New video upgrade coming soon!

[TechCowboy]

I'm willing to wait for the whole audio/video full meal deal.

[+remowilliams]

Just added the UAV board my 800XL and the video quality is... amazing. Thanks Bryan!

[+DrVenkman]

Hey Bryan - I initially installed my first UAV board in an 800XL and it works great. However, after now installing my U1MB board in the same machine, I'm getting the dreaded vertical bands on the screen, especially on the left side. If I understand things correctly, I probably need to pull power directly from a pin near the power supply by soldering a wire directly to the UAV board.

 

Here's where I'm unclear: does it matter where exactly on the +5V pad on the UAV board I solder the power jumper? Or do I need to solder power to one side of the pad and cut the narrow trace between the sides?

[Frankie]

Is there a waiting list for this?

[+Stephen]

Is there a waiting list for this?

No - once produced, the next "batch" will be in the Atari Age store.

[Bikerbob]

Succes!

 

Before and after pictures.

 

Atari 600 XL PAL.

 

It didn't remove the noise pattern, which moves from the bottom of the screen to the top. I believe that is due to a hardware fault / bad design in the 600 XL.

 

http://imgur.com/a/A8DW2

Im not saying that that the sharpness looks better on the UAV..

 

but you lost a lot of your colour brightness?? Why such a change in the colour saturation??

 

James

Edited March 27, 2016 by Bikerbob

[dakidski]

Im not saying that that the sharpness looks better on the UAV..

 

but you lost a lot of your colour brightness?? Why such a change in the colour saturation??

 

James

Those photos are obsolete. I worked further on the mod and used s-video instead. Now the picture quality is perfect.

 

UAV does seem to have lower saturation, but I haven't tried adjusting it. I kinda like it that way.

 

Dakidski

[cjameslv]

Yeah svideo should be main output used anyway. No point in not using UAV to its fullest.

[+DrVenkman]

Yeah svideo should be main output used anyway. No point in not using UAV to its fullest.

 

 

If I had an S-Video CRT TV or monitor with chroma/luma I'd be all over it. In the meantime, I'm pretty darn happy with the composite output on my 800XL. I re-did the ground connection this afternoon and got rid of most of the annoying vertical lines. I might cut the UAV ground pad as Bryan has suggested but we'll see.

[cjameslv]

 

 

If I had an S-Video CRT TV or monitor with chroma/luma I'd be all over it. In the meantime, I'm pretty darn happy with the composite output on my 800XL. I re-did the ground connection this afternoon and got rid of most of the annoying vertical lines. I might cut the UAV ground pad as Bryan has suggested but we'll see.

Look around on craigslist. You can usually score a nice sony trinitron crt for $15 or less, sometimes free! For a nice monitor, i use the commodore 1702. It has seperate luma/chroma as rca jacks but you can solder an svideo connector and it works great. Edited March 27, 2016 by cjameslv

[+DrVenkman]

Look aroubd on craigslist. You cab usually score a nice sony trinitron crt for $15 or less, sometimes free! For a nice monitor, i use the commodore 1702. It has seperate luma/chroma as rca jacks but you can solder an svideo connector and it works great.

 

 

I've been on the lookout for a 1701/1702 for over a year but never seen one locally. The closest has been over 90 minutes away one-way, local pickup only of course. I haven't looked for CRT TV's lately but perhaps I should.

[cjameslv]

Nooooo lol, you replied to my typo ridden post before i edited it!

[morelenmir]

I'm going to first admit that I have not read through every post on this 20 page thread. I got to about page 5 or 6 and then gave up!

 

Can anyone tell me the pin-out of the 'UAV' output terminals? I know one of them has to be patched to an external chip - the GTIA I think?. Also, is there a crib for the jumper-strip somewhere?

 

Again, my apologies if these questions have been directly answered before, just a link to the post in question would be a big help. I really think it would be useful to have a dedicated order thread and then a second for technical support and discussion as with the 'UltimateCart'.

[+DrVenkman]

Information is rather scattered throughout the thread, and installation info varies from machine to machine - I think that's what Bryan's working on before Albert puts these into the AA Store (I know I'm buying at least one more, maybe two).

 

That said, I made notes. Post #281 lists jumper configurations for various machines and has a diagram of the signals on the board.

