exxos's Atari, Amiga & retro forum

Icky wrote: 28 Nov 2018 22:17

• Cyan band has a greenish tinge in the darker part
• Yellow band has a reddish tinge in the darker part
• Magenta has a blueish tinge in the darker part
• White has a greenish tinge in the darker part

At first I was wondering if the RGB bits were mixed up somehow on the board but they are correct.

It kinda looks like there are lower bits of RGB mixed up somehow :shrug:

Icky wrote: 28 Nov 2018 22:17 Main conundrum is why games like Gold Runner, Magic Pockets work fine but the dark background at boot has small white on the TOS message and purple dark themed desktop.

It might be worth checking the RGB digital inputs to the DAC from the shifter and the 32MHz line for noise or bad ringing...

It is odd that the same mix up of colours isn't there with the first 3 RGB bars but only visible when mixing the RGB signals. Does the shifter in the STF output the same signal levels as the falcon one? Could it be the chips response to a voltage level too high etc?

If it was a binary signal id be thinking a bit shifted off the end of something like that.

I will try and get my sockets changed tomorrow afternoon and see if I can kick the STF back into life now that I know where my problem lies.

Normal digital input levels >2V logic 1, <0.8V logic 0. My only though is if the shifter has ringing on its outputs...

Saying that... does the shifter need pull ups on its outputs I wonder...

I just got home, checked the Falcon DAC, all GND, VCC wired as I have them on my DAC board, so no issues there.

exxos wrote: 28 Nov 2018 22:29

Icky wrote: 28 Nov 2018 22:24 Don't know how I missed that one. I'll have a look at trying that out.

I think with a binary input, you should see voltage steps like this on the output (before the big cap on the video board)

So try RED inputs first, 0000 should give for example 1.787V output, and 1111 should give 3.178V... probably about 0.1V per step.

Hmm - odd

0000 = 2.65V, 1111 = 3.17V

Actually the jumper bit makes the biggest difference,

Icky wrote: 28 Nov 2018 23:18 Hmm - odd

0000 = 2.65V, 1111 = 3.17V

Actually the jumper bit makes the biggest difference,

ahh, my bad, higher bits are only going to give a low voltage jump.. so its probably right..( we are only using 3 higher bits) but please check all binary numbers and note the output voltage.. you should see a even step in voltage as binary number goes up..

EDIT:

Ignore the jumper setting, leave it connected to 0v for these tests...

exxos wrote: 28 Nov 2018 23:20

Icky wrote: 28 Nov 2018 23:18 Hmm - odd

0000 = 2.65V, 1111 = 3.17V

Actually the jumper bit makes the biggest difference,

ahh, my bad, higher bits are only going to give a low voltage jump.. so its probably right..( we are only using 3 higher bits) but please check all binary numbers and note the output voltage.. you should see a even step in voltage as binary number goes up..

EDIT:

Ignore the jumper setting, leave it connected to 0v for these tests...

Now I really am confused - Have been using the Blue line as its easier on the board.

B0,B1,B2
000 = 2.64V
001 = 2.66V
010 = 3.02V
011 = 3.07V
100 = 2.61V
101 = 2.62V
110 = 3.06V
111 = 3.03V

Looks like B2 does nothing, and B1,B0 look swapped :shrug:

exxos wrote: 28 Nov 2018 22:48

Icky wrote: 28 Nov 2018 22:17 Main conundrum is why games like Gold Runner, Magic Pockets work fine but the dark background at boot has small white on the TOS message and purple dark themed desktop.

It might be worth checking the RGB digital inputs to the DAC from the shifter and the 32MHz line for noise or bad ringing...

Just for reference. Probe is on x10.

32MHz line:
B0 line:

They should be OK I think...

I can only assume really that your chip is faulty as I can't see anything wrong anywhere. Either that or the pinout in the data sheet is wrong..

Will try and build mine up tomorrow and see what happens..