1084 woes
Linards Ticmanis
I tried to get the Commodore 1084 Monitor to work with a Apple II
eroplus with pal card... it works fine in pure text or pure graphics
mode... however in mixed text/graphics mode the colors are messed up. I
wonder, does the 1084 maybe not support graphics lines and b/w lines in
the same frame?
The Apple does not produce a color burst in text mode, only in graphics
mode. In mixed mode the graphics lines do have a color burst while the
text lines don't.
Thanks,
--
Linards Ticmanis
The Master said, "The business of laying on the colors follows the
preparation of the plain ground."
Mike
> The Apple does not produce a color burst in text mode, only in graphics
> mode. In mixed mode the graphics lines do have a color burst while the
> text lines don't.
My Apple IIc and II+ NTSC machines both produce color-burst in mixed mode
on the text. The characters have alternating purple-ish and green-ish
colors.
I think you are mistaking about mixed-mode.
Mike
David Empson
> I tried to get the Commodore 1084 Monitor to work with a Apple II
> eroplus with pal card... it works fine in pure text or pure graphics
> mode... however in mixed text/graphics mode the colors are messed up. I
> wonder, does the 1084 maybe not support graphics lines and b/w lines in
> the same frame?
This is normal behaviour with any colour composite monitor on an Apple
II.
> The Apple does not produce a color burst in text mode, only in graphics
> mode. In mixed mode the graphics lines do have a color burst while the
> text lines don't.
In mixed mode, the Apple II _does_ generate colour burst for the text
lines.
There is at least one exception: on an Apple IIgs with an RGB monitor,
the VGC chip is able to switch to monochrome mode for the four lines of
text. (It is actually fixed colour mode - on the IIgs, the user can set
the foreground and background colours in text mode, and the border
colour for all modes.)
With a composite monitor and the text colours set to white on black, the
IIgs behaves exactly the same as earlier models - you get green and
violet for the text characters.
--
David Empson
dem...@actrix.gen.nz
Linards Ticmanis
> In mixed mode, the Apple II _does_ generate colour burst for the text
> lines.
OK... probably true. However it seems that the PAL card that I use doesn't.
It took me a while to find out that this card (made for a clone) expects
an active-high TEXT instead of the Apple-provided active-low /TEXT on
pin 23 of slot 7. So in an unmodified state I'd get colored text but B/W
graphics. Exchanging one NPN transistor (2N3904), that shuts down the
PAL quartz when active, with a PNP one (2N3906) that happened to be
lying around fixed the problem - only in mixed mode I'd get the
complementary colors (purple instead of green etc.) in a stripe covering
about the second sixth of the screen, counting from the top, preceded
and followed by a few lines in B/W. Also the text wasn't fringy, as it
should be, derived from what you said. Only the top most line of text
had a bit of a pinkish color to it. Totally removing the transistor gave
color fringed text at all times, but with correct colors in the graphics.
It seems that the /TEXT signal on pin 23 of slot 7 _is_ indeed pulled
low for the four lines of text in mixed mode. Unfortunately I don't have
an Oscilloscope, but a quick multimeter check showed a somewhat
decreased mean voltage on the pin in mixed mode, compared to
graphics-only mode.
I guess either the quartz or the PAL chip TCA650 simply doesn't like to
be switched on and off very quickly. I'll try inserting a 220uF
capacitor between the transistor base and ground - combined with the
base current limiting 4.7kOhm resistor that should delay the activation
of the transistor for about 0.1s (if I calculated correctly), quite
enough to get the quartz across those 4 lines of text alive.
Michael J. Mahon
>I guess either the quartz or the PAL chip TCA650 simply doesn't like to
>be switched on and off very quickly. I'll try inserting a 220uF
>capacitor between the transistor base and ground - combined with the
>base current limiting 4.7kOhm resistor that should delay the activation
>of the transistor for about 0.1s (if I calculated correctly), quite
>enough to get the quartz across those 4 lines of text alive.
