Setting up GBS-8219 CGA to VGA Converter

[Eudimorphodon]

Type in and run the following basic program in BASICA and see what you get:

10 for x=0 to 7
20 color x : print “this is normal “;
30 color x+8 : print “this is intense”
40 next x

(20 ends with a semicolon and 30 doesn’t on purpose.)

If the manual is correct the problem will immediately become apparent.

 

[xpower]

Type in and run the following basic program in BASICA and see what you get:

10 for x=0 to 7
20 color x : print “this is normal “;
30 color x+8 : print “this is intense”
40 next x

(20 ends with a semicolon and 30 doesn’t on purpose.)

If the manual is correct the problem will immediately become apparent.

 

[Chuck(G)]

On a 5153, that "normal" yellow is brown--there is circuitry in the 5153 to handle exactly that case. The EGA version does a similar trick.

 

[Eudimorphodon]

So the bigger problem is there is no distinction between “normal” and “intense” at all, as the photo demonstrates. There should be 15 different colors of text on that screen (the first “normal” is black on black, the first “intense” should be a dark gray), there are only seven. So this thing doesn’t do CGA, meaning “RGBI”, at all.

FWIW, the 7 colors you are getting are roughly correct for the “intense” versions.

 

[maxtherabbit]

Were there ever any devices that used 3 bit TTL RGB? Seems like a bizarre product that doesn't support any extant standard...

 

[Chuck(G)]

Regardless, most converters that I've seen don't take the "I" input. However, driving them with TTL levels will always drive the colors to saturation. You really need an external circuit to convert to RS-170 levels, which, if memory serves is <1.0V peak.

There are several such DIY projects around. Google "RGBI to RGBS" for some hits. Generally, they're pretty simple and can even be built into the D-sub hood of a cable.

Mattis Lind has a nice project, complete with PCB layout.

 

[Eudimorphodon]

Were there ever any devices that used 3 bit TTL RGB? Seems like a bizarre product that doesn't support any extant standard...

Pretty sure there were, yeah, but you mostly have to go back to the 1970’s to find them.

The real issue here is that I bet if you crack this device open you’d find it uses almost exactly, if not the exact, analog-centric scaler chip as those cheap GBS boards, and to implement “digital” support they took an incredibly minimalistic approach, possibly just adding some isolation so a TTL level input won’t burn out the input lines. (*) They also tacked on official support for separate sync in addition to composite, but I’m pretty sure the cheap boards already have the hardware support “present”. These results are exactly what you’d get from connecting one of those cheapo boards to a CGA card with just resistors on RGB. They *could* have built a proper digital converter into it with just a tiny handful of parts but couldn’t be bothered.

I think they’re getting away with it because they’re marketing this to people that need a fix for their 80’s CNC mill or whatever, they don’t care if the colors are wrong.

(*) Or like Chuck said, maybe they literally did nothing in hardware and they’re just letting the 5v input saturate the analog input. The “digital” option on the menu is either a no-op or it just tweaks some gain/contrast settings so it’s not completely washed out.

 

[maxtherabbit]

The regular gbs 8200 absolutely does support separate sync in hardware, it works perfectly with the GBS-Control firmware

 

[Eudimorphodon]

… re: 8-color RGB, I checked and my vague recollection was right, the Sanyo MBC-550 used digital RGB-no-I. Pretty sure the same system wasn’t that uncommon on those various fancy Z-80 machines with color monitors coming out of Japan in the early 80’s, like the earlier NEC PC-88s.

 

[bladamson]

The regular gbs 8200 absolutely does support separate sync in hardware, it works perfectly with the GBS-Control firmware

I keep meaning to mod one of my boards and try gbs-control out, but..... Yet another of those 1000000 projects that I never get around to, lol.

 

[xpower]

The MCE2VGA converter arrived.
It was the Motherboard and pre-programmed FGPA board Fully tested with the 3D printed case.



The box contained all the parts needed to assemble the only thing that was missing was a mini USB to power the board.



There was no instructions but assembly was easy the only issue was that the screws to attach the board to case was slightly obscured by the lid's screw wholes. My computer toolkit didn't have a screwdriver that would fit those screws but managed to tighten the screws enough using a flat head.



Running the basic test produced the difference between Normal and Intense colors.


The only issues I've had is in Hercules graphics mode the left of the screen is cut off. Changing the vertical and horizontal adjustments doesn't make it viewable.
Another issue which can just been seen on the Text programme box is there is shimmering going down the image. Changing the phase option causes the shimmers to move but doesn't resolve them.



As a comparison this is the same screen in TGA mode

 

[bladamson]

Nice, I am glad you got it going.

IIRC Hercules graphics have a 720 px horizontal resolution, but it runs at 50hz so I *think* the dot clock still ought to work with a VGA monitor. Does it do the same thing on other monitors? Maybe the back porch or whatever it's called is a little wonky and that monitor just doesn't like it.