Indivision AGA MK3 - No Signal

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.
  • Hi,

    I just got my Indivision AGA MK3 and ran into the no-signal error which was discussed on the board here a few times already.

    I installed the Indivision on the board, with no additional load (no floppy, no keyboard). The setup is listed below:

    - A1200 REV 1D.4 (recapped)

    - Original Commodore PSM-601-2 (25W, 5V-3A)

    - DELL P1917S (tried both VGA and HDMI, for sure not simultaniously)


    I cleaned the Lisa chip connectors with air and isopropyl alcohol and remounted the indivision quite a few times without any success. After reading most threads here I also took out my scope and started measuring.


    As I do not have a test load for the PSU, I measured the +5V, GND on the floppy connector, while the indivision was mounted. The result shows a ripple of max. 10mv around 5.03 V, which is well within the specs (tmbk). Still, a second opinion would help.


    Further I measured the 22-pin connector of the indivision. Between Pin 1 and 2 I get 4.87 V with a ripple similar to the floppy connector. The reduced voltage shall be ok, according to wiki. Unfortunetelly no other PIN has any signal against A-GND (Pin1), except Pin 15 which shows a ripple-free 4.3V constant voltage, but no sync as intended.


    Important note about my setup


    I do not have any analog 15khz capable monitor to test my Amiga RGB output, but sure I measured with the scope as well and got the color signals and syncs. Even the digital RGB is there. I even went a step further and decoded the Green channel recording with a Python script into an actual image, which looks quite like the expected boot screen.


    On the other hand, I could already proove that my FBAS (Composite) is down. More precisely the composite outputs only the chroma signal and my (hopefully educated) guess is, that the delay line or the 1145M itself passed away. I could not find anything about it, but I do not expect any cross implication to Lisa and still, RGB is working.


    I am thankful for any help.


    Attachments (scope measures):




  • This might be a startup problem of the flicker fixer. We mostly have that with power supplies that don't have a clean startup-ramp.


    Do you have the possibility to run the MK3reset tool? Maybe a second Amiga where you can create a rescue disk?


    BTW, the voltage drop on the 5V output of the Indivision output is due to a Schottky diode that is meant to protect the Amiga from voltage that is sent back from the monitor.


    Jens

  • Thank you Jens for the prompt response. I actually wanted to get around to get a Gotek on the rush, because I am waiting for an original floppy to appear on the market (mine is not functional unfortunately.


    Can recapping improve the startup ramp of my PSU or is this something inherent to the PSU model I have? I actually captured part of the startup in Floppy5V_2ms_div.png, but could also capture the entire startup once if this would help investigating.

  • Can recapping improve the startup ramp of my PSU or is this something inherent to the PSU model I have?

    I don't have a definitive answer to this being a PSU problem or not. I only have a few reports about customers who bought a CA-PSU where the effect of "requires MK3reset in startup-sequence" went away with a proper PSU. However, I was never able to see and measure the effect in person, as not a single customer was willing to send me his board+PSU. I'm pushing those customers, though, and one of them has actually shipped an aftermarket-PSU to me that is currently in transit from Norway, so there's some movement on this front.


    Further, we don't even know if your case is also a "FPGA startup is trashed"-case. Only running the MK3 reset tool or the rescue disk will tell.


    Re-Capping will of course help, but I wouldn't pull out the soldering iron just yet, as you haven't even done the most basic diagnostics. You should also at least connect the defective floppy drive and listen if it starts to click in the typical 2.5-second rhythm while the computer is waiting for a disk to be inserted.


    Where are you located? I do have a few internal floppy drives that are originally packed&sealed from a Commodore service company.


    Jens

  • Hello Jens,


    thanks again for the detailed response.


    I agree that diagnostocs on my side should be finished before considering any further action. Anyhow, I have attached a full capture of the startup which has an interesting knee after 15ms at 4.68 V. I will not assume anything more on these measures until I can run a full diagnostic as recommended.


    The floppy drive clicks as expected, starting a few seconds after power on. But well, it was sold to me as a defective drive and I have no doubt it doesn't work.


    Quote

    Where are you located? I do have a few internal floppy drives that are originally packed&sealed from a Commodore service company.

    I am located in Hamburg, DE. If this is a hint, that you want to sell one of those, I would like to take one for sure ;-)


    Have a nice evening.



  • UPDATE: I got it working! BUT, it does not survive a cold boot.


    What did I do:

    - I created a RDB image with a single bootable partition and copied the content of the rescue disk over (yes, I lack a working floppy for now)

    - I configured my PiStorm for the very first time (Musashi)

    - I started the Amiga with my monitor connected to the Indivision

    - I signed into the Pi via SSH and started emulation with my rescue image


    As a result, the rescue system somehow managed to output the log below to the ssh console and my monitor turned on showing the same within an AmigaDOS window.


    flash sig: 53464450 - ver: 1.0

    Loading firmware file, 177610 bytes...

    About to start flashing.

    Press CTRL-C to abort,

    Waiting 7 seconds...

    Flashing...

    Done.


    Run the GUI tool from WB to restore last saved config,

    or flashtool ? from command line.

    Press CTRL-C or mouse button to continue.


    From now on, I can quit the emulation, change the image, boot into workbench etc. BUT only until the next powercycle.

    I think it gets more valid to assume a PSU issue?

  • UPDATE: I got it working! BUT, it does not survive a cold boot.


    What did I do:

    - I created a RDB image with a single bootable partition and copied the content of the rescue disk over (yes, I lack a working floppy for now)

    - I configured my PiStorm for the very first time (Musashi)

    - I started the Amiga with my monitor connected to the Indivision

    - I signed into the Pi via SSH and started emulation with my rescue image


    As a result, the rescue system somehow managed to output the log below to the ssh console and my monitor turned on showing the same within an AmigaDOS window.


    flash sig: 53464450 - ver: 1.0

    Loading firmware file, 177610 bytes...

    About to start flashing.

    Press CTRL-C to abort,

    Waiting 7 seconds...

    Flashing...

    Done.


    Run the GUI tool from WB to restore last saved config,

    or flashtool ? from command line.

    Press CTRL-C or mouse button to continue.


    From now on, I can quit the emulation, change the image, boot into workbench etc. BUT only until the next powercycle.

    I think it gets more valid to assume a PSU issue?

  • I think it gets more valid to assume a PSU issue?

    At least it's the same "FPGA does not start on power-up" issue that other customers are reporting. I have strong indicators that it's a PSU startup issue, but no hard proof yet.


    I did get your eMail (in perfect German - why are you writing in the English forum here?), will answer that in a few minutes.


    Jens