Jump to content

Photo

video fuzzy even after composite A/V mod

repair fuzzy video a/v mod troubleshooting

40 replies to this topic

#26 Divarin OFFLINE  

Divarin

    Space Invader

  • Topic Starter
  • 22 posts

Posted Fri Apr 26, 2019 2:36 PM

Lohe, yeah I did try that but it didn't help

 

So I have an update. I had given up and when I started the refurb on the 2nd unit I decided to use the power supply it came with (an original) rather than one of my new ones which I had ordered from china.  Since 2600 #1 was still on the bench I decided to check it out with this power supply and it actually looked clean.  At one point I had tried the power supply I use in my living room on the 2600 I keep and play with ... still fuzzy.  That one is also new but a different brand and a bit more expensive.

 

I measured the voltage coming out of the Chinese ps's and it was a bit over 9volts, pretty steady.  Then I measured the voltage coming out of the original power supply and it was like 14!  It's labeled as a 9 volt power supply though.  The voltage coming out of the voltage regular is unchanged but the voltage going into it is higher.  I suspect a capacitor (or something) which is in circuit somewhere before the voltage regulator may be bad but if you pump up the voltage enough the fuzziness goes away.

 

However I may still end up using this as a parts board because the composite mod is really dark.  I've read some posts that say to remove one of the resistors and I did which lightened it up a bit but it's still too dark, in fact so dark that if you go into b&w mode you can't see a thing.  If I removed both resistors then it's way too bright (and washed out).  The strange thing is before, when I was trying to get rid of the fuzziness I noticed it was dark, removed the one resistor, and it looked fine.  I can't explain that.

 

Anyway, I don't know if I can blame the power supply 100%, since it is putting out the voltage it says it is but the fact remains that if I use an original Atari power supply the fuzziness is gone.



#27 Djoulz OFFLINE  

Djoulz

    Space Invader

  • 13 posts

Posted Fri Apr 26, 2019 4:21 PM

Try to look at the jack connector when you put it into the slot (that should be transparent), you will see that the click is different between the two PSU ... I've made this test with 3 PSU and the one with the shortest jack produce fuzziness.



#28 Osgeld OFFLINE  

Osgeld

    Quadrunner

  • 5,988 posts
  • Location:Nashville, TN

Posted Fri Apr 26, 2019 4:49 PM

some power supplies are just shit

 

I had one from a ras pi kit (radio shack) that when adapted for my 65XE looked like a worm farm was crawling over top of the picture, got a different one and its razor sharp



#29 ChildOfCv OFFLINE  

ChildOfCv

    Moonsweeper

  • 407 posts

Posted Fri Apr 26, 2019 4:50 PM

Lohe, yeah I did try that but it didn't help

 

So I have an update. I had given up and when I started the refurb on the 2nd unit I decided to use the power supply it came with (an original) rather than one of my new ones which I had ordered from china.  Since 2600 #1 was still on the bench I decided to check it out with this power supply and it actually looked clean.  At one point I had tried the power supply I use in my living room on the 2600 I keep and play with ... still fuzzy.  That one is also new but a different brand and a bit more expensive.

 

I measured the voltage coming out of the Chinese ps's and it was a bit over 9volts, pretty steady.  Then I measured the voltage coming out of the original power supply and it was like 14!  It's labeled as a 9 volt power supply though.  The voltage coming out of the voltage regular is unchanged but the voltage going into it is higher.  I suspect a capacitor (or something) which is in circuit somewhere before the voltage regulator may be bad but if you pump up the voltage enough the fuzziness goes away.

 

However I may still end up using this as a parts board because the composite mod is really dark.  I've read some posts that say to remove one of the resistors and I did which lightened it up a bit but it's still too dark, in fact so dark that if you go into b&w mode you can't see a thing.  If I removed both resistors then it's way too bright (and washed out).  The strange thing is before, when I was trying to get rid of the fuzziness I noticed it was dark, removed the one resistor, and it looked fine.  I can't explain that.

 

Anyway, I don't know if I can blame the power supply 100%, since it is putting out the voltage it says it is but the fact remains that if I use an original Atari power supply the fuzziness is gone.

So moar powr fixes the fuzziness?  That is consistent with lack of sufficient power.

 

There are two potential causes there:

 

1) Weak power supply.  Multiple good supplies are not working, but your 14V supply (probably a blown regulator) masks this issue by providing way too much.  This will work for a while, but will be taxing the regulator on the Atari board.

2) Short on the board.  The electrolytic capacitors are often the suspect, which is why recapping is popular.  But it's not always the issue.  One Atari system had a shorted ceramic capacitor, which is one you don't see often at all.  Other potential issues come from the solder work, which may accidentally bridge connections when too much is applied.  A quick measurement with a multimeter in ohms mode on both sides of the regulator would give you an impression of its load on the power supply.

 

Either way, the core problem is that the power supply is offering less current than the board wants.  When this happens, it usually results in AC ripple, which is why it may be useful to measure.  Secondly, it's likely the ripple that is the cause of the fuzzy video, but if so, it should show up as AC voltage on the power in plug.



#30 nick3092 OFFLINE  

nick3092

    Chopper Commander

  • 236 posts

Posted Fri Apr 26, 2019 5:18 PM

1) Weak power supply.  Multiple good supplies are not working, but your 14V supply (probably a blown regulator) masks this issue by providing way too much.  This will work for a while, but will be taxing the regulator on the Atari board.


He said the 14v supply was an "original". I'm assuming he means that it is the original Atari linear supply with no regulator in it.

Which would put out about 14v with no load on it. It would not tax the regulator, as once it's loaded the supply drops to around 9.