 

http://atariage.com/forums/topic/246613-new-video-upgrade-coming-soon/?p=3431922

Edited March 28, 2016 by DrVenkman

[Bryan]

Those photos are obsolete. I worked further on the mod and used s-video instead. Now the picture quality is perfect.

 

UAV does seem to have lower saturation, but I haven't tried adjusting it. I kinda like it that way.

 

Dakidski

Saturation is an interesting beast on Atari computers. Here's how it works:

 

The monitor/TV looks at the colorburst to determine the color phase for that line. Modern monitors also set an overall color amplitude according to it as well. For any desired saturation, the amplitude of the color carrier is supposed to scale with the brightness/luma level. This means bright colors need color carriers which have more amplitude than the colorburst level and dark colors need less.

 

The Atari can only produce one color carrier amplitude which means saturation becomes an inverse function of the luma value. So, dark colors will be very saturated and bright ones will be pale.

 

Now, the A8's fixed chroma level varies from machine to machine. A weak chroma signal means the monitor will try to amplify chroma levels which leads to more composite errors like artifacting on text. A carrier which is too strong will increase the chance that a dot pattern is visible in the picture but there will be less artifacting. I played with several levels and chose one just slightly under the NTSC and PAL reference level which optimized the composite text quality.

 

The other part of the equation is the luma strength. If the original output is darker (which some A8's are) then the colors will be less saturated on UAV because of the luma vs. chroma level rule above.** The luma level was chosen by making GTIA color 15 (available in mode 9 only) close to 1V (reference white) in the video signal. Some video upgrades boost the luma beyond 1V which I tried to avoid.

 

Anyway, every adjustment affects something else so I started with the standard levels and tweaked from there.

 

** The 1200XL/1400XL/1450XLD has an extra circuit to produce 2 chroma levels so the burst will be weaker than the rest of the line. This results in a very vivid picture but a lot of oversaturation and bleeding on darker colors and it was dropped on later machines.

[morelenmir]

Information is rather scattered throughout the thread, and installation info varies from machine to machine - I think that's what Bryan's working on before Albert puts these into the AA Store (I know I'm buying at least one more, maybe two).

 

That said, I made notes. Post #281 lists jumper configurations for various machines and has a diagram of the signals on the board.

 

http://atariage.com/forums/topic/246613-new-video-upgrade-coming-soon/?p=3431922

 

That did it Dr.Venkman, many thanks! Going from these data it is the 'COLOR' signal which needs to be patched in to UAV Pin6. Do you know if this is located on Leg9 of the CTIA/GTIA chips on machines other than the 2600? My next impossible question - do you know why there are two GND pins on the UAV output?

 

Saturation is an interesting beast on Atari computers. Here's how it works:

 

The monitor/TV looks at the colorburst to determine the color phase for that line. Modern monitors also set an overall color amplitude according to it as well. For any desired saturation, the amplitude of the color carrier is supposed to scale with the brightness/luma level. This means bright colors need color carriers which have more amplitude than the colorburst level and dark colors need less.

 

The Atari can only produce one color carrier amplitude which means saturation becomes an inverse function of the luma value. So, dark colors will be very saturated and bright ones will be pale.

 

Now, the A8's fixed chroma level varies from machine to machine. A weak chroma signal means the monitor will try to amplify chroma levels which leads to more composite errors like artifacting on text. A carrier which is too strong will increase the chance that a dot pattern is visible in the picture but there will be less artifacting. I played with several levels and chose one just slightly under the NTSC and PAL reference level which optimized the composite text quality.

 

The other part of the equation is the luma strength. If the original output is darker (which some A8's are) then the colors will be less saturated on UAV because of the luma vs. chroma level rule above.** The luma level was chosen by making GTIA color 15 close to 1V (reference white) in the video signal. Some video upgrades boost the luma beyond 1V which I tried to avoid.

 

Anyway, every adjustment affects something else so I started with the standard levels and tweaked from there.

 

** The 1200XL/1400XL/1450XLD has an extra circuit to produce 2 chroma levels so the burst will be weaker than the rest of the line. This results in a very vivid picture but a lot of oversaturation and bleeding on darker colors and it was dropped on later machines.

 

That is really interesting information Bryan! My own 130XE displays both the high-chroma dot pattern and the low-chroma artefacted text through its 'Monitor' composite video. What I find particularly strange is that it shifts between these conditions randomly, which leaves me wondering if there is a dodgy component somewhere. In the past I have concentrated on replacing capacitors as they seem to be the most likely to degrade with time. I wonder if the interference which comes in to the video circuitry along the power tracks from the RAM refresh can also affect the chroma level?

 

On a stock motherboard is the chroma level controlled by the R38 (on an 130XE) potentiometer?

[+DrVenkman]

 

That did it Dr.Venkman, many thanks! Going from these data it is the 'COLOR' signal which needs to be patched in to UAV Pin6. Do you know if this is located on Leg9 of the CTIA/GTIA chips on machines other than the 2600?

 

Actually, per one of Bryan's posts elsewhere in this thread, for my 800XL picked up the COLOR signal from the leg of a resistor just beneath the 4050 IC chip (R45 I think?) My photo doesn't show the markings very clearly, but I've circled the one Bryan indicated for an 800XL.

 

Edited March 28, 2016 by DrVenkman

[morelenmir]

 

Actually, per one of Bryan's posts elsewhere in this thread, for my 800XL picked up the COLOR signal from the leg of a resistor just beneath the 4050 IC chip (R45 I think? My photo doesn't show the markings very clearly, but I've circled the one Bryan indicated for an 800XL.

 

IMG_5760.jpg

 

Interesting!!! Many thanks again DrVenkman!

[+DrVenkman]

 

Interesting!!! Many thanks again DrVenkman!

 

 

You're welcome.

 

If you're installing in a 130XE, Bryan posted instructions with a photo a couple pages back:

 

http://atariage.com/forums/topic/246613-new-video-upgrade-coming-soon/?p=3467649

[morelenmir]

 

 

You're welcome.

 

If you're installing in a 130XE, Bryan posted instructions with a photo a couple pages back:

 

http://atariage.com/forums/topic/246613-new-video-upgrade-coming-soon/?p=3467649

 

That lays it out very clearly indeed - and simultaneously answers my question about where to patch the 'COLOR' signal from!

 

Do you know if the various Chroma, Luma and Composite signals need to go through those inductors, or can they just be connected directly to the appropriate pins on the base of the socket? It seems a bit of a shame to risk potential interference with what are no doubt pristine outputs from the UAV!

 

Another big positive about the UAV is, once installed and providing the composite video signal I can finally remove the big and ugly RF can! This gives a nice amount of extra room for any future mods such as the Ultimate1MB in particular.

[+DrVenkman]

That lays it out very clearly indeed - and simultaneously answers my question about where to patch the 'COLOR' signal from!

 

Do you know if the various Chroma, Luma and Composite signals need to go through those inductors, or can they just be connected directly to the appropriate pins on the base of the socket? It seems a bit of a shame to risk potential interference with what are no doubt pristine outputs from the UAV!

 

Another big positive about the UAV is, once installed and providing the composite video signal I can finally remove the big and ugly RF can! This gives a nice amount of extra room for any future mods such as the Ultimate1MB in particular.

The instructions indicate lifting the legs of those ferrite bead inductors and soldering the wires into the pads vacated by the lifted legs. This cuts the signal from the RF modulator and feeds the UAV signals into the board to feed the DIN jack.

 

In my case, the 4050 was socketed so installing the UAV there in place of the existing IC cuts the RF modulator out of the loop anyway, though I haven't tried to remove it. ASIDE: I have no idea how to get one of those off an A8 board. Seems like it might require a Dremel, pliers and a lot of desoldering.

[morelenmir]

The instructions indicate lifting the legs of those ferrite bead inductors and soldering the wires into the pads vacated by the lifted legs. This cuts the signal from the RF modulator and feeds the UAV signals into the board to feed the DIN jack.

 

Of course - I was reading it backwards! I thought when Bryan said 'Once the lead is out of the board, solder a wire to it.' that he meant solder a wire to the inductor leg that you have just lifted clear! For the sake of neatness I think I will just pull the whole component in each case.

 

 

In my case, the 4050 was socketed so installing the UAV there in place of the existing IC cuts the RF modulator out of the loop anyway, though I haven't tried to remove it. ASIDE: I have no idea how to get one of those off an A8 board. Seems like it might require a Dremel, pliers and a lot of desoldering.

