SX-64 Troubleshooting Help
Roy Stout
Worked great for about 4 hours. :-((
No doubt its been powered off for some time, and my 4 hour "burn-in"
must have cooked something.
On power up, the drive light comes on and goes off as usual.
The screen brightens up, but no familiar **** Commodore 64 Basic V2 ****
No blue background or border, just a white screen- edge to edge.
Powering on with a cartridge inserted does not work either - same
symptons, drive light goes on then off, screen comes on, but is blank.
Typing blind, "Load"$",8" [Return] does not cause the drive to read the
disk.
I've read from the various troubleshooting documents that the culprits
could be: 6510, 6567, PLA, 8701, RAM. I'm leaning towards the 6510.
Can anyone offer any suggestions to narrow the search?
Are there logic signals I can check to eliminate one of these off the
hit list?
Or am I stuck with the old "process of elimination", replacing one at a
time until it starts working?
Thanks,
Roy
Woodsgene
It turned out to be the one of the two 6526
needed to be reseated. Gene!
White Flame (aka David Holz)
news:20020223082705...@mb-fe.aol.com...
> I had a every similar Problem with my SX
> It turned out to be the one of the two 6526
> needed to be reseated. Gene!
IIRC, the 6526s are on a daughterboard that's connected with a pin connector
instead of an edge connector. That can be a flaky and unreliable
connection, so reseating the whole daughterboard (possibly bending the pins
a bit so the connection's more snug) could help out as well.
--
White Flame (aka David Holz)
http://fly.to/theflame
(spamblock in effect)
mykrowyre
the pins aligned correctly.
If that doesnt do it, replace the PLA. I had the same problem with my SX and
I replaced the PLA, and then the VIC to solve the problem. She works great
now.
I love my sx. its just soooo coool!
-tom
"White Flame (aka David Holz)" <whitef...@y.a.h.o.o.com> wrote in message
news:a5883e$1qse$1...@barad-dur.nas.com...
Raymond Carlsen
> brightens up, but no familiar **** Commodore 64 Basic V2 **** No blue
> background or border, just a white screen- edge to edge.
Roy,
Do you get that white screen only if you turn the brightness up on the
monitor? Blank (just dim "raster") is a different symptom than a white
(bright raster) screen. The former can be caused by many different chips,
whereas the latter points to the master oscillator not running. There is a
schematic on www.funet.fi/pub/cbm/ and an SX repair article on my site:
http://landru.myhome.net/rcarlsen
> Powering on with a cartridge inserted does not work either - same
> symptons, drive light goes on then off, screen comes on, but is blank.
Can't learn much from that except that the power supply is working.
> Typing blind, "Load"$",8" [Return] does not cause the drive to read the
> disk. I've read from the various troubleshooting documents that the
> culprits could be: 6510, 6567, PLA, 8701, RAM. I'm leaning towards the
> 6510. Can anyone offer any suggestions to narrow the search? Are there
> logic signals I can check to eliminate one of these off the hit list? Or
> am I stuck with the old "process of elimination", replacing one at a time
> until it starts working?
If you have a scope or logic probe, it could narrow it down somewhat, but
the main problem is that many chips share data lines. If a line is being
pulled down (below 4VPP) because of a bad chip, you still need to isolate
it down to the bad one. I usually start by swapping any socketed chips
beginning with the PLA, the most failure prone IC in there. It runs -very-
hot normally as does the SID. After a 5 minute warmup, the RAM chips should
be barely warm (never hot), the heat-sinked VIC runs very warm, the MPU
runs hot, the CIAs should be warm (not hot), etc. For test purposes, the
computer will come up without a few chips plugged in such as the SID and
CIAs, but you need to know what to expect. The repair article explains
that. As someone else mentioned, the interconnects between boards in the SX
sometimes create problems and should be reseated along with any socketed
chips.
Ray
Commodore repair articles: http://landru.myhome.net/rcarlsen/cbm.html
and mirrored at http://staff.washington.edu/rrcc/uwweb/articles
CBM monitor schematics: http://staff.washington.edu/rrcc/index.html
and -lots- more info at http://www.funet.fi/pub/cbm/
Software and hardware sales:
http://www.centsible.com (Centisble Software)
http://www.vintagecomputer.com
http://www.aye.net/~feelgood (Dr. Feelgoods Crazy C= Sales)
http://www.oldsoftware.com
http://www.jspusa.com/products.html (chips)
Paul Rosenzweig
Have you tried plugging an external monitor into the SX's audio / video
port. Somebody I know gets a black blank screen on his powered
up SX 64 but gets something meaningful on an external monitor.
