1 Dec
2014
1 Dec
'14
3:44 p.m.
I have a Commodore 1950 monitor that came along with an Amiga 3000, and I think it has
defeated me. Its failure mode is that there's no vertical scan; just a single,
horizontal line. Vertical drive is still working, because I can shift the line up and down
a bit using the vertical centering control, but whatever circuitry generates the vertical
scan doesn't appear to be doing it.
Trying to debug it has been challenging, even though I found a service manual online. The
monitor appears to have been designed to be slapped together without much consideration of
service accessibility, and the service manual does't describe the circuitry in much
detail. It includes a block diagram, with no indication of which actual circuitry
corresponds to which blocks. The schematic diagrams show connectivity between the parts,
but don't identify the functionality of most of the ICs.
I think I've narrowed down to one IC that's a likely candidate for the failed
component, but it's still just an educated guess that it's even the IC responsible
for generating the vertical scan. And it's a no-name IC with the monitor
manufacturer's internal part number (IC401, with part number 56A326-1), so I don't
think I'm likely to find a replacement for it unless I find another model 1950 with a
different failure.
I have an LCD monitor that I can use with the A3000, but the A3000 just doesn't look
right with a flat panel on top of it. Now I wish I hadn't gotten rid of my old
Viewsonic 17" CRT monitor all those years ago! I was sure happy to see it go at the
time, but I failed to predict that I'd get interested in retrocomputing a decade
later.
So, I guess I'll just keep my eyes open for a suitable CRT monitor that I like for the
A3000. Maybe some older Viewsonic monitor in the 13"-17" range? If anybody near
southern California might be able to use my failed 1950 for parts, or has a 1950 with a
different failure that they'd like to dump on me, please let me know! Even though I
don't anticipate being able to fix it, I don't want to throw it away, in case
parts appear or some other collector can make use of it. So, I'll stash its carcass
away for the time being.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
1 Dec
1 Dec
4:30 p.m.
Are you just driving it off the built in VGA, or do you have a VGA card in it? I have a
Picasso IV card in mine. Before I got the card I was just using an old 13" Magnavox
VGA monitor, then I switched to a bigger and better SVGA monitor.
One of these days I *will* get one of my Amiga's hooked up and running again!
Zane
On Dec 1, 2014, at 3:44 PM, "Mark J. Blair" <nf6x at nf6x.net> wrote:
I have a Commodore 1950 monitor that came along with
an Amiga 3000, and I think it has defeated me. Its failure mode is that there's no
vertical scan; just a single, horizontal line. Vertical drive is still working, because I
can shift the line up and down a bit using the vertical centering control, but whatever
circuitry generates the vertical scan doesn't appear to be doing it.
Trying to debug it has been challenging, even though I found a service manual online. The
monitor appears to have been designed to be slapped together without much consideration of
service accessibility, and the service manual does't describe the circuitry in much
detail. It includes a block diagram, with no indication of which actual circuitry
corresponds to which blocks. The schematic diagrams show connectivity between the parts,
but don't identify the functionality of most of the ICs.
I think I've narrowed down to one IC that's a likely candidate for the failed
component, but it's still just an educated guess that it's even the IC responsible
for generating the vertical scan. And it's a no-name IC with the monitor
manufacturer's internal part number (IC401, with part number 56A326-1), so I don't
think I'm likely to find a replacement for it unless I find another model 1950 with a
different failure.
I have an LCD monitor that I can use with the A3000, but the A3000 just doesn't look
right with a flat panel on top of it. Now I wish I hadn't gotten rid of my old
Viewsonic 17" CRT monitor all those years ago! I was sure happy to see it go at the
time, but I failed to predict that I'd get interested in retrocomputing a decade
later.
So, I guess I'll just keep my eyes open for a suitable CRT monitor that I like for
the A3000. Maybe some older Viewsonic monitor in the 13"-17" range? If anybody
near southern California might be able to use my failed 1950 for parts, or has a 1950 with
a different failure that they'd like to dump on me, please let me know! Even though I
don't anticipate being able to fix it, I don't want to throw it away, in case
parts appear or some other collector can make use of it. So, I'll stash its carcass
away for the time being.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
5:22 p.m.
On Dec 1, 2014, at 16:30 , Zane Healy <healyzh at
aracnet.com> wrote:
Are you just driving it off the built in VGA, or do you have a VGA card in it?
