Apple IIGS (ROM1) and AppleColor Monitor 100

You'll never get the performance you're expecting with a IIGS and a monitor connected to the composite port.

It will work on if it is in IIe mode, but in IIGS mode, it will always look wrong, blurry and weird colours.

My solution to this problem is to use an old 4:3 VGA monitor, an el-cheapo Gonbes GBS8200 video converter ($20 on amazon) and modify it with an ESP8266 controller and the "GBS Control" firmware (it's on Github).  The resultant picture is excellent and the total cash outlay was less than $50 including the $10 monitor from a thrift store.

It did occur to me that I was unable to find any pictures online of a AppleColor Monitor 100 definitively working correctly connected to a IIGS, so maye this setup never really worked.

The setup was known to Apple and they published a technical article on the subject, which reiterates that the SW100 needs to be in the upper position.

The switch is now in the RGB I (bottom position)

That's all you need to change: put it in the top position.

For a more technical explanation, you need to learn about how color monitors on early computers operated. There were multiple ways of representing color on an electrical level. The AppleColor 100, used with the Apple IIe Extended 80Col/Color card, uses a scheme called "digital RGBI". The Apple IIgs uses a scheme called "analog RGB video". These are quite different in terms of signal levels and even the number of signals. Digital RGBI, for instance, uses four pins to indicate color to the display, each assuming TTL true-false levels: red, green, blue and intensity. Four pins, each of which can be either high or low (two states), makes 2^4 or 16 possible colors, and no more.

A computer producing analog RGB, such as the IIgs, has no way to output any intensity signal. It is just not there in the video hardware. It is not required, because the red, green, and blue pins can assume any number of states between 0 V and 0.7 V: each color component is a true analog value, so an infinite number of different colors is, in theory, possible. The computer's digital color data is converted to this analog signal using a video DAC (digital to analog converter). The IIgs can produce 4096 different colors.

The AppleColor 100 monitor looks to have been designed with a view to provide compatibility with future video formats, and its SW100 switch enables it to operate with analog RGB systems. Absent a schematic, the function of the switch appears to be to pass the r,g,b inputs directly to the channel amplifiers when it is in the top position (labeled RGB2,4,8 for reasons that are obscure. It would make more sense to label the mode "RGB2,5,9" since the color channels are on those pins of the cable). In the bottom position, the 4 input signals are weighted and mixed according to the color palette shown in the first photo, which does not allow for analog colors.

From what I vaguely remember from back in the day the color output from the Apple IIGS is analog.

Well, it is possible that I'm barking up the wrong tree because without schematics it's just guesswork.

One thing I noticed while researching this subject is that there is an inconsistency in the pins assigned to red and green channels. The AppleColor 100 and the IIe 80col/RGB assign Green to pin 2 and Red to pin 5, but the IIgs and the (analog) Apple RGB Monitor assign Red to pin 2 and Green to pin 5. So unless these pins are swapped in the cable, mixing generations of color displays will always give the wrong colors. It must have been a bad day at the office when this cockup was discovered (the technical note from Apple doesn't even mention this problem, so it seems they never tested the combination and weren't aware of it).

To see how the IIe RGB card (digital RGBI) can be adapted to use the Apple RGB Monitor (analog), see this tutorial.

In a way, your problem is just the opposite.

By the way, are you able to read the AppleService Technical Procedures - AppleColor Monitor 100.pdf ? I just see blank pages.

robespierre wrote:

One thing I noticed while researching this subject is that there is an inconsistency in the pins assigned to red and green channels.

 

I wondered that myself. Looking at the resulting dislpay that could explain some of the colors. But the LBLU bar is actually yellow (red+grn) so swapping them would still give yellow which is totally wrong. I suspect it is something else.

Looking more closely at the color bars, it appears that there is no "analog" setting. The RGB1,2,4,8 is a digital 16-color RGB mode, and RGBI is another digital mode where the fourth bit is interpreted as Intensity.