Cameron Kaiser
>Have you tried plugging an external monitor into the SX's audio / video
>port. Somebody I know gets a black blank screen on his powered
>up SX 64 but gets something meaningful on an external monitor.
Make absolutely sure you plug in the monitor cable with the SX powered OFF.
I've fried two SIDs myself this way, and I've heard it happening to others,
too.
--
Cameron Kaiser * cka...@stockholm.ptloma.edu * posting with a Commodore 128
personal page: http://www.armory.com/%7Espectre/
** Computer Workshops: games, productivity software and more for C64/128! **
** http://www.armory.com/%7Espectre/cwi/ **
Roy Stout
>> On power up, the drive light comes on and goes off as usual.
>> The screen brightens up, but no familiar **** Commodore 64 Basic V2
****
>> No blue background or border, just a white screen- edge to edge.
>> Powering on with a cartridge inserted does not work either - same
>> symptons, drive light goes on then off, screen comes on, but is blank.
>> Typing blind, "Load"$",8" [Return] does not cause the drive to read
the
>> disk.
>>
>
>port. Somebody I know gets a black blank screen on his powered
>up SX 64 but gets something meaningful on an external monitor.
Good suggestion, unfortunately the external monitor is blank as well. Thanks
for the feedback.
Roy
Roy Stout
Raymond Carlsen wrote in message ...
>> On power up, the drive light comes on and goes off as usual. The screen
>> brightens up, but no familiar **** Commodore 64 Basic V2 **** No blue
>> background or border, just a white screen- edge to edge.
>
>Roy,
> Do you get that white screen only if you turn the brightness up on the
>monitor? Blank (just dim "raster") is a different symptom than a white
>(bright raster) screen. The former can be caused by many different chips,
>whereas the latter points to the master oscillator not running. There is a
>schematic on www.funet.fi/pub/cbm/ and an SX repair article on my site:
>http://landru.myhome.net/rcarlsen
>
>If you have a scope or logic probe, it could narrow it down somewhat, but
>the main problem is that many chips share data lines. If a line is being
>pulled down (below 4VPP) because of a bad chip, you still need to isolate
>it down to the bad one. I usually start by swapping any socketed chips
>beginning with the PLA, the most failure prone IC in there. It runs -very-
>hot normally as does the SID. After a 5 minute warmup, the RAM chips should
>be barely warm (never hot), the heat-sinked VIC runs very warm, the MPU
>runs hot, the CIAs should be warm (not hot), etc. For test purposes, the
>computer will come up without a few chips plugged in such as the SID and
>CIAs, but you need to know what to expect. The repair article explains
>that. As someone else mentioned, the interconnects between boards in the SX
>sometimes create problems and should be reseated along with any socketed
>chips.
>
>Ray
>
I do remember having to turn the brightness up to see that the monitor
was in fact on. When the unit first failed to start up, the screen was black
blank. Then I adjusted the brightness up to get the white appearance.
I do have a logic probe, but first I will use the "burnt finger"
approach to isolate chips that are not getting as hot as they should. Also,
thanks for the links to the repair sites. I'll browse them too.
With this additional information, do you have any additional
suggestions?
Thanks!
Roy
Nicolas Welte
> Do you get that white screen only if you turn the brightness up on the
> monitor? Blank (just dim "raster") is a different symptom than a white
> (bright raster) screen. The former can be caused by many different chips,
> whereas the latter points to the master oscillator not running. There is a
Wow, I didn't know this! Actually I have an SX that sometimes shows a white
screen when it wasn't powered up for a longer time, but works fine after it
got warm. When it does it next time I'll test the dotclock on the expansion
port, because right now the machine works just fine :)
Nicolas
mykrowyre
-tom
Roy Stout
appear to work fine.
Robert Brenner's "Commodore 64 Troubleshooting and Repair Guide"
suggests the 74LS629 (U31) (a dual voltage controller oscillator) or a bad
crystal.
