MK3 no output

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.
  • I had also some problems when I had the Indivision AGA MK3 in the begin.

    Some errors like no hardware found, bad picture. This was with the original A1200 PSU.

    Then I bought a PSU from a Polish company, did't have a problem for 1 week...

    After 1 week again the same problems, and than on a day... that new PSU was blown up after using for about 20 minutes that day. End of life for that Polish PSU.

    At last, I bought in August of this year the CA-PSU from Jens and now 4 months later, didn't have any problems so far.

    It's maybe not the cheapest PSU on the market, but it's the best I ever used.

    A1200 Rev2B OS 3.2| SanDisk CF64GB | Blizzard 1230 IV 64MB FPU 68882/50 | Indivision AGA MK3 | CA-PSU | Eizo S1933 | Bose Companion 2 series III

    A500 Rev6A OS 3.2| SanDisk CF16GB | ACA500+ V2 / Blizzard 1230 IV 32MB | Indivision ECS V3 | ACE2b | CA-PSU | Eizo S1933 | Bose Companion 2 series III

    A1200 Rev2B OS3.2.2 | Black Edition | SanDisk MSD32GB | PistormLite32 | Indivision AGA MK3 | RTG Picaso96 1280x1024| CA-PSU | Eizo S1933

    iAmiga 27" | OS 3.2 FS-UAE | SDD 1 TB | RTG Picaso96 2560x1440

  • Thanks for the positive referral - Mr. Toast won't help this, as he's in the US, and the 110V version of CA-PSU is not yet available. It'll have to be a rebuilt original Commodore PSU, at least the output cap should be exchanged, and voltage should be adjusted to be within 1% deviation on the board, under load.

  • OK, it was worth a shot. Recapped the PSU with ultra-low ESR Panasonics on the larger caps and the rest were Wurth for the smaller ones. Tested it without load and verified operational before putting it in the A1200. Once powering the A1200, I adjusted the 5V rail as perfectly as I can on an un-calibrated meter under load.

    Exact same outcome: Rescue disk works, warm boot to gorgeous display. On cold boot, I get hardware not found,


    Is it possible I just have a bad unit? I can only think of the flash RAM failing as to why reprogramming would temporarily work and not a power nuance.

  • Configuration of the FPGA always goes through the flash chip. There is no way that the software can push code directly into the FPGA. So the chip would have to remember it's content directly after flashing, but not after power-cycling - almost like RAM does. I haven't heard of such a defect before, but there's always a "first".


    Given the long distance and your obvious soldering skills, would you be willing to give a new flash chip a try? The flash chip is the only 8-pin chip next to the socket. If you feel confident enough to exchange that, I can ship out a new chip today, otherwise I'd just ask you to send the unit back to us after reading this article.

  • Given the long distance and your obvious soldering skills, would you be willing to give a new flash chip a try? The flash chip is the only 8-pin chip next to the socket. If you feel confident enough to exchange that, I can ship out a new chip today, otherwise I'd just ask you to send the unit back to us after reading this article.

    Yes, I have the skills and tools to swap the chip out. I can rework everything except BGAs.


    Let's give this a shot.

  • Let's give this a shot.

    Too late for a shipment today, but I'll ship one with tomorrow's mail. This side of the pond is six hours ahead :-)


    If you have your customer ID (K-number on the invoice) or the order ID at hand, I can prepare the paperwork today.

  • My order number for this was 81634

    I do not see any other unique identifiers on the email invoice received. The order date was August 8th if that helps.


    No worries about the delays. Yes, the time difference presents some challenges!

  • IC arrived a week ago but needed to clear the bench of another project.


    It was a noble effort but replacing the flash has me back at the start. Boot to hardware not found. Reflash and everything works but does not survive cold boot. I'm going to put it under the microscope and give the board some tighter inspection-but, I think we're going to need to start the RMA process.


    Thanks for giving this a shot, Jens.

  • Return arrived, works without problem out of the box.


    We've had other cases where we had trouble reproducing a "no output on VGA and/or HDMI" report. Just a few days ago, we found a race condition when changing PLL settings (this was during debugging the autoresolution modes), which might explain your sequence of "works first, but fails when config tool is saving something to flash".


    Since you have a memory expansion only (and one that is not sold in Europe), the closest thing we come to your config is to operate an ACA1221lc with 14MHz CPU speed (sync) and see if we can reproduce these errors with the old (currently in the Wiki) version, and then see if it has been fixed already with the latest internal changes. Timm is currently on the Indivision AGA MK3 config tool and will take it from here. We'll delay shipping out your unit until we have new findings.

  • Trying to get to the bottom of this, our technician was a bit too fast bringing the unit into "factory default" condition, meaning, the flash was overwritten. That means I have no way to find out what configuration you flashed that made the unit go "blind" after a power cycle.


    Could it be that you've tried to save an autores-mode?

  • Trying to get to the bottom of this, our technician was a bit too fast bringing the unit into "factory default" condition, meaning, the flash was overwritten. That means I have no way to find out what configuration you flashed that made the unit go "blind" after a power cycle.


    Could it be that you've tried to save an autores-mode?

    Sorry Jens-I missed the notifications on this.


    Just for clarity, during my testing I had the exact same results with the memory expansion removed (completely stock).


    If it was in autores, it was by default. I would do the recovery disk and boot off of the floppy or CF card. No attempt was made to configure the card other than defaults in most of my attempts. Configuring through all the available modes also did not change the results (hardware not found on cold boot).


    Chris.

  • Autores is not on by default, as there are quite a few monitors that don't like it. The latest update to the firmware and configtool introduce a resolutoin-switch with the "R" key during the test-screen to make autores available to more monitors, but that won't change anything here.


    We also checked if a plain memory expansion has any influence on writing to the flash, but did not find anything. The one thing left to check is not too easy for us here in Europe, as I don't have a genuine NTSC board available for testing. The working hypothesis is that the higher base frequency and the "always start in NTSC, never see a PAL resolution" situation triggers a bug. I may be able to modify a board with a solder jumper and the right crystal, but work keeps piling up here since the start of the year (way too much administrative work). I fear it will at least take another week until I can take a short at such a modification and send it to Peter for testing (he's working in home office almost three hours away).

  • Autores is not on by default, as there are quite a few monitors that don't like it. The latest update to the firmware and configtool introduce a resolutoin-switch with the "R" key during the test-screen to make autores available to more monitors, but that won't change anything here.


    We also checked if a plain memory expansion has any influence on writing to the flash, but did not find anything. The one thing left to check is not too easy for us here in Europe, as I don't have a genuine NTSC board available for testing. The working hypothesis is that the higher base frequency and the "always start in NTSC, never see a PAL resolution" situation triggers a bug. I may be able to modify a board with a solder jumper and the right crystal, but work keeps piling up here since the start of the year (way too much administrative work). I fear it will at least take another week until I can take a short at such a modification and send it to Peter for testing (he's working in home office almost three hours away).

    I sincerely appreciate the heroics. If it was practical (and it certainly isn't), I'd send you my A1200 and let your team have a working unit to do failure analysis on. I have a scope but no logic analyzer. If your team needs to ask me for measurements, I can be reached at the email provided.

  • Hi guys, my card arrived back. There doesn't appear to be a description of whether the unit has been repaired, replaced, or "could not duplicate". Is there any explanation about what happened?

  • I'd have to look it up in the database, but we generally ship back what's working, and there's usually a short note on the return invoice (which is usually a 0,00 EUR-invoice).

  • The last reply was more than 365 days ago, this thread is most likely obsolete. It is recommended to create a new thread instead.