Re: AppleColor Composite Monitors discussion

MdntTrain wrote:
Although it certainly doesn't hurt to have a fine shadow mask pitch,
0.52mm is really not very fine. The CRT was chosen to support just
enough resolution that it's tolerable to view 80-column text on it.

It's pretty good for a TV CRT, but not very good for a color monitor.


Yeah, it's definitely not as good as the IIgs RGB monitor's .31 or .
35? pitch (for 12") for around 650 trio stripes... but on the other
hand RGB arcade monitors of that period were at .66 for 13" and .82
for 19" so it's not too shabby.

What makes this monitor unique among composite monitors, and able
to display readable 80-column text at all, is that it has a switchable
bandwidth luminance channel that goes into high bandwidth mode when in
monochrome mode.


Yes, I notice it does that automatically with the IIe/IIgs, but
earlier I have to press the button.

Or just send it a video signal without a color burst...

Related to this luminance bandwidth topic, we determined awhile back
that in Apple II 40 col text and HGR, the "dot clock" was 7mhz..
meaning that dots are being sent out at that rate. For 80 col text
or DHGR, the dot clock would be 14mhz. Further, you discovered that
for IIgs 320 mode, the rate was 8mhz, and for 640, 16mhz.

Actually, for lo-res, the dot clock is also 14MHz.

Two questions about this. 1) How would someone with an oscilloscope
go about verifying these dot clock rates, say for starters just on a
IIe? Obviously, the scope gets hooked up to the video out and the
timings set to see one horizontal line.. but what would we need to
have on the screen and then look for in the waveform to count this dot
clock?

Well, for anything but the IIgs, the exercise is more about how to use
a scope than about determining the dot frequency, since the frequencies
are absolutely clear from the schematics. The video design of the
Apple II is one of the most thoroughly described aspects of its design,
and also one of its most novel aspects.

The IIgs is another matter, since the schematic does not clarify this
issue much. ;-)

The easiest way to proceed is to display a pattern/color on the screen
that you wish to investigate. If you want to see the frequency of the
dot clock, you will need to find a pattern/color with alternating 0's
and 1's at the highest rate for the display mode--like gray or, for the
IIgs, the dithered blue of the desktop background.

Then display the video on the scope, setting the sweep to display
a horizontal line. Then magnify the sweep and examine the timing
for, say, ten cycles of the video signal (corresponding to 20 dots)
and take the reciprocal of the time and multiply the resulting
frequency by the number of cycles timed.

If your scope has fancier modes, you may be able to read out the
frequency directly, or use "delayed sweep" to get a more stable lock
on the video undulations and more magnification to time it more
accurately.

Oscilloscopes are all about displaying waveforms vs. a calibrated
time base, so the trick is to adjust the vertical amplitude and the
triggering to get a stable display, then count M events in T time
and do the division.

2) How does dot clock relate to needed bandwidth? If a 14mhz dot
clock is used for 80 columns, how does the Applecolor or IIgs RGB
monitor handle that with an 8mhz bandwidth? Is it because of half
cycles, such that 14mhz requires a 7mhz analog bandwidth?

Exactly. The highest video *fundamental* that can be generated by
a 14MHz dot clock is 7MHz, composed of alternating 0's and 1's.
(Of course, having a wider bandwidth will show sharper "edges" on
the dots, but an 8MHz "3dB bandwidth" will certainly make them all
visible.)

The first-order effect of the high frequency response of the monitor
rolling off is that isolated pixels will not show a full excursion to
black or white, but will only get, say, 70% of the way there.

This is particularly apparent in text, since there are many vertical
features only a single pixel wide, and they are noticeably dimmer than
horizontal line segments, which reach full brightness during their
multi-pixel durations.

The 8MHz bandwidth of the AppleColor Composite monitor is good enough
that the difference in brightness of the horizontal and vertical parts
of characters is not too distracting.

Many monochrome monitors have video bandwidths of 12-18MHz, and on them
the horizontal and vertical features of 80-column characters look almost
the same brightness.

Cool stuff. Thank you Michael!
~ J

You're welcome--it's a very interesting monitor!

-michael

NadaPong: Network game demo for Apple II computers!
Home page: http://members.aol.com/MJMahon/

"The wastebasket is our most important design
tool--and it's seriously underused."
.

Loading