I cannot find a 74LS629 on the SX-64, but could it be one of the ICs on
a daughterboard labeled DOT Clock and COLOR. The daughterboard is soldered
at points to the 251102 Rev A circuit board and also contain 2 crystals.
One is a 74LS74A - dual D flip-flop, a 74LS04P - a hex inverter, and a
74LS161AN - 4-bit counter. None sound like a 74LS629.
I'll take some logic readings on all 3 just to make sure there is some
activity there. And I'll check to crystals.
Any suggestions where I might find the "dual voltage controller
oscillator".
Thanks!
Roy
Woodsgene wrote in message <20020223082705...@mb-fe.aol.com>...
Roy Stout
appear to work fine.
Robert Brenner's "Commodore 64 Troubleshooting and Repair Guide"
suggests the 74LS629 (U31) (a dual voltage controller oscillator) or a bad
crystal.
I cannot find a 74LS629 on the SX-64, but could it be one of the ICs on
a daughterboard labeled DOT Clock and COLOR. The daughterboard is soldered
at points to the 251102 Rev A circuit board and also contain 2 crystals.
One is a 74LS74A - dual D flip-flop, a 74LS04P - a hex inverter, and a
74LS161AN - 4-bit counter. None sound like a 74LS629.
I'll take some logic readings on all 3 just to make sure there is some
activity there. And I'll check to crystals.
Any suggestions where I might find the "dual voltage controller
oscillator".
Thanks!
Roy
mykrowyre wrote in message ...
Roy Stout
Roy Stout wrote in message ...
> I've tested both of the 6526 (and the 6510 and 6567) in a C64 and they
>appear to work fine.
> Robert Brenner's "Commodore 64 Troubleshooting and Repair Guide"
>suggests the 74LS629 (U31) (a dual voltage controller oscillator) or a bad
>crystal.
> I cannot find a 74LS629 on the SX-64, but could it be one of the ICs on
>a daughterboard labeled DOT Clock and COLOR. The daughterboard is soldered
>at points to the 251102 Rev A circuit board and also contain 2 crystals.
> One is a 74LS74A - dual D flip-flop, a 74LS04P - a hex inverter, and a
>74LS161AN - 4-bit counter. None sound like a 74LS629.
> I'll take some logic readings on all 3 just to make sure there is some
>activity there. And I'll check to crystals.
> Any suggestions where I might find the "dual voltage controller
>oscillator".
>
>Thanks!
>Roy
>
both legs of the 14mhz crystal go low and stay there. Whereas on the C64
only one leg goes low and the other does not register anything on the probe,
but when I touch it, the display flickers and distorts a little. Could this
be a bad crystal, or could something else be pulling the crystal low?
hmmmmm
Thanks,
Roy
Raymond Carlsen
> just a short question about terminology: Scope is oscilloscope, ie. the
> thing with the square screen and the many buttons to adjust speed, range
> etc. for the display, that looks like a smaller SX-64, only with much
> more buttons and w/o a keyboard and floppy drive of course, right?
Paul,
Correct. It is more like a television or monitor than anything else.
It has inputs that pick up and amplify signals from electronic devices of
all types. These signals deflect the beam on the tube face to produce a
graphical representation of the variable voltages... a bit like a bar
chart. A 'scope has variable instead of fixed scanning frequencies to test
low and high frequency circuits.
> Then what is a logic probe? Sorry, English is not my mother tongue. :-/
A logic probe is (was) a "pencil" with internal circuitry inside to detect
AC signals on data lines, and flash an LED or sound a "beep". Probes were
popular back in the 1970s when TTL (standard logic devices ie 74xxx series
chips running at 5 volts) was just about all there was. It was a cheaper
alternative to an oscilloscope that anyone could afford, and was a quick
go/no-go test device. I don't suppose they are used much any longer because
of the changes in technology. Such a device would still be useful for
troubleshooting inside our computers though. In most cases, the signals are
either there or not, so a probe would still be useful in that application.
Ray
Raymond Carlsen
> was in fact on. When the unit first failed to start up, the screen was
> black blank. Then I adjusted the brightness up to get the white
> appearance.
Roy,
OK, the master oscillator (a module in the SX) is probably working.
