driven
circuit that will present both positive and negative voltages at the
input?? A bridge of some sort?
Doug
On 2/8/2022 5:22 PM, Jon Elson via cctalk wrote:
The S&H will likely have a FET as the switch element, and those might
be fairly easily damaged by external pulses.? I opened up some
comparable Xincom modules and made minor repairs.
Jon
On 2/8/22 13:34, Douglas Taylor via cctalk
wrote:
I was only using a voltmeter to look at it.
I'm going to try 2 different measurements; (1) use a scope to look at
the S&H signal on the AXV11 and see what it does over time, (2) look
at the same S&H signal on the Data Translation board which seems to
be working.
Update on this:? I did put together a battery and
voltage divider
to test the AXV11.? The label on the A/D module says it brings the
output from the multiplexer to one of the external pins.? I was
able to verify that the voltage applied to a couple of the A/D
inputs makes it through the multiplexer when selected using the
CSR.? The next output available is from the Sample and Hold, and
this is always pegged at +12v.? Am I wrong to assume that the
sample and hold will 'freeze' its output when the A/D go bit is set?
Well, it will probably only hold the approximate voltage for a few
seconds, but should be long enough to see on a voltmeter.
Jon
9:43 a.m.
This is a really interesting thread, I have this darn AF01 A/D converter
for a negative bus pdp8 and your issues and methods with testing would
work well for testing this thing. Bonus it does appear to have a 16 or
64 channel mux capability.
Question: Is it possible to change the driver boards for ones that do
positive bus logic or is it simpler to build a DW08 converter clone to
hook it up to my pdp8/L for testing?
C
On 2/13/2022 12:04 PM, Douglas Taylor via cctalk wrote:
On 2/8/2022 7:36 PM, Jon Elson via cctalk wrote:
On 2/8/22 16:35, Douglas Taylor via cctalk
wrote:
Stopped fooling with the AXV11 for now.? Applied various voltages to the
Data Translation input and recorded the A/D octal values to get an idea
of what the calibration of the board is.? It looks very linear, +/-10v
range.
Using a single battery and voltage divider I was able to generate
voltages on the input of the DT2762 board, however, I had to swap wires
to get negative voltages.? Is it possible to construct a battery driven
circuit that will present both positive and negative voltages at the
input?? A bridge of some sort?
Doug
On 2/8/2022 5:22 PM, Jon Elson via cctalk wrote:
The S&H will likely have a FET as the switch element, and those might
be fairly easily damaged by external pulses.? I opened up some
comparable Xincom modules and made minor repairs.
Jon
On 2/8/22 13:34, Douglas Taylor via cctalk
wrote:
> Update on this:? I did put together a battery and voltage divider
> to test the AXV11.? The label on the A/D module says it brings the
> output from the multiplexer to one of the external pins.? I was
> able to verify that the voltage applied to a couple of the A/D
> inputs makes it through the multiplexer when selected using the
> CSR.? The next output available is from the Sample and Hold, and
> this is always pegged at +12v.? Am I wrong to assume that the
> sample and hold will 'freeze' its output when the A/D go bit is set?
Well, it will probably only hold the approximate voltage for a few
seconds, but should be long enough to see on a voltmeter.
Jon
I was only using a voltmeter to look at it.
I'm going to try 2 different measurements; (1) use a scope to look at
the S&H signal on the AXV11 and see what it does over time, (2) look
at the same S&H signal on the Data Translation board which seems to
be working.
10:03 a.m.
On 2022-Feb-13, at 9:04 AM, Douglas Taylor via cctalk wrote:
Stopped fooling with the AXV11 for now. Applied
various voltages to the Data Translation input and recorded the A/D octal values to get an
idea of what the calibration of the board is. It looks very linear, +/-10v range.
Using a single battery and voltage divider I was able to generate voltages on the input
of the DT2762 board, however, I had to swap wires to get negative voltages. Is it
possible to construct a battery driven circuit that will present both positive and
negative voltages at the input? A bridge of some sort?
You said it: what amounts to a wheatstone bridge. Take say, two 10K R, in series across
the battery forming a 1:1-split voltage divider. The center point shall be the common/0V.
The wiper of a pot across the battery forming a variable voltage divider presents the +/-
test signal.
That would give e.g. +/-6V from a 12V battery.
It also presumes you have a reasonably high impedance on your target input/load as you
need to keep the Rs adequately high to keep from draining the battery.
If you wanted to get +/-12V from a 12V battery, or needed a lower impedance source, a
couple of op-amps could be worked up into an equivalent circuit. E.g. two unit-followers,
one inverting and one non-inverting, both fed, in counter-point, from the fixed divider
and the variable divider, 0V from one output, signal from the other output.
12:40 p.m.
Two batteries in series, using the ?middle? as 0V reference.
The ?+? is V+, the ?-? is V-.
Van: Douglas Taylor via cctalk<mailto:cctalk at classiccmp.org>
Verzonden: zondag 13 februari 2022 18:05
Aan: Jon Elson via cctalk<mailto:cctalk at classiccmp.org>
Onderwerp: Re: DEC AXV11-C analog board
Is it possible to construct a battery driven circuit that
will present both positive and negative voltages at the
input? A bridge of some sort?
Doug
19 Feb
19 Feb
9:03 a.m.
Put one together with back to back 9V batteries and a 5K pot.? Does
exactly what I wanted.
The Data Translation board is set up for SE inputs and +/-10 V input range.
Was able to apply a series of DC voltages to the DT2762 board and record
the A/D value using ODT (tedious, but works OK).
Using Octave (Linux) was able to convert the A/D 2's complement Octal
numbers to decimal and plot input voltage versus A/D decimal value.?
Expected to see a nice clean linear plot.
The one DT2762 seemed to 'drop' bits, while the newer one seemed OK.
Next step is to capture a time varying signal and see if the A/D output
follows the input correctly.? I'd like to use Macro-11 to manage the A/D
board and Fortran to deal with the data.
What is the Fortran or F77 interface with Macro-11 routines?
Doug
On 2/13/2022 3:40 PM, Henk Gooijen via cctalk wrote:
Two batteries in series, using the ?middle? as 0V
reference.
The ?+? is V+, the ?-? is V-.
Van: Douglas Taylor via cctalk<mailto:cctalk at classiccmp.org>
Verzonden: zondag 13 februari 2022 18:05
Aan: Jon Elson via cctalk<mailto:cctalk at classiccmp.org>
Onderwerp: Re: DEC AXV11-C analog board
Is it possible to construct a battery driven circuit that
will present both positive and negative voltages at the
input? A bridge of some sort?
Doug
8 Feb
8 Feb
1:35 p.m.
On Sun, Feb 6, 2022 at 9:20 AM Douglas Taylor via cctalk
<cctalk at classiccmp.org> wrote:
Does anyone have any info on the A/D module? Who made it? Can you open
it up? Does XXDP have a test for this?
http://www.bitsavers.org/pdf/dec/pdp11/xxdp/PDP11_DiagnosticHandbook_1988.p…
PDP11 DIAGNOSTIC HANDBOOK 1988, Page A-7:
VAXAB0.BIC ADV11-C/AXV11-C A/D CONVERTER TEST
807
days inactive
820
days old
18 comments
9 participants
participants (9)
-
bhilpert＠shaw.ca -
cz＠alembic.crystel.com -
dj.taylor4＠comcast.net -
elson＠pico-systems.com -
g4ajq1＠gmail.com -
glen.slick＠gmail.com -
henk.gooijen＠hotmail.com -
jsw＠ieee.org -
paulkoning＠comcast.net