Could a damaged quad bridge rectifier (1200XL) leak AC voltage to the system and cause faint vertically scrolling snow bars on the video?

[ACML]

The big square rectifier is a quad bridge type.  If say one of the legs has failed, wouldn't that bleed AC voltage to the board?  So instead of 11.5VDC to the large filter cap, you get 9.5VDC and 1-2VAC.  If everything is working, you should get no VAC to the big cap.  If that is the case, could that cause video snow on the display which looks like thick vertical scrolling bars of faint static?

 

[AndyR]

I won't try to guess at exactly what the effects of a bad 12V rail would be.

 

But, is there an end to this story you're imagining that doesn't involve fixing your power supply?  If you know it's broke, it's broke.  Fix it.

 

Seems like if you really want to know, get a bench power supply (or just an ATX box from an old PC) cabled to the board instead and see if that fixes the video issue.

[ACML]

39 minutes ago, AndyR said:

Seems like if you really want to know, get a bench power supply (or just an ATX box from an old PC) cabled to the board instead and see if that fixes the video issue.

That makes perfect sense.  Thanks.  Until you said it, I forgot that I have a spare ATX switching power supply on the shelf that indeed does have +12VDC.

 

As to "is it broke"?  Don't know.  Not knowledgeable enough to test a quad bridge rectifier with a simple multi-meter. 

 

Edited February 24, 2023 by ACML

[Peri Noid]

4 hours ago, ACML said:

The big square rectifier is a quad bridge type.  If say one of the legs has failed, wouldn't that bleed AC voltage to the board?  So instead of 11.5VDC to the large filter cap, you get 9.5VDC and 1-2VAC.  If everything is working, you should get no VAC to the big cap.  If that is the case, could that cause video snow on the display which looks like thick vertical scrolling bars of faint static?

 

If one of the channels fails, what you get is half of a rectified "sine" wave at its output - every second "heel" is present. Which means half of the current delivered to the big cap and 7805 converter. It means, that the 7805 may be less effective and may give less stable output. And this, as a result, may influence the picture quality or even board stability. 

Edited February 24, 2023 by Peri Noid

[Rybags]

Sounds very likely that mains noise could cause what you describe.

The mains is 50/60 Hz and generally kept very near to that but the Atari video signal is non-interlaced so doesn't meet either the broadcast or mains frequency which is why such noise would move like you say.

 

In theory given the Atari is under mains frequency the noise should come earlier each frame so be moving upward.

You should even be able to do a quick calculation to work out what speed it would run at.

 

59.94 is broadcast, 59.9227 is hardware which is .02887% faster.  So I'd guess such interference should repeat about every 345 frames or about 5.75 seconds.

Edited February 24, 2023 by Rybags

[Peri Noid]

That's exactly what you're getting - the 7805 is fed not constantly but in intervals. If the big cap is not able to sustain the power, you actually have a 50/60 Hz noise "embedded" at the 5V rail since there is no power delivered to the converter at this frequency. 

[ClausB]

2 hours ago, Rybags said:

59.94 is broadcast, 59.9227 is hardware which is .02887% faster.  So I'd guess such interference should repeat about every 345 frames or about 5.75 seconds.

NTSC broadcast frequency doesn't enter into it. The US mains are 60.00 Hz on average, though that can vary quite a bit depending on grid load.

[ACML]

5 hours ago, Rybags said:

In theory given the Atari is under mains frequency the noise should come earlier each frame so be moving upward.

59.94 is broadcast, 59.9227 is hardware which is .02887% faster.  So I'd guess such interference should repeat about every 345 frames or about 5.75 seconds.

Yes, the bands of faint snow are indeed moving vertically upwards.  As to the frequency. I think the next new snow bar is more like every 1.5 seconds apart.  

[ACML]

5 hours ago, Peri Noid said:

If the big cap is not able to sustain the power, you actually have a 50/60 Hz noise "embedded" at the 5V rail since there is no power delivered to the converter at this frequency. 

That begs another question I posed on the AA board regarding the big 10,000uF cap.  The OEM one is about the size a "C" battery.  New 10,000uF caps are much smaller, probably less than half the volume.  Shouldn't a new production (much smaller) 10,000uF cap perform just as well as the OEM?

 

[Rybags]

Error in my calcs earlier... the movement would be somewhat quicker so 1.5 does sound about right.

[Peri Noid]

3 hours ago, ACML said:

That begs another question I posed on the AA board regarding the big 10,000uF cap.  The OEM one is about the size a "C" battery.  New 10,000uF caps are much smaller, probably less than half the volume.  Shouldn't a new production (much smaller) 10,000uF cap perform just as well as the OEM?

 

If its specs are the same, it should. 

[+bob1200xl]

There should be no ac voltage on the big cap. If there is, bad cap or bad rectifier bridge.

 

The stock cap is 10,000 @ 16v. I use 10,000 @ 35v, which is close to the original size.

 

You might want to check the solder lugs on the rectifier for cracked solder.

 

 

Bob

 

 

[Kenshi]

The voltage on the capacitor will fluctuate if there's any load on it. The regulator is constantly pulling charge from it but the rectifier can only recharge it near the peaks of the AC voltage or every 1/120 of a second. The numbers you gave are about right if the load is an amp or two. If you lose a diode in your rectifier, it'll only charge half as often and the fluctuation will be twice as bad. It's up to the regulator to clean up the fluctuations and make the voltage constant.