Talk:Motorola 6847

Latest comment: 1 year ago by 4throck in topic Colorset and available colors issues

Differences between the AMI S68047 and the Motorola MC6847

edit

Its not sure whether the MC4847 was a clone of the AMI S68047 or vice versa, but the chips were very much alike, only with different colors for text mode and some semigraphics mode, the MC6847 used a green background for text (as can be seen in TRS-80 Color computers which used the MC6847, and many other motorola chip, as the CoCo was a Motorola design, not a Tandy design

Translation from korean site http://retrospc.blogspot.com (memories of the SPC-1000) :

(Note: The colors in the table are randomly added by me based on the color name. The actual expression may differ.)

After discovering that the S68047 is functionally identical to the MC6847 but with a different text color, I took a closer look. As a result, he noticed quite a few differences. The colors of the semi-graphics are also different, and there are modes with different resolutions in the graphic mode (128x192 Resolution Graphics-3 vs 256x96 Graphics 5). The color palette and background color in graphics mode are also different. And 128x64 4-color graphics mode (Color Graphics-2 and Graphics 2) and 128x96 4-color graphics mode (Color Graphics-3 and Graphics 4) have the GM2--GM0 settings staggered. The graphic modes in these problems are either hardware not available in SPC-1000 (when GM2=0), or the SCREEN command is not supported in HuBasic (because different resolutions are obtained with the same signal in both VDGs, this is I think the mod is blocked).

For graphic modes that are not supported by HuBasic, use GMODE (Address I/O 0x2000).

OUT &h2000, &h08 ' 128x96 Color Graphics-3 (6847) or 128x64 Graphics 2 (68047), palette #0 OUT &h2000, &h88 ' 128x96 Color Graphics-3 (6847) or 128x64 Graphics 2 (68047), palette #1

OUT &h2000, &h0C ' 128x192 Resolution Graphics-3 (6847) or 256x96 Graphics 5 (68047), palette #0 OUT &h2000, &h8C ' 128x192 Resolution Graphics-3 (6847) or 256x96 Graphics 5 (68047), palette #1

I think you will be able to enter it by doing so (it is necessary to check whether it can be done in MESS and in the actual machine).

MC6847T1?

edit

Would it be appropraiate to add details about the T1 version of this chip on this page, versus making a separate page just for it? Allen Huffman (talk) 04:59, 30 November 2021 (UTC)Reply

Colorset and available colors issues

edit

There is a lot of inaccuracy in this article around the colors which the 6847 can output.

I have corrected the initial dubious count of possible colors and clarified the circumstances which make available the 'extra' alphanumeric character mode colors somewhat.

However, the article omits detail on the light orange color entirely from all of the later tables, and substitutes the incorrect core-colorset orange in some places where it is not appropriate. The YPBPr signal levels table also appears to be extremely dubious, and the shown values can't be ascribed to the datasheet as a source. Further research and correction / verification is needed here. Bluearcus (talk) 17:55, 17 February 2023 (UTC)Reply

The YPbPr table is straight from the datasheet, coming from this link https://www.datasheetarchive.com/pdf/download.php?id=76ae30fa50e5b4996304d289f7fb181a685f9b&type=M&term=MC6847 .
See table "DC (STATIC) CHARACTERISTICS " on page 3 of the datasheet. Lines "Chroma A Voltage", "Chroma B Voltage" and "Luminance Y Voltage" show signal levels. The YPbPr table show the "typical values" columns (unless there's some copy/paste mistake).
I'll try to update the reference to make this source clearer. But there's no doubt here, the values are copy/paste from the datasheet.
As for the conversion to sRGB colors it's just an approximation for illustration purposes. 4throck (talk) 11:14, 18 February 2023 (UTC)Reply

I'm a little unclear as to how the colour YPbPr per colour table can be attributed to the datasheet when the levels are only ever listed in isolation and never in a colour X has levels ABC form. There are obviously dubious elevenses in this made up table, for example true black and dark green / dark orange having the same Y value. I'm unable to find a table of this form in the datasheet and can only assume its been made up. Bluearcus (talk) 19:34, 2 March 2023 (UTC)Reply

Indeed, the values are only shown in isolation, because all the colors are generated by combination. All you have to start with is Y 0.72V, 0.65V, 0.54V, 0.42V and chroma A (R-Y or Pr) or B (B-Y or Pb) 1.0V, 1.5V, 2.0V. There are no other values. So my claim to the values being from the datasheet is only related to these starting values.
The table is a simple listing of permutations, where I tried to attribute the most likely corresponding color. The problem here is that the values are not normalized and seem to be inverted, so I've assumed that R-Y 2.0 B-Y 1.5 is blue and R-Y 1.5 B-Y 2.0 is red. You can call it guesswork, no problem in admitting that. But it doesn't change that somehow those initial voltages generate all the colours ;-)
Now, to make the article clearer, feel free to correct or remove the table from the article, no worries.
I'll explore this further on my sandbox and keep discussing it with you. I'll try to convert those values to normal ranges and into RGB and see what I get. 4throck (talk) 23:22, 2 March 2023 (UTC)Reply