#31 Divarin OFFLINE  

Divarin

    Space Invader

  • Topic Starter
  • 22 posts

Posted Fri Apr 26, 2019 6:17 PM

Right an original Atari power supply. So is it normal for those to put out 14v when not under a load? I've never measured those before.

#32 nick3092 OFFLINE  

nick3092

    Chopper Commander

  • 236 posts

Posted Fri Apr 26, 2019 6:51 PM

Yes. Old power supplies are made using a power transformer, which is why they are so much heavier than newer switching power supplies that are all solid state.

Transformers always put out higher voltage when there is no load on them. They are designed so they put out the proper voltage when the specified load is added. Newer switching power supplies are regulated and put out the exact voltage (or very close to it) regardless of load.

The problem with newer switching supplies is they are electrically noisy compared to the old linear/transformer supplies. Which older electronics like the 2600 don't tolerate the switching noise because they don't have the proper filtering built in.

If the filtering is done properly in the switching power supply, then it will work fine with older electronics. But as Osgeld mentioned above, there are a fair share of cheap supplies out there that are very noisy because the filtering is inadequate.

Edited by nick3092, Fri Apr 26, 2019 6:54 PM.


#33 Divarin OFFLINE  

Divarin

    Space Invader

  • Topic Starter
  • 22 posts

Posted Sat Apr 27, 2019 8:32 AM

Ah okay thanks. So I got a lot of 10 of these power supplies, I'd like to be able to use them on 2600's as I refurbish them (if I pick up one that doesn't have a power supply) maybe I can build a small filtering circuit or something. Or chuck them and pick up some quality power supplies.

If I could fix the dark composite mod issue then this 2600 can have a second life.

#34 ChildOfCv OFFLINE  

ChildOfCv

    Moonsweeper

  • 407 posts

Posted Sat Apr 27, 2019 10:14 PM

Though honestly, I would expect the 2200uF capacitor to be enough filtering.  The next step might be to add a choke.  And again, if filtering is needed, it should show up as AC ripple.

 

Where did you buy those power (or are they poser) supplies?



#35 Divarin OFFLINE  

Divarin

    Space Invader

  • Topic Starter
  • 22 posts

Posted Sun Apr 28, 2019 4:06 AM

https://www.ebay.com...ug/192439401874



#36 nick3092 OFFLINE  

nick3092

    Chopper Commander

  • 236 posts

Posted Sun Apr 28, 2019 9:40 AM

I always thought switching supplies by nature needed more filtering than just a big cap? That works good for the 60hz ripple on a linear supply. But I thought you needed a different approach for the high frequency and other noise/interference generated by a switching supply.

#37 Divarin OFFLINE  

Divarin

    Space Invader

  • Topic Starter
  • 22 posts

Posted Sun Apr 28, 2019 2:23 PM

Hmm, I don't know. I've used these power supplies on other 2600's a couple of times without any issue but maybe this one has a bad cap or a bad ground or something that makes it more susceptible to noise.

#38 ChildOfCv OFFLINE  

ChildOfCv

    Moonsweeper

  • 407 posts

Posted Sun Apr 28, 2019 3:17 PM

I always thought switching supplies by nature needed more filtering than just a big cap? That works good for the 60hz ripple on a linear supply. But I thought you needed a different approach for the high frequency and other noise/interference generated by a switching supply.

 

I guess it's possible.  The filter does need good ESR, whatever you use.  Electrolytic caps do lack in the fast response category, and you'll notice that the 2600 also has a ceramic cap in parallel with the 2200uF one, likely to pick up that slack.

 

The frequency of a switching supply is likely closer to around 100KHz, so especially for a low-quality capacitor with bad internal resistance characteristics you may see some "fuzz" in the output.  A series choke might help, something around the inductance of the ones that ColecoVision uses on the main power inputs (200mH or so?)

 

This is just me guessing at the theoretical power supply's construction, but I guess it might cause those "worm tracks" due to the high frequency fuzz, but then why does it only happen for a small portion of the screen?  Maybe because the switching is only at work during the peak times of the AC cycle?  Let's see, some napkin math says that a 100KHz switcher would cause about 6 "worms" per line.  400KHz would of course cause around 24.

 

Yeah, I suppose that could explain the odd interference.



#39 Osgeld OFFLINE  

Osgeld

    Quadrunner

  • 5,988 posts
  • Location:Nashville, TN

Posted Sun Apr 28, 2019 4:32 PM

Though honestly, I would expect the 2200uF capacitor to be enough filtering.  The next step might be to add a choke.  And again, if filtering is needed, it should show up as AC ripple.

 

Where did you buy those power (or are they poser) supplies?

big caps only smooth out slow changes, small caps filter out higher frequencies (based on maf) 



#40 Lohe OFFLINE  

Lohe

    Space Invader

  • 17 posts
  • Location:Graz, Austria

Posted Sun May 5, 2019 6:38 AM

Divarin, just to ask, did you try the 2nd power supply on a different socket maybe?

#41 ChildOfCv OFFLINE  

ChildOfCv

    Moonsweeper

  • 407 posts

Posted Mon May 13, 2019 1:22 AM

Revisiting this subject since I've had a chance to get my hands on one of these power supplies.

 

It's an interesting design.  It uses a transformer, but not as a step-down.  It uses it for buck conversion.  It uses the zener on the output side to tell the input side to stop oscillating when the voltage reaches the target.

 

While it may explain this would cause problems (in spite of the ceramic filter capacitor on the Atari), it is still strange that it only affects a portion of the screen at a time.  The oscillation would be happening at a more constant and higher frequency rate.

 

But oh well, if the old step-down and linear regulator method works better, then so be it.

 

Attached Files







Also tagged with one or more of these keywords: repair, fuzzy video, a/v mod, troubleshooting

0 user(s) are browsing this forum

0 members, 0 guests, 0 anonymous users