Image Map Image Map
Page 2 of 3 FirstFirst 123 LastLast
Results 11 to 20 of 22

Thread: Rev 4 Apple II with weird colour issue

  1. #11

    Default

    RAM tests won't change colors in your case. Check sockets of those three ICs. Check which one of them changes colors. Check once again ICs are good and they are as specified (don't rely on an IC tester!). BTW the good picture you provided is an output of an emulator, that's why it is not "pixelated". AppleWin in NTSC moitor mode shows extremely accurate picture.

  2. #12
    Join Date
    Jan 2011
    Location
    Vancouver, BC
    Posts
    4,633
    Blog Entries
    3

    Default

    After checking those ICs, I had no luck. So then I started replacing the entire board around the RAM. When I got to the 74ls153, 74ls195, and 74ls20 and C1, C2 and D2 and changed them.. suddenly I had a slight change in hue. When I adjusted the color trim and the color on my monitor.. it was perfect.

    I swapped the original chips back in one by one.. no change. Swapped *all* the original chips back... same thing.. colors stayed good.

    I can only put it down to maybe loose chip in socket? All the other IIs gave correct colors without touching the monitor and still do now.. and no amount of dialing the color trim was changing anything prior to the chip swap..

  3. #13
    Join Date
    Dec 2011
    Location
    Dallas, TX
    Posts
    248

    Default

    That looks like part of the video/DRAM timing generator circuit. Most failures here would result in all-or-none failure modes. The section does output a color burst enable signal. Perhaps the color burst was not getting enabled, resulting in color sync issues.

  4. #14

    Default

    Quote Originally Posted by dfnr2 View Post
    That looks like part of the video/DRAM timing generator circuit. Most failures here would result in all-or-none failure modes. The section does output a color burst enable signal. Perhaps the color burst was not getting enabled, resulting in color sync issues.
    The aging sockets of vintage Apple ][s are very problematic and I suspected them as I have written in several messages in this topic. No IC was defective in this case. Up to //e I prefer some clones for better PCBs. From //e onward I am fan of original Apple 2s

  5. #15
    Join Date
    Dec 2011
    Location
    Dallas, TX
    Posts
    248

    Default

    Quote Originally Posted by george View Post
    The aging sockets of vintage Apple ][s are very problematic and I suspected them as I have written in several messages in this topic. No IC was defective in this case. Up to //e I prefer some clones for better PCBs. From //e onward I am fan of original Apple 2s
    Since it began working after just removing and replacing the chip, you are right that it's probably a socket issue. I was only speculating on the cause of the color issue, since the problem seemed to be coming from the section of the circuit you described. A hardware failure of any type, including a bad socket, can be a good opportunity to learn about the circuit, if you are able to drill down to the specific cause of the symptom, such as what signal was not getting from one pin to the next. In this case, since the problem vanished, it will be hard to reproduce conveniently, but we can speculate.

    That's why hardware nerds will often try to scope out the signals first, before starting to swap out chips.

    Dave
    Last edited by dfnr2; October 10th, 2019 at 12:28 PM.

  6. #16

    Default

    Quote Originally Posted by dfnr2 View Post

    That's why hardware nerds will often try to scope out the signals first, before starting to swap out chips.

    Dave
    I guess you are a nerd then. Theoretical one. Where would you connect your scope's probe then? To the IC pins? To each one of them? To the solder pads below the socket? To all ends of each track? And what will you see on the scope? A fully missing signal? Or partly attenuated by the socket-pin's resistance but still very close to what it should be? Where would you connect your next probes if any? Rhetorical questions...

  7. #17
    Join Date
    Dec 2011
    Location
    Dallas, TX
    Posts
    248

    Default

    Quote Originally Posted by george View Post
    I guess you are a nerd then. Theoretical one. Where would you connect your scope's probe then? To the IC pins? To each one of them? To the solder pads below the socket? To all ends of each track? And what will you see on the scope? A fully missing signal? Or partly attenuated by the socket-pin's resistance but still very close to what it should be? Where would you connect your next probes if any? Rhetorical questions...
    It's a good question. To a beginner, the options seem limitless. A scope only has two or four traces, and they should be meaningful. Hobby time is limited. You must be judicious.

    The key, not just for hardware, but for many kinds of problems, is to exploit the underlying structure to come up with some kinds of constraints on the range of problems. For example, a circuit board is not just a random bag of chips, pins, discreets, and wires. The chips have pins which have specific functions. That's a constraint imposed by structure. The circuit is made by connecting components to interact with one another, usually with a flow of functionality from basic (e.g., a counter) to specific (e.g., a train of pulses, or a seqence of chip enable strobes).

    Just as replacing caps or swapping chips is not the optimum starting point, neither is probing pins. The first steps take place in your head.

    You start by looking at the symptom as systematically as you can (e.g., Is the video is all there, but is the wrong color? Is everything OK spatially, or does some of the video look garbled? Etc.) Try to form some basic theories of what could cause the symptoms (It's OK if they are simple or wrong; this is an iterative process). Then, take a look at the schematic. Focus on the areas that could cause the problem based on your working theory. If you see a circuit pattern you haven't seen before, try to understand how it works.

