Rasterlist Archives: Re: Breakout PCB & Motorola monitor

From: Rodger Boots <rlboots_at_cedar-rapids.net>
Date: Thu Sep 28 2000 - 03:27:28 EDT

<x-html>
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
There is one other thing to try.  You said the regulator was running
quite hot, but you didn't say if the big resistor was running hot (should
run hotter than the regulator).  If the resistor was bad the regulator
would run extra hot and might be causing the humbars due to overload.
Also, I had worked on some Atari games years ago that had the wrong
resistor installed.  This caused the regulator to be overloaded.
ALSO, that resistor, depending on manufacturer, may have two windings
in it and one might have opened.  Check it with an ohmmeter (be sure
to touch the test leads together to measure lead resistance and subtract
that from the resistor reading to get the REAL resistance reading).
  
Paul Sommers wrote:
<blockquote TYPE=CITE> 
Rodger,
Tried the second transformer
to no avail. It made no difference what so ever.
Here's my strategy now -
Rip out the 7805 and the big
resistor and wire a switcher to feed +5.
Breakout still needs the AC
to feed lights and sound, so I'll leave that connected - but I'll just
bypass the regulator with +5v.
If that doesn't fix the problem
- I don't know what I'll do.
Thanks for the help. I'll 
let you know how I go.
Paul
<ul>----------
 From:  
Rodger Boots
 Reply To:      
rasterlist@synthcom.com
 Sent:  
Tuesday, 26 September 2000 17:45
 To:    
rasterlist@synthcom.com
 Subject:       
Re: RASTER: Breakout PCB & Motorola
monitor - questions for everyone!!!!
If I remember correctly, the H series of voltage
regulators were good for 5 amps.  No need to change it.  You
might try using BOTH the 8000 AND the pair of 4700 uF capacitors together
to minimize the valleys of the filtered DC ripple.
OR you could rewind the existing transformer, if
it isn't currently running very hot.  (I'm not necessarily joking
about this, but don't try it unless you think you're up to the challenge).
If there is still room on the transformer for additional
turns (not usually).  On a large transformer they might be getting
out as much as a volt per turn, so we aren't talking much here.
Another possibility is to boost the primary voltage
to the transformer.  You want 16.5, you have 15.  That's a boost
of 16.5*100/15 = 110% of the current line voltage.  So take a small
(at least one amp) 12 volt transformer, connect its primary to the line
and the secondary in series with your large transformer's primary. 
If you get the phasing wrong it will further reduce the output of your
large transformer.  If you get it right, though, the output should
slide right up where you need it.
  
  
Paul Sommers wrote:
<ul> 
Rodger,
I pretty much came to the same
conclusion last night. Here's what I did...
Turned it on with no game board
- there was small ripples on  screen edges (no sync or signal) but
no humbars.
I replaced the old 8000uf with
2x4700uf. No difference.
I went back and looked at the
schematic for the paddle problem. I then tried adding a 1uf lytic between
pin 5 of the 555 and GND. There is still paddle flutter, but it is constant
across the screen at about a half inch rather than up to 2 inches on the
extremes.
I then tried a 10uf in the
same spot but it went back to the big paddle flutter. I suspect the 10uf
was big enough to affect the timing.
There was nothing left to look
at except the voltage regulator - UA78H05S (of which I can find no info).
Checking against the schematic this is not the original - it should be
a LM323 which is rated at 3amps. All the 7805's that I could find were
rated at 1 amp or 500ma.
As I can't get a LM323 in Australia
I started to look on the web for an equiv and that's when it hit me.....
The original transformer put
out 16.5vac to the board. The transformer I was able to find is only 15vac.
I thought it would be enough as most voltage regulators have a bit of tolerance.
Checking the 5v at the chips
gave me 4.94v... pretty close.
I did a check of the local
electronic distributors catalogue and they have one at 18vac. I'm thinking
I should try this. The spec sheet on the common 7805 and the LM323 says
that it should handle both voltages, but maybe it doesn't have enough grunt.
But now looking at your mail
I could run into the problem if I go too high as well....
No-one has a 16.5vac transformer.
Why on earth did Atari used such weird voltages????
Thanks for working with me
on this one.
Cheers
 Paul
  
        ----------
 From:  
Rodger Boots
 Reply To:      
rasterlist@synthcom.com
 Sent:  
Tuesday, 26 September 2000 03:16
 To:    
rasterlist@synthcom.com
 Subject:       
Re: RASTER: Breakout PCB & Motorola
monitor - questions for everyone!!!!
        You know,
I didn't really think about it, but the problem might be the replacement
transformer!  Check your +5 volts on the game board and see if it
is too high or too low.  Voltmeter OK, but scope is even better. 
None of this applies if you have only ONE humbar.
        Possibility
one, too low.  In this case either the capacitor is bad OR the transformer
voltage is too low, causing the regulator to drop out of regulation.
        Possibility
two, too high.  In this case, since this is an early Atari board,
the voltage peaks (if high enough) can cause the regulator to quit regulating. 
What happens is these old boards run most of their current through a large
resistor connected across the regulator with the regulator picking up the
last little bit of current.  Otherwise the regulator will overheat
and shut down.  Now, IF the transformer voltage were too high, the
regulator current would drop to zero and ALL the current would go through
the resistor.  This would happen during voltage peaks and cause the
+5 volts to increase (but only during the peaks).
        What got
me thinking about all this is that the problem didn't start until the transformer
got changed.  If you do have a high voltage condition and you are
lucky enough to have a higher voltage primary transformer setting available
(example, you currently are using the 110 volt tap and switch to using
the 120 volt tap) you can fix it that way.
  
  
  
        Rodger
Boots wrote:
<ul> 
        Paul Sommers
wrote:
<ul>               
The 8000uf at 80v won't fit on
the board (the one that is their is surface mount) so I'll get a couple
of big ones in series - that should do it. BUT, before we get too
carried away with this, I have a question.  Is there ONE hum bar or
TWO?  Very simple rule here:
                   
ONE hum bar = bad rectifier (or fuseholder in series with rectifier)
     TWO hum bars = filter capacitor
problem.
        That's
a good tip to remember. No hum bars. It's a continuous wave.
        Imagine
a grayscale light to dark - then put another on with the light side butted
to the dark edge. There is no break - it just keeps coming. It's a vertical
monitor - so it is going across the screen(or down/up if it was horiziontal).
I guess if two hum bars mean a filter cap problem - 8 could mean a bad
filet cap problem.
  
  </ul>
        OK, you're
saying you have TWO dark waves?  Or two greyscales?  Whatever,
if there are two of anything the capacitor (8,000 uF) is bad.</ul>
        --
 Windows:
        32 bit
graphical interface for a
 16 bit patch for an
 8 bit operating system written for a
 4 bit processor by a
 2 bit company that can't stand
 1 bit of competition.
  </ul>
--
 Windows:
32 bit graphical interface for a
 16 bit patch for an
 8 bit operating system written for a
 4 bit processor by a
 2 bit company that can't stand
 1 bit of competition.
  </ul>
</blockquote>

--
 Windows:
32 bit graphical interface for a
 16 bit patch for an
 8 bit operating system written for a
 4 bit processor by a
 2 bit company that can't stand
 1 bit of competition.
  </html>

</x-html>
Received on Thu Sep 28 08:27:30 2000

This archive was generated by hypermail 2.1.8 : Tue Dec 02 2003 - 17:29:00 EST

Re: Breakout PCB & Motorola monitor - questions for everyone!!!!