Just the built-in VGA. As I understand it, plain old 31 kHz monitors will work with the
flicker-fixer turned on, right?
On Dec 1, 2014, at 16:42 , Geoff Oltmans <oltmansg
at gmail.com> wrote:
Well, if you do replace it with another CRT monitor, don't toss the 1950! Monitors
that sync that low are few and far between, and although the Commodore monitor ain't
the best, it is a brand match and will work with the A3000 as well as the A1200 and A4000
in all unflicker-fixed modes.
Yeah, I'd love to use the brand-matched monitor even though it's not the greatest
monitor. Maybe another parts-donor will come along, and I can build a working monitor out
of the pair?
An oscilloscope would go a long way to helping with
the repair.
It sure would, if I knew what any of the pins on the chip were supposed to be doing!
It's just a 20-pin DIP with no explanation of what the chip's function is in the
circuit, other than what I can infer from circuit topology.
I really don't enjoy working on monitors like this one, or else I might poke around
some more... lots of pointy high-voltage stuff, and very poor circuit accessibility
without reducing it to a pile of loose boards and a tangle of cables next to a naked CRT
rocking on the bench. Pretty much the whole thing needs to be taken to pieces to get to
the solder side of the main board. Grr!
If I could identify and procure that IC, and perhaps a few others in the vertical drive
circuit, I'd happily tear it all apart, swap the chips, and cross my fingers. Hmm,
maybe I should expend some more effort to try to find out if the OEM is still around (said
to be a company called AOC, in Milpitas, CA), and whether anybody there might be able to
identify the IC in question? I think it's a long shot, but maybe I'll get lucky.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
5:37 p.m.
Oh, this is neat... I have an address for the OEM from some old documentation:
AOC International
311 Sinclair Frontage Rd.
Milpitas, CA 95035
Google Earth shows that whatever used to be at that address has now been replaced by
recent-looking residential construction.
I found another possible address online in nearby Fremont, in what is an
industrial/commercial area. Maybe I'll try some phone work during business hours
tomorrow to see if the company still exists.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
5:48 p.m.
Hmm, looks like they might still be around!
AOC International
47490 Seabridge Dr.
Fremont, CA 94538
http://www.aoc.com (Duh!!)
I'll try calling them during business hours tomorrow to see if anybody there might
know something about a 24 year old monitor!
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
4:42 p.m.
On Dec 1, 2014, at 5:44 PM, Mark J. Blair wrote:
So, I guess I'll just keep my eyes open for a
suitable CRT monitor that I like for the A3000. Maybe some older Viewsonic monitor in the
13"-17" range? If anybody near southern California might be able to use my
failed 1950 for parts, or has a 1950 with a different failure that they'd like to dump
on me, please let me know! Even though I don't anticipate being able to fix it, I
don't want to throw it away, in case parts appear or some other collector can make use
of it. So, I'll stash its carcass away for the time being.
Well, if you do replace it with another CRT monitor, don't toss the 1950! Monitors
that sync that low are few and far between, and although the Commodore monitor ain't
the best, it is a brand match and will work with the A3000 as well as the A1200 and A4000
in all unflicker-fixed modes.
An oscilloscope would go a long way to helping with the repair.
9:48 p.m.
I have a Commodore 1950 monitor that came along with an Amiga 3000, and I think it has
defeated me. Its
failure mode is that there's no vertical scan; just a single, horizontal line.
Vertical drive is still working, because I
can shift the line up and down a bit using the vertical centering control, but whatever
circuitry generates the
vertical scan doesn't appear to be doing it.
It is not uncommon for the vertical shift control to apply a DC bias to the yoke, and thus
the output amplifier
stage could still be the problem.
Trying to debug it has been challenging, even though I
found a service manual online. The monitor appears to
Do you have a URL for this manual? I might give it a look and see if I can deduce
anything.
have been designed to be slapped together without much
consideration of service accessibility, and the service
That is quite normal, alas.
manual does't describe the circuitry in much
detail. It includes a block diagram, with no indication of which
actual circuitry corresponds to which blocks. The schematic diagrams show connectivity
between the parts, but
don't identify the functionality of most of the ICs.
I think I've narrowed down to one IC that's a
likely candidate for the failed component, but it's still just an
educated guess that it's even the IC responsible for generating the vertical scan.
And it's a no-name IC with the
monitor manufacturer's internal part number (IC401, with part number 56A326-1), so I
don't think I'm likely to
find a replacement for it unless I find another model 1950 with a different failure.
