Hi -
I ran into a problem with my Apple II (a Taiwanese clone, my original Apple IIe still is in disrepair). The problem is that its color killer circuit is not good enough so I always get color artifacts in text screens. My Apple IIc does not have the problem with that same TV which appears to have a very sensitive color burst amplifier. Almost no burst amplitude, just a bit of wiggles, barely visible on the scope, are enough to get full color ! Alas, the phase lock on such a feeble signal is imperfect, and so the color in the text shifts, making the problem worse. Much like a LSD trip. Text with always shifting and moving colors all over the rainbow. No good for focused work on software development ! How people could have worked under these conditions on the Apple II back in the day eludes me. Must have been an ordeal. Apple IIc seems to have a 'color killer' circuit which works. Alas, it's a hybrid circuit in which everything is hidden under a layer of coating. And I don't want to destroy a perfectly good Apple IIc just to find out what is inside it.
So this question arises: did anyone analyze the 'color killer' circuit of the Apple IIc ? Are there any good circuits out there who can do the job ? --- Note that I tried the known mod to gate off the color burst in the digital logic, and it is gated off in text lines, but still the damn TV thinks there is a color burst. I can see in the oscilloscope that regular hires vertical stripe patterns can excite the LC tank in the Apple II color circuit during the 'live scan' portion of the TV scan lines, and at the moment I have the hypothesis that the whole machine is so polluted with harmonics and subharmonics of the 14.31818 MHz master clock that this LC tank may pick up some of that crap and turn it into a 'color burst' signal all the time, and the damn TV picks it up. So it's not a trivial solution which will bring good B&W text. At the moment I have lifted one leg of the inductor to disable the LC tank, and then I get B&W text, but this is not a solution as it also deprives me of color graphics.
Any ideas / links to good 'color killer' circuits ?
- Uncle Bernie
Perhaps not the answer you're looking for, but have you considered modifying the TV instead to give it a color/b&w switch? The necessary signals should already be there for compatibility with over-the-air black and white broadcasts.
I know you're a stickler for details, so I'll start by saying that I hate when poeple refer to a "color killer" circuit in the Apple II. The only circuit that deserves that name is actually in the TV/monitor; it is a colorburst detector that turns off the color decoding for monochrome video signals. So let's agree that a typical Apple II video signal ALWAYS has color (chroma) in it, albeit as an artifact of the video signal being switched at the color subcarrier frequency.
So now, when we want to supress that color information, we need to remove the color burst signal and expect that the TV/monitor will then stop decoding it. The early Apple motherboards (after Rev 0, which had no provision to remove colorburst) did this by grounding the video signal during the colorburst time using a simple transistor. In some cases the transistor is not driven hard enough and some color burst can remain which confuses the monitor.
Later boards, including the //c, do this by only gating in the colorburst signal when displaying graphics. This should be sufficient to work correctly with most TV/monitors. In your case however, there appears to be enough stray subcarrier floating around to trigger the monitor. I'd say your options are the following: 1) turn down the COLOR control of the monitor when displaying text only; 2) try turning down the video level control of the clone (if it has one); 3) see if you can eliminate the stray subcarrier with some filtering; or 4) perhaps add an additional burst killing switch as close to the output jack as possible to clamp any stray subcarrier present there.
Hope this helps.
In post #3, JeffMazur wrote:
"I'd say your options are the following: 1) turn down the COLOR control of the monitor when displaying text only; 2) try turning down the video level control of the clone (if it has one); 3) see if you can eliminate the stray subcarrier with some filtering;"
Uncle Bernie comments:
Thanks for these ideas. Essentially, they confirm my own lines of thought. I already made a weak attempt to do the filtering, but it did not help much. It helped a little, so with adjusting the video level down, I could get a B&W text picture with only occasional color "assaults", i.e. when I give the monitor program a 'hexdump' or 'disassembly' command, and while the screen scrolls, the unwanted colors are back. But then, when it idles waiting for a key input, it reverts to the wanted B&W.