> I do have a logic probe, but first I will use the "burnt finger" approach
> to isolate chips that are not getting as hot as they should. Also, thanks
> for the links to the repair sites. I'll browse them too. With this
> additional information, do you have any additional suggestions?
I would be going after the most failure prone chips in there, the PLA
for starters. It's the same as the one in a C64. If any chip should have
been heat sinked, it's that one... and the SID, and the MPU... well, you
know. ;-) Blank screen is the most frustrating symptom because any one of
a dozen chips can produce that symptom. You can't always go by temperature
to find bad chips, but if any that are supposed to run cool (CIA, RAM,
etc.) are getting hot, they're bad. If any that normally run warm are stone
cold after 5 minutes... you get the idea. Troubleshooting is grunt work for
sure. Some of us have a perverse idea of "fun". I do it for a living. :-S
Ray
Raymond Carlsen
Nicolas,
You get the same white screen symptom if the VIC is completely dead,
but failing chips usually work when cold, and barf after warmup. If I had
an intermittant like that, I would be reseating all the connectors and
socketed chips just for luck. Bet that's it.
Ray
Raymond Carlsen
> both legs of the 14mhz crystal go low and stay there. Whereas on the C64
> only one leg goes low and the other does not register anything on the
> probe, but when I touch it, the display flickers and distorts a little.
> Could this be a bad crystal, or could something else be pulling the
> crystal low?
Roy,
The capacitance of the probe may be loading the circuitry down so it
appears there is no signal. That's where a scope is better, but it sounds
like the oscillator is running, at least. Good enough. Make sure -all- data
lines are active (pins 30 through 37 of MPU). If one is low or absent, that
will produce a blank screen. Since the MPU is the heart of the system, see
if it's getting what it needs to work. AEC pin 5 should have a clock signal
(generated by VIC), RDY pin 2, IRQ pin 3, and NMI pin 4 should all be logic
high (+5V) 3 seconds after powerup.
Turn the volume up on the monitor and power up the computer. The
strange noise you hear can sometimes be used for troubleshooting, but of
course you need to know what it sounds like when it's working. As the drive
LED comes on and goes out (about 3 seconds), there will be one kind of
"whine", and it will change as the opening screen appears. If that sound is
normal, I suspect a failing VIC (no video output). If the whine never
changes, it can be just about anything... PLA, VIC, RAM, etc. RAM problems
are not as common in the SX as in the C64 (better power supply).
Ray
Roy Stout
And thanks to everyone else who contributed advice.
Roy
Roy Stout wrote in message ...
Ditdat
Do you have a procedure for adjusting the convergence on an SX-64? I
received a nice unit in trade a few months back, and the only real
problem is a bit of misconvergence causing the displayed text to
appear a bit fuzzy.
Thanks,
Rick
Paul Foerster
> A logic probe is (was) a "pencil" with internal circuitry inside to detect
> AC signals on data lines, and flash an LED or sound a "beep".
... ah, ok. Thanks very much for enlightening me. I wonder when I will ever
get all the English terms straight. :-)
--
cul8er,
Paul
oo
pa...@gmx.net ~( "> paul_f...@s.maus.de
Raymond Carlsen
> > detect AC signals on data lines, and flash an LED or sound a "beep".
> ... ah, ok. Thanks very much for enlightening me. I wonder when I will
> ever get all the English terms straight. :-)
Paul,
Ever the perfectionist, I should add something I overlooked in my
statement. A logic probe obviously must also be able to detect a logic high
or low in addition to sensing data on a line. The high or logic 1 must be
over a certain level (+2.4V IIRC) and low or logic 0 must be under +0.8V to
be "valid". Any DC voltage "floating" between those two points was
considered noise and was called a "disallowed state" in logic terms. This
goes only for old TTL chips. Newer equipment runs on lower voltages and at
much faster speeds, so a scope or some other type of device must be used to
test those circuits. More than anyone wanted to know. :-)
Ray
mykrowyre
I have the same problem...also interested in a solution.
-tom
Raymond Carlsen
> > Rick
Tom/Rick,
Here is your answer... long winded I know, but knowing why is at least
as important as knowing how, especially if something doesn't work as
expected. This info is generic and therefore good for just about any
monitor. Good luck and be careful!