    For example, in this case, if one set of colors isn't being formed, you might look at the circuit that selects one color set over the other. In this case, there is a flip flop that delays the video line by one cycle of a 14 MHz clock. Either the delayed or non-delayed data is selected on a per-byte basis, so I'd start looking to see if the data stream going in and going out look identical but shifted. I'd fill the screen RAM with one color set and then the other, and see which stream (regular or delayed) matches the output.

    If that is all OK, I'd look at the color burst and see if it's there. I'd check the data streams relative to the color burst to see if they appear shifted.

    Now, don't get me wrong. I do think it's reasonable, for an old computer, to try and re-seat the chips first, and avoid and fiddling around if possible. But if a problem persists, I would try trouble shooting. If re-seating a chip made the problem go away, as was the case here, I'd look at what the chip does, and try to form a theory for what happened. In some cases, I might even try to test the theory by trying to recreate the issue.

    I also do believe in checking power first, to make sure there's not a power issue. Some things fail more than others, like cables. Scoping a signal at both ends of a cable can help determine if the cable is at fault.

    Getting good at debugging circuits does take time and experience, but you get that by going through a process each time. You can start with no knowledge, and at first it takes a lot of time to work through things, but the pleasure you get from figuring something out will keep you going, and over time you get much faster. You will see certain patterns over and over and that will help you more quickly evaluate a new circuit.

    Look up the data sheets for the chips, including truth tables, enable signals, etc. Try to understand each chip. For analog circuits, if you see a new circuit topology, try to guess what it does, look it up, and do a little reading about it. This knowledge accumulates over time; each project is easier than the last.

    The problem with blind chip swapping is that you miss the chance to learn, and also it may not solve tricky problems, or problems where chip swapping is impractical (e.g., chips are all soldered in, or no substitutes are available). If you take the time to learn something from each failure, then eventually you get good at fixing stuff. If you get lucky, great. But if you rely on luck to fix stuff, there will be a lot of stuff forever out of reach.

    Dave

  8. #18
    Join Date
    Dec 2011
    Location
    Dallas, TX
    Posts
    248

    Default

    Quote Originally Posted by george View Post
    I guess you are a nerd then. Theoretical one. Where would you connect your scope's probe then? To the IC pins? To each one of them? To the solder pads below the socket? To all ends of each track? And what will you see on the scope? A fully missing signal? Or partly attenuated by the socket-pin's resistance but still very close to what it should be? Where would you connect your next probes if any? Rhetorical questions...
    And to answer more literally, where to connect the probe depends on the board. Sometimes there are convenient test points In many cases, you will probe a ship or discrete device directly holding the probe by hand. Other times, you'll clip the probe to a lead. Sometimes, when a board is not easily accessible when plugged in (and no extender available) you'll solder some wires to the points you're interested in to bring out the signals of interest.

    What you're looking for will obviously depend on the circuit. As a first pass, don't look for the malfunction. Look for the correct signal. The spectrum of failures will surprise you, but if you know what you expect to see, you will most likely be aware if something is wrong.

    Some failures are subtle or intermittent. If you suspect a subtle failure like a glitch, or subtle noise coupling, then you will need to set up a special experiment to catch the failure. But that will come from understanding the circuit and expecting a certain type of failure. I can tell you that professional engineering teams can on occasion spend months tracking down this type of bug. But, this is more common in new designs than in old, proven designs.

    Dave
    Last edited by dfnr2; October 11th, 2019 at 07:29 AM.

  9. #19

    Default

    Quote Originally Posted by dfnr2 View Post
    And to answer more literally, where to connect the probe depends on the board. Sometimes there are convenient test points In many cases, you will probe a ship or discrete device directly holding the probe by hand. Other times, you'll clip the probe to a lead. Sometimes, when a board is not easily accessible when plugged in (and no extender available) you'll solder some wires to the points you're interested in to bring out the signals of interest.

    What you're looking for will obviously depend on the circuit. As a first pass, don't look for the malfunction. Look for the correct signal. The spectrum of failures will surprise you, but if you know what you expect to see, you will most likely be aware if something is wrong.

    Some failures are subtle or intermittent. If you suspect a subtle failure like a glitch, or subtle noise coupling, then you will need to set up a special experiment to catch the failure. But that will come from understanding the circuit and expecting a certain type of failure. I can tell you that professional engineering teams can on occasion spend months tracking down this type of bug. But, this is more common in new designs than in old, proven designs.

    Dave

    Dave
    Hey, amateur, don't explain me how to repair or reverse electronics or software. Go into someone else's shoes and explain them remotely. Scope is needed when you lack of brains in case you need to measure near DC electrical contact. Many of us started when scope was luxury and had to deal without it. By thinking. This reminds me of applefritter forum that some guy despite my advice bought new contemporary scope (the old one was more than perfect for his Apple1). Those kind of thinking is when one is stupid enough that thinks that money would replace good thinking. In our country their is a saying "To usa a cannonball to kill an ant".

  10. #20
    Join Date
    Dec 2011
    Location
    Dallas, TX
    Posts
    248

    Default

    Quote Originally Posted by george View Post
    RAM tests won't change colors in your case.
    The high bit determines the color group, so a stuck bit could produce the symptom, under some circumstances. A faulty bit could be a bad RAM chip, but also a faulty register.

    Dave

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •