Atom Colour Board Thread

[Prime]

Aha!

Found the problem, I was feeding the Y signal into the pad that the emitter for TR2 would be connected to, however I had failed to notice / disconnect R6 a 470R also connected to that pin. So I guess that was pulling Y down, and stopping the other circuitry work

Mind I've also now got a 74LVC1G66 in there, they are *SMALL* and really fiddly to solder, might try the slightly bigger version, or replace it with a 74LVC4066, which has the same on resistance, and though it has 3 extra switches I don't need it would be much easier to mount on the board.....

Now to prototype the changes.

Cheers.

Phill.

[oss003]

Phill, it's new years eve ...........
Happy newyear ..,...

Greetings
Kees

[hoglet]

Found the problem, I was feeding the Y signal into the pad that the emitter for TR2 would be connected to, however I had failed to notice / disconnect R6 a 470R also connected to that pin. So I guess that was pulling Y down, and stopping the other circuitry work

Yes, that would definitely mess things up. Especially with the higher on-resistance of the 4016. You might actually find that with R6 removed the 4016 does in fact work.

Dave

[Prime]

So I was initially just working on a section of breadboard, but decided that wasn't a very long term solution.......

So Now I have this on a little pice of stripboard plugged into the MK5.1 where the Colour VR would normally be, having first removed the Colour & Mono pots & resistors.

Things to note, the Y amp has been removed as Y is now coming in from PL4 pin 9.
As mentioned above I no longer need separate asjustments for Colour / Mono graphics modes, so those components are removed.
Added is Dave (hoglet) 's back portch clamp circuit, the CPLD has it's code modified to generate this.

This is just a tempory solution to do some more tests, I have an etched board with the above changes ready to drill and build, just wanted to check the changes first, this will need one less comparitor chip too, but can't do that on the MK5.1 board due to the way I used the comparitors within the chips.

Modified MK5.1 colour board.

Cheers.

Phill.

[hoglet]

Hi Phill,

So Now I have this on a little pice of stripboard plugged into the MK5.1 where the Colour VR would normally be, having first removed the Colour & Mono pots & resistors.

I love the dead-bug mounting of the 74LVC1G66 - it really is rather a pig to solder.

Are you finding this is an improvement? I hope so, as you've put a lot of effort into this.

There is one more trick that might help if you are still having noise issues...

In the CPLD you can legitimately ignore the luminance (Y) signal for certain colours.

Specifically, you only actually need to consider luminance for:
- green
- orange
- buff

Why do this?

The 6847 puts out 4 different levels of luminance (ignoring sync):
- WB (Black, Green Text Background, Orange Text Background)
- WL (Red, Blue)
- WM (Green, Cyan, Magenta, Normal Orange)
- WH (Yellow, Buff, Bright Orange)

With your current CPLD all colours are qualified by luminance, so you need to set the threshold to discriminate between WB and WL (which on the 6847 is 0.72V and 0.65V).

If you limit where you use Y (to just green, orange, buff) then you only need to discriminate between WB and WM (which on the 6847 is 0.72V and 0.54V).

Doing this gives you much more immunity to noise.

So if you are finding it hard to eliminate the effects noise, this might be worth a try.

Dave

[Multiwizard]

Hi,

I'm so following this Topic with lot's of excitement...

I could use a few boards...


Greetings, Wim...

[Prime]

I love the dead-bug mounting of the 74LVC1G66 - it really is rather a pig to solder.

Not really a dead bug mounting, I made some little adapter boards, though I did use the wrong footprint for the damn chip....gonna order some more of the bigger footprint ones to try, as Mouser carry them.

Are you finding this is an improvement? I hope so, as you've put a lot of effort into this.

The adjustment is certainly easier, and it only needs one comparitor / pot / resistors so is less components, also not needing to build the Y amp helps as that's 2 transistors, 3 resistors an inductor and a cap less.

There is one more trick that might help if you are still having noise issues...

Right I'll look at that, may certainly help.

Cheers.

Phill.

[TheCorfiot]

Phill,
This is looking awesome, can't wait to upgrade my card

Thanks for persevering

[Prime]

Right, next stage is complete, prototyping new board, on self etched board

AtomColour 5.3.

Seems to be working fine now, though it's taken a few days for me to track down a bug that was caused by A0 being connected properly to the 6847 but not (reliably) to the Atom Bus leading to odd results where writes would get corrupted & therefore wrong data displayed. Tat'll teach me to pay attention when working out which pins need top soldering!

Next stage is to send Mk 5.4 off to ALL-PCB to get some test boards made. Hopefully the new layout should be less of a pain in the ass to build, as I've done things like laying all the resistors flat and re-arranging the board somewhat in places (as can be seen from the picture). Now hopefully the replacement 74LVC1G66s from Mouser will be the bigger foorprint and also not such a problem to solder

Hopefully I'll have at least one with me at the meetup later in the month.

Cheers.

Phill.

[hoglet]

Nice work Phill. I'm looking forward to seeing this next week.

[TheCorfiot]

Congratulations Phill,
I knew you would get there

Looking forward to testing a board in my Atoms

[Prime]

I'll hopefully have the boards by the meetup, I ordered 5 today from ALL-PCB. With any luck I should have at least one built

I've also recieved some of the slightly bigger 74HCT1G66s from Mouser these :

https://www.mouser.co.uk/ProductDetail/ ... VC1G66DBVR

Whilst they are still small they at least match the pads on the board! I managed to desolder the smaller one on my test board and replace it with this.

Cheers.

Phill.

[hoglet]

The ones I am using are larger SOT-23 parts (74LVC1G66DBVT).

[Prime]

Gotta say I'm impressed with all-pcb, ordered these Monday 7th Jan, arrived Friday 11th Jan, including the journey from China!

Anyway here's the Mk 5.4 board seems to be working so far (this is the second one built), which Kees has kindly aggreed to test for me.

Mk 5.4 colour board.

I'm not currently using the picel latching code but the results are still pretty damn good.

Cheers.

Phill.

[jms2]

Phill,

I know you were never able to reproduce the screen rolling instability problem with version 5.1, but is there reason to believe that the new version 5.4 might perform better in that regard?

John

[Prime]

Phill,

I know you were never able to reproduce the screen rolling instability problem with version 5.1, but is there reason to believe that the new version 5.4 might perform better in that regard?

Well, I was able to reproduce the problem with "Cowboy shoot out", this board doesn't seem to have it. The difference.....there will be two resistors soldered to the back of the Mk5.1 board, they will be either 22R or 33R (I think). On the 5.3 I replaced them with 470R, and on the 5.4 with 270R, both of those values seem to help cure the cowboy shootout problem at least for me. So that may be worth trying.

Cheers.

Phill.

[jms2]

OK, I'll try that. I've been trying to follow the details of the v5.4 board to understand how it improves on previous versions, but I must admit the details eluded me! In a nutshell, is the main benefit improved ease of manufacture, or does it do other things such as give more faithful colours? I'm happy with my existing board but I'm interested to understand why you've changed the design.

[Prime]

OK, I'll try that. I've been trying to follow the details of the v5.4 board to understand how it improves on previous versions, but I must admit the details eluded me! In a nutshell, is the main benefit improved ease of manufacture, or does it do other things such as give more faithful colours? I'm happy with my existing board but I'm interested to understand why you've changed the design.

The new board is much easier to build, as it moves some of the patches (like the resistors for /hs and /vs) onto the actual board, also the protection resistors that I was having to insert (manually) into the base circuit of the output transistors.

Because I'm using Dave's back porch circuit, you only need one adjustment that works for all modes, and it stays adjusted. This also means a reduction in required components, 2 less pots, one less LM319 comparitor, and no need for the zenner diode & resistor to set the comp voltage.

The input video is taken from PL4, so there's no need for the onboard preamp, 2 less transistors, 3 less resistors, and no need for the inductor / capacitor circuit.

Cheers.

Phill.

[jms2]

Nice... I can see why you're so pleased with the new design.

[1024MAK]

Gotta say I'm impressed with all-pcb, ordered these Monday 7th Jan, arrived Friday 11th Jan, including the journey from China!

For some reason, I find that December 27th through to mid January are always good times for items coming from China to the U.K., with a lot of items making the journey in 7 to 14 days.

But your delivery date is even more remarkable!

Mark

[Multiwizard]

Anyway here's the Mk 5.4 board seems to be working so far (this is the second one built), which Kees has kindly aggreed to test for me.

I'm not currently using the picel latching code but the results are still pretty damn good.

Cheers.

Phill.

Awesome nice...

Hi Phill,

was the Mk 5.1 board the only version with this Issue at the "Cowboy shoot out" game?


Greetings, Wim...

[sirmorris]

Beautiful work as always, Phill!

[Prime]

was the Mk 5.1 board the only version with this Issue at the "Cowboy shoot out" game?

Hi Wim,

It was the only version it was reported on, I think you have one of the late 4.x series (which became the 5.x), so yours may have the problem
I think your board is the one that has an LS chip on the outputs instead of the 4 transistors, but doesn't do orange correctly (which is why I swapped to the 4 transistors).

Cheers.

Phill.

[bprosman]

Is the full design somewhere published Phill ?

Kind regards, Bram

[Multiwizard]

was the Mk 5.1 board the only version with this Issue at the "Cowboy shoot out" game?

Hi Wim,

It was the only version it was reported on, I think you have one of the late 4.x series (which became the 5.x), so yours may have the problem
I think your board is the one that has an LS chip on the outputs instead of the 4 transistors, but doesn't do orange correctly (which is why I swapped to the 4 transistors).

Cheers.

Phill.

Hi Phill,

I have a MK 4.7 Colour board it works fine, I'm still very pleased with it...

I just uploaded a "short" clip of the board running, COWBOY SHOOT OUT:

https://www.youtube.com/watch?v=gBuBdyFz-mQ


Greetings, Wim...

[TheCorfiot]

Looking Fantastic Phill

[Prime]

Right I think I've got the colours working correctly now :

8 Colour test.

And Mono mode looks good too.....

Mono mode....

Cheers.

Phill.

[Multiwizard]

[oss003]

Nice job Phill, looks great....
Did you also test colours in textmode and in colourset2?
?#B002=8

Greetings
Kees

[Prime]

Nice job Phill, looks great....
Did you also test colours in textmode and in colourset2?
?#B002=8

Yep I believe your Manic Miner screen uses them and displays OK but will check.

Cheers.

Phill.