SETUP ADJUSTMENTS FOR INLINE CRT'S
Advances in technologies along with cost-cutting measures by all
manufacturers have produced a display tube that needs less support hardware
and fewer setup adjustments than earlier dot-matrix types. The most common
Cathode-Ray Tube (CRT) now has three inline (side by side) electron "guns"
rather than the triad arrangement of the earlier types. Setup adjustments
are done by moving several sets of rings on the neck of the tube. These
rings are made of magnetized material to bend the electron beams to the
desired locations on the tube face. The two major alignments are screen
"purity" and "static beam convergence". The term "static" simply means beam
correction with magnets rather than with electronic circuitry. The latter
is usually referred to as "dynamic" convergence and is only used in
high-end monitors and large screen TV sets.
There are three pairs of rings, each designed for a specific
adjustment. The ones closest to the yoke (the large deflection coil
assembly, mounted near the bell of the tube) are for purity alignment.
Purity adjustments are necessary only if a white screen shows "blotches" of
color. The next pair of rings are a four pole type that controls the RED
and BLUE static convergence. The last pair of rings are 6 pole to control
the static alignment of MAGENTA (RED + BLUE) with GREEN. Behind that pair
of rings there is often a locking device. Manufacturers usually put a line
of glue across the assembly, and/or there is an ink line drawn across all
the rings to indicate their position after factory alignment. That's useful
as a "marker" in case you get "lost" when attempting realignment. Each ring
has tabs that sticks out to allow for adjustment with your fingers.
A picture tube is said to be in convergence when all three beams
(primary colors of RED, GREEN and BLUE) overlap. Misconvergence shows up as
color "fringing" around the edges of objects anywhere on the screen. The
convergence adjustments are normally "roughed in" before purity is adjusted
because they interact with each other somewhat. Unless someone has been
"diddling" and has it completely out of alignment, coarse settings will not
change and you can just "tweek" the convergence to optimise it.
All ring pairs share one thing in common: when the adjustment
tabs are set together (aligned with each other), the magnetic fields of the
two rings cancel each other out. If those rings are rotated together,
nothing happens. If a picture tube and yoke were perfect from the factory,
they would need no correction, but that never happens. When pairs of rings
are offset with respect to each other, a magnetic field is generated in a
specific direction relative to the tabs. One convergence adjustment is done
by splitting or separating the tabs of a pair of rings in the necessary
direction. That moves the beams in a horizontal direction and so affects
vertical lines on the screen. The other adjustment is done by rotating both
rings together, and that moves the beams in a vertical direction, affecting
horizontal lines. Of course these adjustments interact with each other, so
you must go back and forth between the two settings for optimal
convergence. It takes very little movement of the rings to affect the image
you see on the screen. All this may seem confusing at first, but it becomes
clear when you see for yourself how movement of the rings changes the
screen.
To be able to see the effect adjustments are having on the tube face,
test patterns are recommended, the most common being a "crosshatch" pattern
of a dozen or so horizontal and vertical white lines on a black background.
Some techs are more comfortable with a dot pattern, but it doesn't really
matter. A screenful of zeros on the monitor will work as well. The color
control must be turned down so there is no residual color on the screen to
confuse the readings.
IMPORTANT FIRST STEP!!!
If a lock is used on the convergence ring assembly, it needs to be
unscrewed (rotated counter-clockwise, looking at it from the rear of the
tube) to unlock it. Otherwise you will break the rings attempting to move
them. A locking ring will be similar in appearance to the adjustment rings
and will have tabs for "adjustment", but it will be thicker than the others
and will be the last one on the stack towards the rear of the tube. Back
off the lock ring about one-half a turn. If it is left too loose, the
adjustment rings will slip out of adjustment easily while you are
working... too tight and they will not move. If there are any wires around
the assembly, they must be moved out of the way so they don't snag on any
of the ring tabs underneath. The glue (if it was used) on the rings holds
them even if the lock is released, so you need to break the glue bonds to
be able to make adjustments. Use a sharp knife or flat blade screwdriver to
gently pry between rings to free them from the assembly and from each
other. Don't scrape the glue off. It's useful as a marker to see where the
adjustments were... just in case. If there is no glue or marks of any kind,
draw one with a magic marker or felt tip pen across the entire assembly
before you begin.
Since you will be working on a live chassis, it is assumed you already
know your way around inside electronic equipment and will observe the
necessary safeguards.