This would be enough to avoid driving me nuts when working on software development in that Apple II. But being a perfectionist, I sought a full remedy for the problem. From the sparingly responses, although competent, it now appears that no really good solution ever has been found, at least for the first generations of Apple II, which were extremely noisy EMI wise. I don't want to dare a guess how much $$$ Apple (the corporation) had to sink into that engineering rathole to get them FCC clean and legal to continue selling them.
- Uncle Bernie
Besides reduced chip count driving costs down, one of the main reasons Apple came out with the //e was to build a more FCC compliant machine. The ][+ had been jerry rigged for compliance about as far as they could. Replacing a whole lot of 74LS chips with two big ASICs was a win on both fronts. It was also more close resistant, albeit it really didn't take the Tiawanese and Hong Kong based cloners long to build thier own equivalents of the IOU and MMU chips. By mid 1984 there were a number of //e clones on the market.
Note that Apple themselves called the circuit, the "color killer modification" - see the June 1979 issue of Apple's Contact Magazine. The reason it doesn't always work is that enough of the color burst signal will leak through the color killer tranistor even whe nthe transistor is switched off to enable some color monitors to recover the color burst. A better solution might involve blocking the color burst signal in the digital part of the circuit when in text mode with a logic gate. I think it might require a extra chip to do this, which is probably why Apple used a transistor.
I haven't thought about this enough to come up with the actual circuit, but it probably could be done with one or two gates using the existing TEXT MODE (F14, pin 4) to conditionally block the COLOR REF signal (B1, pin 3).
-Mike Willegal
What TV are you using and can you post some pictures how it looks? Is it similar to what I have described here: https://www.applefritter.com/comment/99368#comment-99368
I would suggest trying a different TV. Out of the 4 CRT TVs/monitors with composite input that I own, this happens with only one of them (the Sony Trinitron).
I think that Mike's information and idea for a proposed solution is probably the closest to what Uncle Bernie is looking for so far... Does anyone know if something similar is what was used in newer machines like the //e and //c? Probably internal to the IOU if that's how they did it.
The Apple IIe and above that have the IOU chip use a CLRGATE signal generated there to gate IN the color burst signal rather than trying to remove it. This does indeed provide a cleaner solution and works with almost all montors. This circuit could certainly be added to earlier Apples if one was so motivated.
The gate for /COLOR REF was present starting on the Rev 7 and RFI Apple II+. It added a 74LS02 quad NOR at A14 as follows:
RFI color gate.png
Schematic taken from Winston Gayler's essential "Apple II Circuit Description".
In post #7, CVT wrote:
"What TV are you using and can you post some pictures how it looks? "
Uncle Bernie answers:
Sorry, I can't take a picture which shows the problem. It's incredibly difficult to snap such a picture with a still camera. The ill effect is dynamic: the unwanted color moves all the time. Think of a "color glow". No way to capture that with a still camera. I do have a video camera but can't find the power pack.
But I think the first photo in post #6 of the thread:
https://www.applefritter.com/comment/99368#comment-99368
mentioned by "CVT" in post #7 above may show a hint of this "color glow" or "color wobble" effect but since it's a still picture we can't tell fuer sure.
There also is a lot of noise visible in that photo (the specks in the background) which should not be there for direct video input. So maybe a RF modulator was involved which can bring its own set of ill effects. Or some other defect.
About my TV in question, it's a "Symphonic Model SC3913", Made in Malaysia in January 1999. It still works fine and it has the all important "GAME" mode where video and sound enter through front side RCA jacks. It certainly does not use the TV chip set me and my team developed in the mid-to-end 1990s. This was the last chip set ever developed for CRT based TV chassis and despite the CRT was on the way out (think: dying dinosaurs) it still sold very well. And it did have proper detection of "no color" conditions, unlike the problem at hand with this "Symphonic" TV.
A QUICK HACK REMEDY FOUND:
In the meanwhile I have found a remedy, but it's a hack, and does not weed out the root cause of the problem:
Understanding the nature of the ill effect, I just disconnected the "color burst killer" transistor, so I get the same picture "quality" as seen on early Apple II before this half-baked "color burst killer" circuit was added.
