Graphics corruption

[rvo_nl]

I cant write applications.. otherwise I would.

@Karlos I did thoroughly clean connectors of bvision and power circuits. Sound is now at least always working, but the artifacts remain. Ive noticed they consist of 20 or-so lines spaced evenly across the screen. Everything they slice through becomes distorted. Most notably in menu's and interface buttons.

BTW, I never used serial so had no idea this was affected, too. My a1200 connects over PCMCIA.

Everything is now back in the tower and I've done a complete backup of the system, in case reinstalling doesnt help. I honestly dont know where to continue afterwards.. but we'll see.

[amiga1260]

Haven't you try to assemble your Amiga 1200 piece by piece?

Maybe there is a problem with one piece of hardware.

Did you also try without the IDE Express controller? Sometimes this piece of hardware can make your Amiga unstable. I saw this problem when I visit a friend of mine.

[Homer]

Thinking that nscaleworld had a real bad attitude that reminded me of "military spec A2000's", I checked out that name on the web. Take a look at http://nscaleworld.com/

Well, a quick WHOIS confirms it is indeed Doomy (Hans Campbell). What is the 2012 Amiga.Org position with him ?

[Karlos]

Well, a quick WHOIS confirms it is indeed Doomy (Hans Campbell). What is the 2012 Amiga.Org position with him ?

Until I am told otherwise by the admin, doomy is banned indefinitely. Some further checks will be necessary however.

-edit-

Checked and unsurprisingly, a big overlap with previously banned Doomy sockpuppets.

[Homer]

Well, if it is him, I would like to say that I am an Electrical Engineer that works with real trains, not childrens 'n' model ones. Am I now guilty of trolling Doomy ? :roflmao:

[rvo_nl]

please allow me to stay on topic with another update. I didnt get round to reinstalling 3.9 and for some reason I couldnt get the 4.0 installer to work, so that will come later I did however find out something that is quite interesting:

I only have artifacts in 16 bit modes. 8 bit and 24 bit screens have no troubles. I did use CGXmode to change the configuration for the main 16 bit mode I use to something on the very safe side, but that didnt make any difference.

What do you people think is wrong here? PPC? Bvision? Software? I still fear for the PPC..

[Karlos]

Do those corrupt vertical lines stay at the same position on the screen when you move the Edit Mode window, or do they stay at the same position *in* the Edit Mode window if you move it?

If they are fixed relative to the screen itself, this suggests a possible DAC problem or issues with the pixel clock for that screenmode.

If they get blitted around with the rest of the window, then something is definitely corrupting the Video RAM.

[rvo_nl]

they are fixed relative to the screen. they are always in the same position. even when I change the pixel clock for the mode myself. youre talking about the DAC, that is part of the bvision, right? how can it be that they all have the same issue? doesnt make sense.. interference? ..the kind that only appears in 16 bit modes? Im lost..

[Karlos]

they are fixed relative to the screen. they are always in the same position. even when I change the pixel clock for the mode myself. youre talking about the DAC, that is part of the bvision, right? how can it be that they all have the same issue? doesnt make sense.. interference? ..the kind that only appears in 16 bit modes? Im lost..

Yeah, it does seem strange, to be honest.

Can you tell us exactly which versions of all the CGX components you have?

Perhaps I can compare them with my CGX 4.2 install which I know is working just fine.

[danbeaver]

Fixed artifacts in certain modes = graphics board problems, an electrical problem if you will. Then of course one has to rule out confounding hardware issues such as a monitor "near out of sync" or similar. Think, "it doesn't move," so not software.

[danbeaver]

Well, if it is him, I would like to say that I am an Electrical Engineer that works with real trains, not childrens 'n' model ones. Am I now guilty of trolling Doomy ? :roflmao:

Clarity: n scale = tiny trains = tiny brains

[Karlos]

Fixed artifacts in certain modes = graphics board problems, an electrical problem if you will. Then of course one has to rule out confounding hardware issues such as a monitor "near out of sync" or similar. Think, "it doesn't move," so not software.

Since all the memory and pixel clocks on the Permedia2 are programmable, such issues can still be down to software and the improper configuration thereof.

[danbeaver]

Uh, the fact the the hardware is programmable does not make it a software problem any more so than a resistor can be "programmed" by varying the voltage applied. "Firmware" is software imbedded in hardware but it, like "software" runs on hardware. The software/firmware will not alter the properties of the transistors themselves.

Ya can't fix a bad resistor by changing the software.

[Karlos]

Uh, the fact the the hardware is programmable does not make it a software problem any more so than a resistor can be "programmed" by varying the voltage applied. "Firmware" is software imbedded in hardware but it, like "software" runs on hardware. The software/firmware will not alter the properties of the transistors themselves.

I think you misunderstood me. *If* the corruption is caused by an "almost out of sync" type problem, it could be symptomatic of a misconfigured clock. Trust me, I've poked around in those registers myself and gotten into trouble for it.

Three separate cards with the same, specific hardware defect seems a bit of a stretch.

[danbeaver]

One person's glitch is another man's feature.

(Ask any IBM salesman)

[delshay]

I am inclined to agree that it's a configuration (or power or both) issue simply because the chance that 3 successive bvision cards all suffer the same specific defect seems remote.

Make sure any overclocking tooltypes are disabled in the monitor driver (hint: 2AIDEMREPYMNONAFADELLATSNII, MELTDOWN)

If done correctly overclocking has nothing to do with this problem. I have finished the Ultra Bvision which is clocked 125Mhz and sets a new world record.

A Master Bvision is now also complete,but I am awaiting driver update for this card,but even without the driver update has set another world record with its Ultra Low Latency SGRAM (speed will never be revealed).

Both Bvision work on 37.5Mhz+ bus regardless if the PCI bus has jump ship.