1. First, note that a IIgs never outputs Intensity. The pin (10) is not connected at its video output port.
2. So if the signal from the IIgs to the monitor is interpreted as digital RGB, there only can be 8 colors displayed. Both sets of bars from the IIgs show 8 colors. When the IIgs tries to show a dim color, it will simply appear black since a digital signal has only two allowed states.
3. The red-green confusion is apparent with the IIgs feeding the RGBI mode. The bar for 'GREN' is shown in red, for instance. 'PURP', which is red plus blue, is shown as aqua, green plus blue. 'AQUA', which is blue plus green, is shown red also. You shouldn't get confused by this not being magenta: remember that in analog mode, each color channel is any value between 0 and 0.7 V. So even though blue is present doesn't guarantee that it is above the level needed by a TTL gate to register as true.
4. Likewise, 'YELL', which is red plus green, is shown as magenta (red plus blue). Why? Because a lighter analog color contains all three primaries. Similarly 'LBLU' is shown as yellow (red plus green). But the light blue color contains all three primaries.
5. Based on this data we can infer that either the analog signals for red are "hotter" than green, likewise hotter than blue, or that the TTL gates in the monitor that register these signals have a lower threshold for red. Maybe a combination of the two.
6. Turning our attention to the IIe color bars, it's apparent that neither mode interprets them as "analog", since there is no Intensity signal in analog color. The IIe outputs 16 different combinations of the four signals, so an analog display would show those as 8 colors repeated twice.
7. In the RGBI mode, the upper and lower (without and with Intensity set) are dimmer and brighter versions of the same 8 colors, as would be expected if the Intensity bit controlled brightness. But something else striking is that the color below Black is still just black. In a typical RGBI screen, the Intensity bit is mixed in with each channel which would make this display dark gray. So even with the switch in this position, the AppleColor 100 evidently does not act like other RGBI monitors.
8. In the RGB1,2,4,8 mode, the IIe color bars look pretty close. But the screen fails to show a convincing Magenta, and the two Gray bars are strongly tinted and not actual gray. There is no particular relationship between the upper and lower bars in this mode, other than perhaps Pink being a brighter Purple and Light Green being a brighter Dark Green.
9. This implies what "RGB1,2,4,8" means: there really is no "red" or "blue" signal! The four "color" signals are just arbitrary bits indexing into a color map inside the monitor. As the IIe color bars are displayed, from left to right, the computer is just counting through a 4-bit register beginning with '0000' and ending up at '1111'. The bits don't correspond to primary colors; they are just binary digits of the color number in the IIe palette.
10. By the correspondence of colors shown by the IIe in RGBI mode to RGB1,2,4,8 mode, it's evident that the pin documented as "red" is actually color weight 1, the pin documented as "green" is actually color weight 2, so on for "blue" weight 4, and "intensity" weight 8. The names are deceptive and conceal their real purpose. The labeling of the switch is, however, very accurate as to its function!
11. So the tutorial I linked earlier, showing a IIe RGB card adapted to an analog RGB display, is sorely mistaken, because the meaning of the signals output from the card are not primary colors.

From the above we can see that the SW100 has a strange purpose which is totally unrelated to support for analog RGB. Either the Apple memorandum was completely wrong, or is there some other switch inside somewhere? It describes "a slide switch labeled '16 color/8 color' on the main PC board, almost directly under the neck of the CRT". Does that describe a different revision?

Also confusing is the specification in the AppleColor 100's manual: "16 color semi-analog input; 8 color TTL RGB input; (negative-going TTL level composite synch switch set at factory)".  What "synch switch"? Looks like the technical staff sent in some verbiage that got mangled in the editing process.

Unfortunately, it looks like the Technical Procedures manual is encrypted.

At one time I had a third  party RGB card in a IIe that gave the wrong colours on a IIgs monitor. I ended up making a custom cable for it.

Apparently tis possible judging by these photos

Apple IIGS RGB to AppleColor Monitor 100

This is a rare digital RGB monitor, capable of showing some 16 colors. //GS has analogue RGB output capable of showing 4096 colors on analogue monitor. I recommend usage of the appropriate much cheaper and not rare //GS monitor, or one that supports by design analogue RGB input  (some SONY PVM,  some NEC MultiSync, others with SCART interface). These pictures are very doubtful -- why the second/right side magenta stipe is darker than the left one?

Yep it is hard to beat the official Apple IIGS RGB display. 

Not my photos, but another angle showing the AppleColor Monitor 100 with IIGS - the magenta stripe looks about the same in this one.  It's unclear how the connection was accomplished. 

Apple IIGS RGB to AppleColor Monitor 100 (#2 screenshot)

It looks still like it's interpreting it as digital RGB. Strange. It could also be just that the pinout is not correct from the IIgs to that monitor.