Older Commodore monitors were Philips inside. I have an idea this 1950 is an AOC chassis,
but I might
be wrong. That IC number doesn't look like a Philips number, certainly, It is quite
possibly a standard part,
house coded. The vertical section of a monitor is much simpler than the horizontal
section, and standard
ICs are common. Often they are also used in TVs (the deflection rate, etc, is the same
after all).
Often the back emf from the vertical deflection coils during retrace is used to power the
output
stage of the vertical amplifier. Look for a diode and capacitor connected to the output
IC. And
check them, they often cause this sort of trouble.
-tony
2 Dec
2 Dec
11:33 a.m.
On Dec 1, 2014, at 21:48 , tony duell <ard at
p850ug1.demon.co.uk> wrote:
It is not uncommon for the vertical shift control to apply a DC bias to the yoke, and
thus the output amplifier
stage could still be the problem.
Thanks, that's good to know. I don't have a lot of experience with analog monitor
circuits yet.
Do you have a URL for this manual? I might give it a
look and see if I can deduce anything.
I found it here:
http://elektrotanya.com/aoc_adc_commodore_1950_cm314_monitor.pdf/download.h…
I'll appreciate any insight you might offer, because I'm not very good at making
sense of monitor circuitry yet. My first suspect is IC401, followed by possibly IC402 or
IC553.
have been designed to be slapped together without
much consideration of service accessibility, and the service
That is quite normal, alas. Older Commodore monitors were Philips inside. I have
an idea this 1950 is an AOC chassis, but I might
be wrong.
You're right: It's an AOC CM314. I'll try calling AOC shortly on the slim
chance that some old-timer might have a dusty pile of documentation in his or her office.
If not, one of the notable surplus dealers on our list has contacted me about possibly
supplying a working monitor. Finally, there's a 1950 on eBay, but I'll look for
cheaper options first.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:45 a.m.
On Dec 2, 2014, at 11:33 , Mark J. Blair <nf6x at
nf6x.net> wrote:
I'll try calling AOC shortly on the slim chance that some old-timer might have a
dusty pile of documentation in his or her office.
Well, it was a long shot, and it missed. The friendly tech support agent at AOC wasn't
able to help me with this 24 year old monitor. At least we had a good chuckle over it. :)
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:50 a.m.
Do you have a URL for this manual? I might give
it a look and see if I can deduce anything.
I found it here:
http://elektrotanya.com/aoc_adc_commodore_1950_cm314_monitor.pdf/download.h…
I'll appreciate any insight you might offer, because I'm not very good at making
sense of monitor circuitry yet.
My first suspect is IC401, followed by possibly IC402 or IC553.
You're right: It's an AOC CM314. I'll try calling AOC shortly on the slim
chance that some old-timer might have a
dusty pile of documentation in his or her office. If not, one of the notable surplus
dealers on our list has
contacted me about possibly supplying a working monitor. Finally, there's a 1950 on
eBay, but I'll look for
cheaper options first.
AOC is 'Admiral of China' IIRC. I remember reading about US-made Admiral radios
and TVs from the
1950s. I guess this is a far eastern company that got the name.
To me IC401 looks like a sync processor/oscillator. IC402 looks like the vertical output
stage.
I think pin 2 of IC402 is the output, pin 4 may well be the drive input (maybe a ramp
voltage
here? D404 and C410 may well be that boost circuit I mentioned and are worth checking
first.
Maybe then change IC402 if you can get one.
-tony
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
2:07 p.m.
On Dec 2, 2014, at 11:50 , tony duell <ard at
p850ug1.demon.co.uk> wrote:
OK, got it... What a confusing manual. It's roughly what I would expect a service
manual to be
(schematics, parts lists, some faultfinding info) but it is not well organised... Perhaps
I am spoilt
by HP and Barco manuals :-)
Good, it's not just me being dense then. :) I'd imagine it makes more sense to
somebody who has seen a lot of monitor circuits, but I have a tough time making sense of
it without some additional hand-holding.
To me IC401 looks like a sync processor/oscillator. IC402 looks like the vertical output
stage.
That's pretty much what my guess was. Some of the components south of IC401 look like
they might be something like PLL filter components, and my guess was that IC401 is
probably responsible for generating the vertical ramp. I didn't probe it beyond the DC
voltage checks mentioned by the troubleshooting chart, though, because it's a bit hard
to get to without major disassembly.