So I now do get a color burst in all text lines, and yes, there are "artifact colors" in the characters, but at least now the color subcarrier oscillator in the TV has a stable color burst to lock on, and the artifact colors are stable and don't "glow", "warble", or move. The text is readable now without the "LSD trip" effect, and I can now work for hours on the machine without straining my eyes too much. Despite the artifact colors in the text look ugly and take contrast out of them. Here is a photo to show how text looks now:
PDRM3540.JPG
Color reproduction is not great with this camera, but you can get a hint how all the vertical lines in each text character get a green or magenta tint, depending on whether they are on an even or odd pixel x-coordinate. The camera does not really reproduce how bad it looks - the colors are much more green-ish and magenta-ish on the real TV screen.
This is the very same color artifacting effect used by the Apple II to produce its "six colors" in hires graphics mode. You have to carefully place each pixel on the right spot to get the desired color. A clever trick by Woz thinking out of the box. Which not always is the right thing to do. True, in 1977 it got us six colors in Apple II hires "essentailly for free" (the phase shift which was added later cleverly turns green and magenta artifact colors into orange and blue), but it also produced a lot of toil and trouble down the road. Just think about the PAL card woes. And the extra effort to program nice looking graphics.
As for the text artifacting problem, I think that early Apple II owners must have complained about the unwanted colors in text. So Apple added the "color burst killer" circuit, the one with the sole transistor. Which works fine if the TV does not have high gain for the color burst signal (or any random crap which comes in its place in the TV scan line). But with TVs that have a very sensitive burst amplifier, this "color burst killer" circuit actually makes matters worse: the reason is that the color subcarrier oscillator in these sensitive TVs more often than not never acquires proper phase lock, and so the artifact colors turn into ever moving rainbow colors which are very annoying and a strain on the eyes.
CONCLUSION
This sometimes dysfunctional "color burst killer" circuit is another example where a half-baked "solution" that seems to work for some, causes a lot of headache (literally) for others. I got a headache from these warbly and instable colors. Too much strain on the eyes. I could only work for half an hour or so without pause to avoid that ill effect.
At the moment I explore Apple II color graphics. I never programmed them myself. So even I, at my age, still have to learn ;-) This is why permanent disabling of the color or use of a B&W monitor was not an option.
I hope I can soon accomplish my goal to port Li-Chen Wang's "Kaleidoscope" program he wrote in Intel 8080 assembly code for the Cromemco Dazzler to the Apple II. Of course this is a pastime sort project, just to kill time. But I always wanted to do this. This is the early microcomputer program which, as the story is told, back in 1976 did cause a major traffic jam and drew crowds in NYC. Drivers would leave their cars parked in the road to get out and see this "Kaleidoscope". Hence, the traffic jam. And all this just because one computer shop had the thing placed and running 24/7 in its storefront window. To attract customers, I think. And it did . . . but how many buyers where among the bedazzled spectators lured by it ? Back in 1976 this must have been an effect never seen before by the people. When the microprocessor (and yours truly) was still young !
- Uncle Bernie
I did exactly the same hack when I was using my Apple II+ with a Sony Trinitron TV, but then I got a JVC TV for it that did not have this problem, so I reversed the hack. You don't even need to disconnect the transistor, just lift one of the legs of the resistor connected to its base. You can even hook up an external switch that does that.
One of my early A2's does this (I think Rev.3 or 4).
The transistor killer circuit is not enough and it worked maybe 25% of the time to make the TV switch off color on any of my monitors.
I believe that I was able to piggyback a second resistor into the circuit and improve it to 90% or more. This was a fix I found here on the forums or maybe just reprinted from very old magazine articles.
Found it:
Apple II Plus color killer circuit | Applefritter
I have an excellent book titled Enhancing your Apple II Vol 1, if you still info on this the book contains a whole Chapter on archieving this, I can scan or photograph the 12 pages detailing all that is needed to Install a software Color Killer program, let me know if this is usefiul or not.
It will be very useful.
A full scan of the book from Don Lancaster:
Enhancing your Apple II Vol 1
can be found here : https://archive.org/details/enhancing-your-apple-ii-volume-1