SUCCESS!!! (EGA on Sanyo MBC-775 Internal Monitor)

[T-Squared]

I have done it! It took three video cards, but I seem to have found the correct type that outputs proper EGA! Paradise Autoswitch PEGA1A doesn't seem to follow the standard exactly. Now I sort of wish I had kept the Tecmar 480 EGA Master video card I bought...

Instead, what I have in my system is a Twinhead CT-8090 EGA video card, and it seems to work fine! (Though it badly needed a new 32-pin feature connector socket, and some solder-resist touchup beforehand.) I'm a bit disappointed that I can't get the 64-color-palette-based games to work, but I'm glad that EGA works nonetheless!

It's a bit buggy, though. I may have to adjust the intensity pots on the CRT socket harness, because the colors are not as distinct as I'd like them to be. (as you can see on the Space Quest example, I'm certain there's supposed to be a light blue in that logo somewhere.) Something may be funky with the intensity bit. I seem to be having that problem with all of my games too. What would be light yellow seems to be dark yellow (technically brown), so if I'm assuming correctly, all of the other colors seem to be muted as well.

Unfortunately, I also don't have the proper amount of memory to run PCPaint in EGA mode. (241kb, while I have 256kb at the moment with the interim board that I'm using.)

If it comes to it, I may be selling off that Paradise EGA card.

 

[chjmartin2]

Awesome!

 

[T-Squared]

Looking at the CRT drive board, there are glued RGB pots, as well as a B-Drive pot.
(I've seen the B-Drive in one Youtube video, but I can't, for the life of me, remember what this is for. Is it an input voltage for the CRT?)

Also, something that just came to mind... would there be a contrast dial to adjust these color intensities? I'm noticing that all of the colors, light and dark, are not very distinct. There's only a sliver of color difference between the two intensities.

Since that's almost out of the way, the only thing I need to really focus on is getting the 640k memory decoding/addressing working.

 

[T-Squared]

Yeah, there's something weird with this CRT's intensity bit. With the intensity wire, the color does change, but it doesn't seem to change enough to produce enough of a distinction. (Taking out the intensity connector did make only the brightest 8 colors show up, though.)

I did find a circuit that seems to produce the proper brown color, if it need be, (Not sure if it will work, but I could try.)

Turning the contrast dial didn't do anything substantial.

 

[Eudimorphodon]

I'm a bit disappointed that I can't get the 64-color-palette-based games to work, but I'm glad that EGA works nonetheless!

Is the internal monitor capable of displaying the 350 line EGA modes? Because so far as I’m aware there is no such thing as a “64 color palette” game that runs in 200 line mode.

(okay, I googled it, there is, well, one, that unofficially supports it but it only works on nonstandard monitors. Conventional EGA monitors interpret their input as RGBI in 200 line mode, they literally don’t care about the state of the secondary color lines that don’t correspond with Intensity in the CGA pin out.)

 

[T-Squared]

Ah, I see. (What I meant was any EGA game using 16 colors other than the simple base RGBI color palette, anything within the 64-color palette that the EGA standard provides.)

I may have to add a circuit to the monitor as well, because it doesn't seem to do the CGA+gray+brown colors properly. (although that would require an IC that I would have to buy, since I don't have it in my collection of components)

 

[kdr]

Some CGA monitors (in particular the IBM 5153) use the Intensity bit to control the contrast - when the I bit is inactive, contrast is set by the knob on the front; when the I bit is active, contrast is forced to 100%. On that kind of monitor, setting the contrast knob too high will remove the distinction between the regular and intense colours. Maybe something similar is happening with your display?

 

[T-Squared]

Huh. I guess it does support them, looking back at the Checkit Diagnostics. But I'm still not quite seeing a distinction on some colors. Maybe if I try using the contrast pot while the brightness is turned down?

What is the diagnostic program that uses the standard color block display and text labels that I've seen so many times? (I know Checkit 1990 has one, but I can't use that because of the lack of memory in my system.)

 

[retro-pc_user]

I have to say that is amazing to see EGA on the Sanyo MBC-775 and if there was a way to add a color display inside of a Compaq Portable 1 system, that would be amazing as well since I have an EGA card installed (haven't tested the Genoa Super EGA 4880 in my Compaq since I had adjustments that were needed on the card and it wasn't working very well from the start).

I also have a Paradise Basic EGA card in the Compaq system, which does work with the internal monitor after making a logic board to support the internal monitor, whereas the Genoa is in my Wyse WY-2108.