I think pin 2 of IC402 is the output, pin 4 may well
be the drive input (maybe a ramp voltage
here?
I might as well tack on a test point wire or three while it's discharged and then poke
around to see if anything ramp-like is coming out of IC401. I had assumed that the
vertical drive must be fine since I can adjust the centering of the scan line, but I
suppose that the bias circuit to the final drive transistors could still be good with the
previous driver stage being bad. And especially since you have identified a possible cross
for IC402 in your next message, so I might actually be able to replace it if it's
bad!
D404 and C410 may well be that boost circuit I
mentioned and are worth checking first.
Maybe then change IC402 if you can get one.
Sounds like a plan. I still have the fall-back plan of getting another monitor to play
Frankenstein with from Cindy or eBay, but it'll be great if I can fix the monitor that
I already have.
Page 37 of the manual recommended checking for 24VDC at pin 3 of IC402, which also
connects to the junction of D404 and C110. I probed it with my DMM at C410 since that was
easier to get to with the monitor assembled, and I did find 24V there. Next (maybe tonight
if I'm not too tired after work) I'll try probing that probably drive input to see
if IC401 is generating a ramp. If I see a ramp, then I'll try digging up a replacement
for IC402.
If any experienced monitor experts can suggest possible identities for IC401, that would
sure help! It wouldn't surprise me if it's a fairly common part monitor working
incognito in this monitor.
If I do manage to fix this beast, then I should write up what I've learned on my blog
so that maybe the next person trying to fix a similar monitor will stumble across my
notes.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
2:46 p.m.
On 2014-Dec-02, at 2:07 PM, Mark J. Blair wrote:
A couple comments:
- The supply for the vertical drive section comes via a rectifier and filter from the
pins-11/12 secondary of the multi-coil transformer down the page (part of some switching
power supply). I was going to suggest you should probably find around 25V on that supply
line, looks like it's OK as you found 24V at IC402.
- Q402, Q403 form a complementary driver for the 'supply' side of the V coil,
controlled by the vertical centering pot. It looks like they just form a DC driver for the
coil. They would form a sort-of H bridge with the output transistors of IC402.The V around
the junction of the Q402,Q403 emitters (see also R414,R413,C422) should probably be around
the mid-point of the supply, +12V or so, and should vary with the V centering control.
(The objective is so the IC402 output can drive current in both directions through the V
coil so you get both up & down deflection.)
If, for example, Q402, R414 or R413 failed open there would be no current supply for the V
coil. You said you were getting some variation with the V centering, so, yes, the problem
may well be elsewhere.
On Dec 2,
2014, at 11:50 , tony duell <ard at p850ug1.demon.co.uk> wrote:
...
I think pin 2 of IC402 is the output, pin 4 may
well be the drive input (maybe a ramp voltage
here?
I might as well tack on a test point wire or three while it's discharged and then
poke around to see if anything ramp-like is coming out of IC401. I had assumed that the
vertical drive must be fine since I can adjust the centering of the scan line, but I
suppose that the bias circuit to the final drive transistors could still be good with the
previous driver stage being bad. And especially since you have identified a possible cross
for IC402 in your next message, so I might actually be able to replace it if it's
bad!
D404 and C410 may well be that boost circuit I
mentioned and are worth checking first.
Maybe then change IC402 if you can get one.
Sounds like a plan. I still have the fall-back plan of getting another monitor to play
Frankenstein with from Cindy or eBay, but it'll be great if I can fix the monitor that
I already have.
Page 37 of the manual recommended checking for 24VDC at pin 3 of IC402, which also
connects to the junction of D404 and C110. I probed it with my DMM at C410 since that was
easier to get to with the monitor assembled, and I did find 24V there. Next (maybe tonight
if I'm not too tired after work) I'll try probing that probably drive input to see
if IC401 is generating a ramp. If I see a ramp, then I'll try digging up a replacement
for IC402.
If any experienced monitor experts can suggest possible identities for IC401, that would
sure help! It wouldn't surprise me if it's a fairly common part monitor working
incognito in this monitor.
If I do manage to fix this beast, then I should write up what I've learned on my blog
so that maybe the next person trying to fix a similar monitor will stumble across my
notes.
2:54 p.m.
On Dec 2, 2014, at 14:46 , Brent Hilpert <hilpert
at cs.ubc.ca> wrote:
If, for example, Q402, R414 or R413 failed open there would be no current supply for the
V coil. You said you were getting some variation with the V centering, so, yes, the
problem may well be elsewhere.
