Mk3 horizontal glitches only at startup

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser. Please create an account and log in to see all content by default. This is a limitation of the forum software.


Also users that are not logged in can not create new threads. This is a, unfortunately needed, counter measure against spam. Please create an account and log in to start new threads.

Don't Panic. Please wash hands.
  • Same here - If I stick to the standard Amiga resolutions - Hi Res Interlaced for example - there are no issues. I know Jens has documented already that we can expect to see this but, like you, am intrigued if anyone has found a solution/workaround for it?

  • Up until now I've assumed it's something to do with timings that settle down once everything is warmed up. The behaviour does seem to change depending on the ambient temperature. For instance normally the Amiga has to be on for a minute or so before it starts to glitch, then it will glitch for five minutes or so and will then settle down and stop.


    At the moment we have very high temperatures in the UK and the room my Amiga is in is around 10C warmer than normal, and when I switch my Amiga on it is immediately glitching as soon as it boots into Workbench, and then also settles down much quicker. I am not using my Amiga for extended periods at the moment

  • As we pointed out in other threads before, these gfx glitches on some A1200s are not Indivision-specific, but computer-specific. Indivision only has a few options to reduce the effects.


    If switching on the pull-up on CCK makes a difference, but does not provide a satisfactory solution, then try the CCK capcitance - that's the solution for more computers.


    In all these cases, please keep in mind that Indivision is not the cause - it merely makes a weakness of your specific board visible. The output looks exactly the same on the DB23 output, but since you probably didn't have a monitor that displays these modes, you never noticed it before.

  • As we pointed out in other threads before, these gfx glitches on some A1200s are not Indivision-specific, but computer-specific. Indivision only has a few options to reduce the effects.


    If switching on the pull-up on CCK makes a difference, but does not provide a satisfactory solution, then try the CCK capcitance - that's the solution for more computers.


    In all these cases, please keep in mind that Indivision is not the cause - it merely makes a weakness of your specific board visible. The output looks exactly the same on the DB23 output, but since you probably didn't have a monitor that displays these modes, you never noticed it before.

    I do have a monitor that will display those screen modes, and I do not get the glitching out of the DB23 output.


    Saying that, I am fine with it; it only lasts a couple of minutes and then goes away so as long as I know no harm is being done then I am happy. It's a minor miracle that a 30 year old computer is able to have a pure digital video output at all IMO.

  • I do have a monitor that will display those screen modes, and I do not get the glitching out of the DB23 output.

    Wow - that would be a first, absolutely worth investigating, as Indivision takes it's input from the very data that Lisa is sending out to the DAC that serves the DB23 port.


    Can you take pictures of a side-by-side setup? We might be talking about different types of glitches after all.

  • Wow - that would be a first, absolutely worth investigating, as Indivision takes it's input from the very data that Lisa is sending out to the DAC that serves the DB23 port.


    Can you take pictures of a side-by-side setup? We might be talking about different types of glitches after all.

    I'll go one better and will take some videos and will post some Youtube links at the weekend.

  • Wow - that would be a first, absolutely worth investigating, as Indivision takes it's input from the very data that Lisa is sending out to the DAC that serves the DB23 port.


    Can you take pictures of a side-by-side setup? We might be talking about different types of glitches after all.

    It seems I am mistaken. I believed the RGB port was fine, but when I connected it to my OSSC I got exactly the same glitching in DBLPAL HiRes No Flicker. Now I think harder, I used to run in Productivity when I was using RGB output, which is a slightly lower resolution.


    My apologies.

  • No need to apologize - we all make mistakes, me in cluded.


    The AGA chipset is "on the edge" when it comes to chip ram timing, and especially the double-CAS modes are affected by temperature drift. To be honest, I'm not looking forward to tweaking this for the A1200 Reloaded, although I have way better possibilities today than what Commodore hat in the 1990s.

