Sega Genesis USA VA6, brightest blue is... dim.

[eatnumber1]

Has anyone seen this before?

The brightest blue on my genesis is... not bright. That's affected the brightest white too, which is tinted yellow as a result.

I just recapped the board, so unless I accidentally installed a defective capacitor, it's unlikely to be caused by that. The board was non-functional before recapping. I also checked the blue traces into and out of the CXA1145P video encoder and there are actually no capacitors at all on those traces.

[eatnumber1]

one bump in the hopes that someone has any ideas

[paulb_nl]

It looks like blue is too strong on all levels. It peaks too soon and the white color bars have a blue tint.

Does this also happen with the composite output? If it does then the blue input to the encoder might be too strong.

[James-F]

First thing first, got another tv in the house?

[AndehX]

all levels look too bright. By the looks of it, you need to bring the contrast of your TV right down.

[viletim]

eatnumber1,

Does the composite video look bad too?

Yes - check the connections/components between the graphics processor and the video encoder an the blue line.

No - check your AV connector and cable for corrosion. Also check your RGB cable has the correct resistors/caps inside.

[eatnumber1]

Thanks for everyone's suggestions:

I don't have a composite cable with which to test. I'll pick one up and report back.
I have a second Genesis which looks fine using the same cables, but hasn't been recapped.
Turning my TV's contrast down dims the whole screen, and doesn't change the relative differences in colors, so the dim blue is still dim.
I don't have easy access to another TV. I'll bring my gear over to a friend's house when I get the chance.

[mikejmoffitt]

We would do well to put this on a scope and monitor the blue channel from the VDP outwards, to see where in the chain this error occurs.

[eatnumber1]

There's a scope at work I can probably use. That will take me a bit longer to sort out, but I'll do that too and report back.

[FBX]

Looks to be an internal issue. Normally when the brightest blue is being clipped, it just simply blends as the same color of the next shade of blue. In this case, the only way it can be darker is an actual internal malfunction. I can't imagine it being a trace issue from the VDP to the CXA becuase then you'd get no blue at all. So I'm more inclined to believe something bad happened to the CXA. Only way to know for sure is to bypass the CXA with a mod board (Voultar recently released one on OSH Park). If that mod were to show the same problem, then you're looking at possible chip failure.

[eatnumber1]

I was considering CXA chip failure too. I Didn't know about Voultar's board. I've been thinking about building Arcade24's bypass board, but Voultar's looks simpler to do.

Unfortunately, although I found Voultar's post about it https://twitter.com/voultar/status/1013 ... 17?lang=en, his oshpark link is 404. Anyone have a working link?

[Syntax]

Overloading the rgb lines will have strange effects to the color bars. I had bar 7 for all colors the same as bar 2 and it was because I had an rgb bypass and stock amp connected at the same time.

Inspect the 3 pull up resistors and 3 in lines resistors before the cxa. Should be able to test resistance in circuit.

[FBX]

I was considering CXA chip failure too. I Didn't know about Voultar's board. I've been thinking about building Arcade24's bypass board, but Voultar's looks simpler to do.

Unfortunately, although I found Voultar's post about it https://twitter.com/voultar/status/1013 ... 17?lang=en, his oshpark link is 404. Anyone have a working link?

https://oshpark.com/shared_projects/i9DVwUR4

You can also visit here for the BoM:

https://www.mouser.com/ProjectManager/P ... 3954234850


-FBX

[eatnumber1]

I picked up a composite cable, that shows the issue too.

I borrowed a friend's oscilloscope, connected leads to the input & output of the RGB amp for both blue and red channels. Here's what I saw:



Purple and teal is the blue line, blue and yellow are the red lines (input and output respectively). You can clearly see the voltage drop at the highest step for the blue in both output and input.

I then tried tapping the blue line of the video processor (SEGA 315-5313) directly, and got this:



There's one surface mount capacitor between the CXA chip and the 5313, which I desoldered and it still shows the same issue.

I also took a shot of a good color line. Here it is for comparison:

[Syntax]

Did you isolate the blue line by removing the inline and pullup resistors before you did the last test?

[eatnumber1]

No but I put the scope on the VCC line and it showed a stable 5 volts

[FBX]

No but I put the scope on the VCC line and it showed a stable 5 volts

You need to make sure it's not the pull-up and series resistors on the blue line. The scope plot is showing something is pinching off the peak amplitude for the blue channel.

[Fevian]

Hello! I had this same problem just now. I reflowed the joints on all the resistors between the VDP and the RGB encoder on the blue line, and it's fixed! Hope this helps.