Well, i got the MC6847 working, but only in b/w. I bought two 2N2222 transistors and i've built an video amp, but it works only with b/w signals, when i tried to run color signal from my iptv receiver to the tv, the picture was heavily distorted. How could i build a reliable video amp using 2N2222 and bfr90 transistors?

I've added the video memory, and tested it with the EhBasic(stolen from the sbc) eeprom, and it works(i am seeing random characters of course)!
But still i didn't find a solution for the color amplifier, and color in general...

So what have you done so far to address color in your design ? A schematic would assist us in helping you further. In any case you need a color modulator if you intend to have a composite video signal. MC1372 is such a device, MC13077 is another candidate but a bit more difficult to find and to find the external components for.

PauF also had some interesting ideas (See some posts earlier in this thread) on how to decode Y, phiA and phiB to generate an RGB signal for display on a TV device. This will almost certainly give you the best quality picture.

I've tested my colour circuit and got all the bugs out. It decodes the Y, phiA and phiB signals and passes them through a programmable palette to generate 4-bit colour (R, G1, G2, B) that can then be fed to a SCART-equipped TV or analog monitor.

The picture quality is excellent and the palette device allows the colours to be redefined - thus you can have white on blue text and pixel-mapped graphics in any four colours you like.

The circuit uses 10 ICs and 6 transistors, so it's not for those who want a simple design but the picture quality is so far above the results obtained using a modulator that I've turned my back on modulators entirely!

I'm writing up the design at the moment and hope to post it to this site sometime this week. The document I'm preparing also includes circuitry for other enhancements, such as text underlining, flashing text and a hardware cursor.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

Paul,
Sounds really interesting. Would the digital part of it be suitable for a CPLD design do you think ? If so it would be much more manageable on a small board such as the one I'm doing for the uKim.

/P

There's no reason why you couldn't use a CPLD to implement the logic functions. This would replace 7 of the ICs. With the CPLD, you would need 3 LM309 dual comparitors and 6 transistors to provide the output drive for 75 ohm loads.

_________________
Shift to the left,
Shift to the right,
Mask in, Mask Out,
BYTE! BYTE! BYTE!

Well, i have the modulator already, and i am trying different schematics for it, but i get always black screen(when i bypass it it works but then i got b/w)... And my 2N2222 video amp does not like colors either... So i need new schematics... And i don't have much space left on the board, so it has to be quite simple...

Schematics we found so far:

http://www.brielcomputers.com/phpBB3/viewtopic.php?f=13&t=937
http://homebrewcpu.blogspot.com/2009/02/this-is-mc6847-character-generator.html(this one works perfectly in b/w)

Well the first schematic is mine and the MC1372 implementation on it works, although alphanumeric and numeric characters does not look very good. But you should be able to use that schema to get your design working. Leave the pot's in the center position and the trim from there to get the best picture. R9 will trim the color saturation and R5 will adjust the offset of the resulting composite signal.

/Pontus

Well, the fs signals the end of the last scanning line, but how would i write between the end and start of the next line, since the picture is not using the entire screen?

Well, i think i got it right, the image shows the logic for controlling the crtc interface to the cpu. I would appreciate if someone could check it for bugs.
The MS signal also controls the tristate circuit that connects the data bus form the cpu to the crtc ram.
S is select, for fast mode where the cpu does not wait for the crtc to finish displaying.

Sorry for the messy schematic, i patched it several times...

4069 typ prop delay at 5V is 50ns, max 90ns.
4071 typ prop delay at 5V is 100ns, max 250ns.
4011 typ prop delay at 5V is 120ns, max 250ns.
From phase 2 to WE\, your propagation delay could be as much as 590ns. That's more than the entire phase-2-high time at 1MHz. See why 4000-series logic is not suitable for computer glue logic? At least use 74HCxx.

Well, i can easily use 74 logic, i used them here only for testing the circuits(i found them first in the menu). Apart from that i hope it works.
74xxx logic is getting a bit hard to find at local stores...

Even here in southern California where we have many electronics stores in the area, I still have to order most of my parts. The walk-in stores only have very basic things. It used to be a little better 25 years ago when there were more hobbyists (before SMT drove them away) and schools were teaching more component-level projects, but even then I had to order most of my parts. The cost of shipping and sometimes a minimum order amount often made it worth buying more parts than I needed, and/or buying other ones that I did not need yet but thought I might need in the near future.

Can i mix 74ls, 74hc, 4000, 6502, and other sbc ic. Do the voltages of LO and HI match? I know there are some ic that don't match...

Your question is primarily Manufacture specific, but I came across this in my files, back when I was researching a similar question as yours...

It may help you, good luck Dajgoro.