  • I too see this on 2 of my mk3 AGA when starting in highgfx mode, Its just as nuttie described, both on a AT 1D4 and a C= 2B board. I haven't tested the CD32 in highgfx mode yet. Would it be Is possible to change the configuration program or the flashtool to allow the CCKLine and PCLK settings to be set from the command line, leaving the reset of the configuration intact. That way a simple iconx script could be called rather than having to navigate through the program to alter the advanced settings.


    Probably of topic but as it was mentioned will the A1200 reloaded have things like indivision mk3, rapidroad and micromys technology neatly integrated into the motherboard. design

  • You could generate different configs with these expert settings set/cleared and save config to a file, which in turn could be loaded with the tool - please see Wiki for command line options. Saveapply argument was among the latest additions to the tool.

  • I have finally completely resolved this issue with my A1200 & MK3.


    The solution is to remove resistor E127R from the underside of the PCB and to then short the pads. You could leave the resistor in place and solder a small piece of wire across it.


    More details here:

    [link removed]

    From that simple fix my display is now rock solid from the moment I switch on my Amiga and through several hours of use.


    As Jens says, it’s an Amiga problem, not an Indivision one; the Indivision just exposes the problem by allowing easy access to the higher display modes.

  • I have removed your link, as it is a) a bad mod, and b) a commercial site that I don't want to be associated with in this forum.


    The mod is bad because it loads the clock lines way too high by just shorting a resistor that is part of an RC pair. The better solution is to remove the capacitors as described in our Wiki for over a decade:


    http://wiki.icomp.de/wiki/ACA1…uired_board_modifications


    Please note that the source of this information is a Commodore technical bulletin that was distributed to all service centers worldwide. While my WIki entry is mostly about accelerators, a wrong board assembly will of course also affect AGA glitche/stability.


    Jens

  • Could you be more specific, that is, provide links to unsolved support cases? I mean, the solution has been in the Wiki for >10 years and from a logical point of view: It's a weakness of the AGA chipset, not of the flicker fixer. The flicker fixer merely makes this weakness visible.


    Here's an experpt from the 2012 short manual of Indivision AGA MK2:

    Quote

    Should you observe this behaviour, there are simple ways to fix this issue. Most of the

    times, a small chip-cooler on the Alice-chip of your computer solves the issue. In rare

    cases, you may need to do soldering on your mainboard. Please ask a technician or your

    reseller for closer information on this.

    Since introduction of Indivision AGA in the year 2008, we've solved such problems, although they are not *caused* by our flicker fixer. I believe that's a reason to say "thanks" - I see nothing that I could do better here (other than making a proper AGA main board, which is on the Agenda).


    Jens

  • including your good self suggesting it as a fix on EAB several years ago:

    Did you look at the date? That's >14 years ago, and I have learned a lot since then. I certainly don't recommend "just bridging" any more.


    We're dealing with 30-year old chips that aren't made any more. Reducing resistance to a capacitor means to increase the load to the pin, which may lead to a premature defect. It's not worth it - better remove the caps.


    You did not answer my question about any unresolved cases in this support forum. I guess you didn't find any - so please stop claiming that this is the case.


    Jens

  • I've already removed the caps, and you know full well I am not going to read through this entire forum.


    Just two years ago on this forum, written by you and saying much the same:


    "You won't damage anything if that's what you're worried about. I'd prefer to try a ferrite bead in this position, as the original design with E127R and E127C intended to make this a filter for higher frequencies. The resistance may cause trouble with the rise/fall time, so reducing it's resistance may be worth a try, but since you still want to maintain the "block high frequencies"-property, a ferrite bead is the better part to try. Go for the standard 56 Ohms @100MHz as a first shot in the dark."

  • I am not trying to go out of my way to pick a fight with you on this. I am simply wanting to make the point, please do not shut down your own customers with little explanation for a fix that you yourself have recommended in multiple places previously.


    You have an absolutely fantastic product, that gives me immense pleasure to use with my Amiga.


    Apologies if I have come across as confrontational.