 

[T-Squared]

I'd call that low-intensity "yellow" a brown color, wouldn't you?

Also, see what I mean about the colors not being quite distinct? Also, if it's not the way it's supposed to display, I'm suspecting that the dark gray color is also missing. (EDIT: Looking at it, they're more distinct on-camera. Somehow, though, they are not, to my eyes, in real life. The light gray color also resembles the white.)

@retro-pc-user: Try looking up a color tube of the same size on eBay, because there are also universal TTL CRT control chasses (the flyback transformer, the color inputs, CRT focus controls, etc. all in one package)

 

[T-Squared]

Nope, I found a problem. Not only is there a color contrast imbalance in most of the colors (including what appears to be a not-brown dark yellow), but I just now saw the dark gray color on the monitor. In the image above, the dark gray is invisible because of the brightness of the surrounding colors, I found it while trying to find a way to bring what I thought was the balance back. I was adjusting the bias on one of the colors, when the text between the gray and blue columns in the center popped out. Yeesh.

 

[T-Squared]

Yeah, so it seems I have the "naive" version of the CGA color palette:

https://www.aceinnova.com/en/electronics/cga-and-the-brown-color-in-ibm-5153-color-display/

 

[T-Squared]

Actually, now that I think about it, I may not have a naive version of the CGA color palette. In fact, it may actually be an oddity. The brown may actually be the one being driven properly, and the dark gray is the one that isn't. The way I see it, there's two LS05s on the original 5153 schematic, and only one in my Sanyo. The gray is the only one that is non-existent when the screen is black, and the brown color looks extremely close.

 

[T-Squared]

Indeed, it seems like the Sanyo MBC-775 Is an oddity when it comes to the CGA standard! Thanks to one piece of info from Hugo Holden's little document on upgrading the 5153 (IC Label Q201 on the schematic), it seems that the Sanyo MBC-775 only uses one LS05 IC to fix the brown color, and not the dark gray color! I'll have to map out the area around it, but I think I may have found my missing color problem! Plus, thanks to that, I was also able to map out which LS05 drives which color on my schematic! Now I have to delete the brown color connection, and then figure out what to do to contrast the colors more. (They aren't bad, but I could use some brightness on the lighter colors.)

The 5153 has an extra LS05 to drive the dark gray color when the intensity bit is set. Normally the intensity bit is ignored for black, giving only 15 colors out of 16. (Like the ZX Spectrum did)

 

[VileR]

For what it's worth, the brown in your photo looks very similar to the brown on my IBM 5153. It's not the "naive" dark yellow (#AAAA00), but not exactly the "canonical" brown (#AA5500) either... rather something close to the latter, but still in between.

About translating RGBI colors to 8bits/channel (s)RGB: the general consensus is to map each channel to values of 00, 55, AA or FF hex; but it's worth remembering that this scheme derives from IBM's translation of CGA colors to the *VGA* palette (although the latter was 6 bits/channel). So it's canonical in the sense that it follows exactly what IBM did with VGA, but that in itself may not have been a 100% accurate rendering of the 5153's color generation. Ever since getting my 5153 earlier this year (and seeing enough decent photos of other 5153s showing the same general appearance), I'm becoming more and more convinced that this is the case.

 

[T-Squared]

Yeah, that's what I thought. It was odd. It looked like brown in the picture, and yet, there's not a good contrast between the normal and intense color modes, so I figured it was a trick of the camera.

It actually may be simple enough to find the PCB points needed to put Hugo's detector board together for my system.

 

[T-Squared]

Actually, now that I think about it, is there a simple way to adjust the color contrast? (Like changing capacitors, or resistors?)

 

[T-Squared]

Again, Hugo Holden comes to the rescue! I know this is simplified, but I'm sure I can figure out what to do about this. The problem I'm having is that the colors on my system are indeed working properly (sans the Dark Grey), but the intense colors are too close in vibrancy to the original non-intense colors. Like I've said before, they probably only used the maximum vibrancy (i.e. FF0000, 00FF00, 0000FF, rather than FF5555, 55FF55, 5555FF) as the baseline for the max brightness, so the colors need to be amplified more.

 

[T-Squared]

Okay, if its anything to go by, looking at the simplified schematic, just to guess, to get a brighter color on my system, the resistors in line with the intensity line need to be swapped out for a lower value; the intensity line adds current to get brighter colors. (With the schematic here being the goal I need to get to)

OR... The RGB lines need to have higher-valued resistors in order to specify the darker color.

I may be wrong, but that's all I can guess right now.