That's about what I was thinking. The horizontal line is near the center of the
screen, and can be moved up and down with the centering control. If just one side of the H
bridge was broken, then I'd expect to see a badly distorted raster on one half of the
screen. So, that's why I suspect that the ramp isn't getting generated. I'll
do some more poking around to see if any signal is being fed to IC402 by IC401.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
9:09 p.m.
Well, I feel pretty dumb now. IC401 is marked "LA7850" on top with no
manufacturer logo. Mr. Google suggests it's a Sanyo LA7850 sync chip, and those are
available on eBay. So, things are looking better, and I gave up way too soon!
The LA7850 datasheet suggests that IC402 may be an LA7832, LA7833, LA7837 or LA7838. I
can't see the chip markings yet, but I should be able to see them once I get the
mainboard out to add some test point wires onto the suspicious IC401. It's tucked
under the CRT, with the side that I need to probe facing in the least convenient
direction, of course. :)
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:02 p.m.
I pulled the main board out in order to solder some test points onto IC401 pins 15
(vertical drive output) and 12 (horizontal drive output). While I had it out of the
monitor, I could see that IC402 is a Sanyo LA7830.
With it back together, here's what I see on the drive outputs of IC401. Trace 1
(yellow) is the vertical drive, and trace 2 (blue) is the horizontal:
https://www.dropbox.com/s/m7a97lypdqknofx/TEK00000.PNG?dl=0
https://www.dropbox.com/s/0c80wuqkbr245n7/TEK00001.PNG?dl=0
The vertical drive looks pretty dead, and the horizontal drive doesn't look anything
like what I expected.
Now, I took these measurements with no video input to the monitor, with the expectation
that the drive oscillators should free-run on an analog monitor this old. Is that a
reasonable expectation?
I think that IC401 is pining for the fjords, so I'll order at least one replacement
and a DIP socket to stick it in. Maybe I should order extra chips, and one or two
replacements for IC402 as well, and tuck them inside the case to help out the next person
to repair this monitor? :)
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:58 a.m.
On Tue, Dec 2, 2014 at 2:33 PM, Mark J. Blair <nf6x at nf6x.net> wrote:
Yes, and it's a big part of the reason that I haven't been motivated to
get much experience fixing monitors so far! I had previous bad experience
trying to fix a 19" Hitachi monitor for my Sun 3/60 20ish years ago
(thought I had it narrowed down to a $70 transistor; replacement blew, too;
ended up taking it to a shop and paying real money for somebody with a clue
to fix it).
I did have much better experience with a DEC VR201 recently. Its circuitry
was a lot easier to get to for maintenance, and it had a simple failure
mode: Blew its fuse immediately. So I powered it through a current-limited
supply and started looking for hot spots with my IR thermometer. Hottest
spot was a diode feeding power to a subcircuit, and second-hottest spot was
a capacitor from that power rail to ground. Replaced the cap, monitor
became happy.
When I was in high school and was just coming up in the field, I had the
opportunity to work at a local computer repair shop in the area that
maintained the ability to do board-level repair (ca. late 90s). There was a
guy there who was just brilliant at repairing monitors... I imagine he had
been doing it for a while so he probably had a good idea of all the tricks
of the trade and things to look out for but I was always impressed... I
felt like there was maybe a bit of art to the whole process. I don't recall
there ever being a monitor he couldn't fix given parts availability and the
budget for T&M.
Certainly watching him do it emboldened me, LOL. I remember taking on a
totally generic 15" VGA display from a trash heap some time ago and I got
lucky... easy fix... cold solder joint or a burnt out resistor or something
that you could just pick up visually. Sometimes it turns out to be easy
stuff! Too bad not in this case!
I have a Sony GDM17 that's been sitting torn apart on my bench forever that
needs a new cable spliced onto it, and one of my NeXT Megapixel Color
Displays has pooped out on me so that will have to go under the knife
someday. Not to mention my VT100! Yikes. I definitely enjoy the discussion
of diagnosis & repair of the analog side of things. Many things to learn!
Best,
Sean
3671
days inactive
3673
days old
16 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
healyzh＠aracnet.com -
hilpert＠cs.ubc.ca -
nf6x＠nf6x.net -
oltmansg＠gmail.com -
scaron＠umich.edu