 

Have you never removed one then DrVenkman? I installed two VBXE's and both times needed the space the RF can occupied to put in a DIN13 socket for the RGB out. On the 800XL it is held in place by three connections on the bottom edge nearest the keyboard - which is where the signals come in to be RF'ified - and then four thick L-shaped legs that go right through the motherboard and are soldered down to the ground plane on the other side. As you can imagine it is a real bastard of a job and requires exactly as you guessed - a Dremmel and a diamond cutting wheel. My technique began by levering down the segment of perforated skirt directly above each of the four big legs. This not only gave access for cutting but some degree of protection to the motherboard if (when) the cutter slipped... Even then you had to be extremely careful and very sure you vacuumed up all the metal dust afterwards. If you try to use the standard abrasive cutting disks you end up with an even bigger mess. Once the solid chunk of metal had gone it was comparatively easy then to desolder the stump of each leg and clean up the slots they occupied. Maybe if you have a really powerful soldering iron you can do it non-destructively, but I found the thermal mass of both the can itself and the ground plane just sucked all the heat away from even a 60W iron before it would melt the solder fixings.

 

Altogether a horrible task, although I must admit It is much nicer on the 130XE. I discovered that at 450°C you could just about solder-suck/braid away enough to get the can free whole and undamaged. Which is just as well because after removing mine I discovered the XE required it to produce a composite video signal!!! Much wailing and gnashing of teeth followed!!!

[Bryan]

On a stock motherboard is the chroma level controlled by the R38 (on an 130XE) potentiometer?

 

The pot controls the color phases, so it's like a HUE knob. Basically, you have 15 hue choices and that controls how many of the delay elements the color carrier goes through before exiting GTIA. The knob is a voltage input to the elements that controls delay by changing their switching speed.

 

Do you know if the various Chroma, Luma and Composite signals need to go through those inductors, or can they just be connected directly to the appropriate pins on the base of the socket? It seems a bit of a shame to risk potential interference with what are no doubt pristine outputs from the UAV!

 

It doesn't really matter if you use the beads or not since they have a very low inductance. Their purpose is to reduce the amount of RF inside the shield that escapes via the connected cables. If you lift the end connected to the monitor jack, you just solder to the pad on the board. If you lift the other end, solder to the lead you lifted. I don't think you'd notice a difference either way.

[morelenmir]

 

The pot controls the color phases, so it's like a HUE knob. Basically, you have 15 hue choices and that controls how many of the delay elements the color carrier goes through before exiting GTIA. The knob is a voltage input to the elements that controls delay by changing their switching speed.

 

I think I see what you mean... I really must look in to the technical nature of how video works! Certainly on a bit of experimentation earlier I saw that adjusting it did make a big difference to the colours shown. However at the time I also thought it cut back on the interference I was seeing as well. Going from what you say above then maybe it just changed the colour of the background so that it was harder to see the interference. When the screen was in attract mode and other colours than light-blue were set as background the interference sprang back to visibility.

 

It doesn't really matter if you use the beads or not since they have a very low inductance. Their purpose is to reduce the amount of RF inside the shield that escapes via the connected cables. If you lift the end connected to the monitor jack, you just solder to the pad on the board. If you lift the other end, solder to the lead you lifted. I don't think you'd notice a difference either way.

 

Excellent Bryan!!! Many thanks for clearing that up for me.

 

'Cutting the ground pad' was mentioned several posts up by DrVenkman. I understand the UAV board has pads for attaching dedicated +5V and GND power inputs in case the native wiring is too noisy for even the new LP5907MFX-4.5 voltage regulator to deal with. However I do not understand what cutting these pads would achieve, unless you were physically cutting the native input from the 4050 socket in order to use the clean power directly from the switch?

[Bryan]

I think I see what you mean... I really must look in to the technical nature of how video works! Certainly on a bit of experimentation earlier I saw that adjusting it did make a big difference to the colours shown. However at the time I also thought it cut back on the interference I was seeing as well. Going from what you say above then maybe it just changed the colour of the background so that it was harder to see the interference. When the screen was in attract mode and other colours than light-blue were set as background the interference sprang back to visibility.

 

The easiest way to think about it is that the pot changes the angular distance between the colors on the wheel:

 

 

When the pot is set correctly, the 15 color options are distributed evenly around the wheel.