COLOR PURITY ADJUSTMENTS
Screen color purity is normally the first CRT setup adjustment to be
made, but it doesn't normally drift very much over time and therefore
doesn't require routine adjustment. Purity is said to be good if, when each
color is turned on and the other two turned off, the entire screen is all
one color. Bad purity will show up as wrong colors on what should be a pure
color field, or color blotches on what should be a white screen (all three
guns turned on). If you can already get a good white screen, don't fool
with the purity adjustments. But, if you must...
There are internal controls for each CRT "gun" to set the brightness
level for that color. To adjust green screen purity (the center gun and the
most common one to view during purity setup adjustments), turn down the red
and blue guns with their respective controls and/or turn up the green. If
you can, somehow mark the position of the controls beforehand so you can
put them back when you're finished. Each TV or monitor has it's own
terminology for these controls. Some call them "low light", others "screen"
or "G2" or "cut off", and still others "bias". These controls need to be
reset to their proper points when purity adjustments are completed or the
resulting offset "white balance" will show as "tinting" of one color over
the entire screen when the brightness is turned down.
An alternate way to do the setup without having to adjust any CRT
level controls is with a generator or via a program in the computer feeding
the monitor. For example, a simple BASIC program can be written to set
forground and background color to GREEN only. Patterns of lines or dots can
be likewise generated in a program.
To adjust purity, set up a green screen as indicated above. Loosen the
yoke clamp and pull the yoke back towards the convergence assembly as far
as it will go. If there is any glue or tape holding it, it may be necessary
to gently twist the yoke to break it loose from the tube. A wide vertical
bar of solid green should appear in the center of the screen. If it's off
center, the purity rings should be rotated and/or separated to center the
green bar. Then slide the yoke forward to get an overall green screen
without contamination by red or blue. Use a bar or crosshatch pattern to
make sure the yoke is straight (rotational misalignment will cause the
picture to be tilted), and then gently tighten the yoke clamp. It's a good
idea to check the red screen purity, then blue. Slight adjustments of the
rings or yoke position may be required to optimize the purity of each of
the three color fields. Some compromise may be necessary, but it's usually
not important to get it perfect. It has to be pretty far off to show up on
a white screen.
If you turned any of the level controls to do the purity adjustments,
reset them to their original spots. To check for proper "grey scale" or
"CRT tracking", turn down the brightness control and see if the white
screen goes to any color as it darkens. If it does, adjust the level
controls a bit to get a neutral grey dark screen.
CONVERGENCE
As stated before, the purity rings are the first set on the
convergence stack behind the yoke. Don't adjust those unless you have to.
The second pair of rings is used to converge the RED and BLUE lines at the
center of the screen. Separating the rings will move the beams horizontally
(side to side), so for that adjustment, you must look at the vertical lines
of the crosshatch to see the effect. Holding and moving both rings together
moves the beams vertically. To see that effect, you need to watch the
horizontal lines at the center of the screen.
The third set of rings set the convergence of MAGENTA (RED+BLUE) AND
GREEN. AS before, Looking at the center of the screen, separating the rings
allows for horizontal beam movement (observe vertical lines) and rotating
both rings together moves the beams vertically (observe horizontal lines).
If there is any misconvergence at the screen edges, you must tilt the
yoke up or down, or side to side for compensation while observing a
crosshatch pattern on the screen. Smaller tubes don't normally require
adjustment, but larger screen sizes can show quite a bit of misconvergence
at the edges. When adjustment is optimised, small rubber or plastic blocks
are used to hold the yoke in place. They are fitted with double-sided tape
or glue to keep them in position between the yoke and CRT bell.
Ray Carlsen CET
CARLSEN ELECTRONICS... a leader in trailing-edge technology.
Questions or comments are welcome, especially if you spot a mistake here.
Thanks!
email: rcar...@myhome.net or rr...@u.washington.edu
Ditdat
<rr...@u.washington.edu> wrote:
>Tom/Rick,
> Here is your answer... long winded I know, but knowing why is at least
>as important as knowing how, especially if something doesn't work as
>expected. This info is generic and therefore good for just about any
>monitor. Good luck and be careful!
Outstanding! Thanks Ray, your expertise is much appreciated!
Rick