No crystal oscillator can be switched on quickly. That is the nature
of a high Q circuit--it takes a while to settle down.
Instead, the output of the constntly-running oscillator can be gated
(ANDed with a control signal) very quickly, and this is the usual
means of switching a clock signal on or off.
Even if you can stabilize the color reference, it must b e supplied
on each line--including during vertical blanking--so that the color
monitor can maintain correct phase lock. This is another "high Q"
circuit which must be kept supplied with a stable phase reference
or color rendition will suffer--usually after just a few lines without
a burst.
This is why the Apple does not switch the color burst off during
the text part of a mixed display--to do so would cause unspecified
behavior of the the "color killer" circuit in the monitor, and poor
rendition of the color for many lines after the burst was restored.
-michael
Email: mjm...@aol.com
Home page: http://members.aol.com/MJMahon/
Linards Ticmanis
>>I guess either the quartz or the PAL chip TCA650 simply doesn't like to
>>be switched on and off very quickly. I'll try inserting a 220uF
>>capacitor between the transistor base and ground - combined with the
>>base current limiting 4.7kOhm resistor that should delay the activation
>>of the transistor for about 0.1s (if I calculated correctly), quite
>>enough to get the quartz across those 4 lines of text alive.
>
>
> No crystal oscillator can be switched on quickly. That is the nature
> of a high Q circuit--it takes a while to settle down.
OK, then my guess was basically right.
> Instead, the output of the constntly-running oscillator can be gated
> (ANDed with a control signal) very quickly, and this is the usual
> means of switching a clock signal on or off.
This is obviously not the solution in this case, because...
> Even if you can stabilize the color reference, it must b e supplied
> on each line--including during vertical blanking--so that the color
> monitor can maintain correct phase lock. This is another "high Q"
> circuit which must be kept supplied with a stable phase reference
> or color rendition will suffer--usually after just a few lines without
> a burst.
And it did suffer. Quite badly in fact.
> This is why the Apple does not switch the color burst off during
> the text part of a mixed display--to do so would cause unspecified
> behavior of the the "color killer" circuit in the monitor, and poor
> rendition of the color for many lines after the burst was restored.
Basically the circuit looks like this....
Quartz
etc. Ground
E C
\ /
\ /
===== PNP transistor 2N3906
B |
---
| | 4.7kOhm resistor
| |
---
|
|
/TEXT from slot.
I wonder what would be the easiest solution to delay the activation of
the transistor? I thought about adding a someting like a capacitor
(220uF maybe) in between B and Ground.
The capacitor would be charged through the 4.7k in Graphics mode
and then upon the switch into text mode it would slowly discharge
through the resistor again. So both switching directions would have a delay
introduced. The mean voltage in mixed mode should be high enough to keep
the transistor inactive at all times then, so the quartz gets its
working voltage and stays happy.
Unfortunately my knowledge of electronics is limited, so I have no idea
if this would cause my Apple to burn...
Linards Ticmanis
> I wonder what would be the easiest solution to delay the activation of
> the transistor? I thought about adding a someting like a capacitor
> (220uF maybe) in between B and Ground.
>
> The capacitor would be charged through the 4.7k in Graphics mode
> and then upon the switch into text mode it would slowly discharge
> through the resistor again. So both switching directions would have a delay
> introduced. The mean voltage in mixed mode should be high enough to keep
> the transistor inactive at all times then, so the quartz gets its
> working voltage and stays happy.
To answer my own questions (maybe for the benefit of future Google
users?), it did work fine. Color is delayed about a second after
switching to Graphics mode, but that is fine with me as long as things work.
Also I moved the quartz's fine-tuning adjustable capacitor (5-30 pF or
so) to the outside of the case and connected it wiht some wires - that
makes it possible to adjust the fine-tuning when the picture starts to
wobble from the interference of the mainboard color clock and the PAL
card's. Since getting rid of this depends on very exact tuning of the
quartz, it has to be readjusted once in a while.