RGB to VGA Cable - Issue with display

[rabidgerry]

So I have been using a GBS 8220 board for a while and also the custom firmware gbs-control which all works beautifully so far and I'm very happy with the picture apart from faint jailbars which can still be seen.

I have been using a rgb connector wired to the GBS8220 board using the plug that came with it since the start and I figured I would eventually make an RGB to VGA cable myself someday and this would be a neater and handier way to connect the board to my Amiga. So today I finally made the cable (had made a test cable one time before using cheaper butchered VGA) and made sure I got the pins all correct and the 680ohm installed on the composite sync. Wired it all up making sure I was doing as neat a job as I could and then finally plugged it in. I was expecting perhaps a little picture quality improvement given that I had just used a heavily shielded short VGA to make this cable. But instead when I plugged it in and turned on my Amiga I got insane vertical lines.

I have no clue where I went wrong. Even unsoldered the resistor again and turned it around the other direction.

Could not shift these lines. Way to strong to ignore.

So I am wondering can someone tell me where I have went wrong or what might be the issue?

Just for the record, I was not getting anything like this (as in vertical lines) when using my old method of connection. Sure I can go back to this but I'd very much like to connect to the board using a VGA connection because when I finally get a housing for the board it will be easier to access via VGA in as opposed to the plug header I was using before.

See the attached photos.

[lesta_smsc]

Is that not a problem with the RGB on board? Usually capacitor issue. One of the colours probably out of tolerance. A colour chart may highlight which one.

Although someone had same issue in menu or something with WHDLoad and the fix was different though can't recall what!

[rabidgerry]

Not sure I understand what you mean? If I switch back to the previous method of connection the picture is fine again. So I was just thinking it was my cable. The lines change from worse to a little less worse if I move the cable about. Also running a rgb to scart to my tv works fine also so I don't think it's the Amiga at all (thankfully).

[Kin Hell]

Your cable is short a few wires....

Have a look here fella.

https://www.ikod.se/rgb-to-vga/

[rabidgerry]

Originally Posted by Kin Hell

Your cable is short a few wires....

Have a look here fella.

https://www.ikod.se/rgb-to-vga/

I see, but in that image it's showing the hsync and vsync in use. I'm using the composite sync with the 680ohm resistor in place to protect the amiga.

On a VGA if using c sync you connect that to pin 13 so I shouldn't be missing anything really

The earth wire I am lacking doesn't really matter because the braided shield around around the wires is actually the third earth/ground wire and it's touching those other ground wires so it's still grounded despite not being soldered.

I dunno

[robinsonb5]

I'm unclear what your original solution looked like - I'm not familiar with the GBS8220 - but it sounds like it's a pin header at the GBS8220? What was at the Amiga end?

Are you sure that pin header and the VGA port are exactly equivalent on the GBS8220?

It's just about possible that your new cable is in fact giving better signal quality, but the old one had enough parasitic capacitance to smooth out the lines! You could test that theory by soldering a small value capacitor (maybe 1nf or 2.2nf ceramic) between each of the R/G/B signal pins and ground.

I don't like the mixing of signal grounds and shield ground, but I guess if the VGA cable you used connects them internally then it shouldn't cause an issue.

[rabidgerry]

Originally Posted by robinsonb5

I'm unclear what your original solution looked like - I'm not familiar with the GBS8220 - but it sounds like it's a pin header at the GBS8220? What was at the Amiga end?

Are you sure that pin header and the VGA port are exactly equivalent on the GBS8220?

It's just about possible that your new cable is in fact giving better signal quality, but the old one had enough parasitic capacitance to smooth out the lines! You could test that theory by soldering a small value capacitor (maybe 1nf or 2.2nf ceramic) between each of the R/G/B signal pins and ground.

I don't like the mixing of signal grounds and shield ground, but I guess if the VGA cable you used connects them internally then it shouldn't cause an issue.

Rather than rebuild the original cable again myself(meaning dismantle the VGA-RGB cable again), take a look here at Ian Steadmans page

https://ianstedman.wordpress.com/201...iments-part-2/

scroll down to the "Synchronising the GBS-82XX"
part and you will see the exact solution I had deployed. There is obviously info regarding the resistor and why use the c sync also. Hopefully explains it a little.

[rabidgerry]

hmmm the issue may be because the VGA in on my board only accepts 31khz signal as opposed to the board header behind it which accepts 15khz. I need to get that confirmed though as I am near positive I wired a VGA cable up before and it worked fine.

[th4t1guy]

I get similar artifacting through RGB->VGA (using official Commodore branded adapter) on my amiga as well when connected to lcd monitors (tested on 2 separate monitors). However, all these artifacts go away if I use an RGB->HDMI adapter.
These artifacts also don't appear on my CRT monitor (using dbl screenmodes).
I wonder if this is just something specific to LCD monitors?

[Kin Hell]

Originally Posted by rabidgerry

I see, but in that image it's showing the hsync and vsync in use. I'm using the composite sync with the 680ohm resistor in place to protect the amiga.

On a VGA if using c sync you connect that to pin 13 so I shouldn't be missing anything really

The earth wire I am lacking doesn't really matter because the braided shield around around the wires is actually the third earth/ground wire and it's touching those other ground wires so it's still grounded despite not being soldered.

I dunno

Those Vertical Lines are usually indicative of Sync not being right.

Each Colour has it's own -ve or return. Always best to use each one rather than a common. A fault on the common will fault on R,G & B & is harder to diagnose issues by looking at the screen.

[rabidgerry]

I'm not actually using any adapters you see, I'm just trying to run the regular Amiga signal output into the up-scaler board. The board does all the conversion or signal manipulation. All I was trying to do was feed the board via the VGA input as opposed to the header plug because it will be handier once I house the board to plug into and the cable looks a dam site better than some rope plug that comes with the board.

[rabidgerry]

Originally Posted by Kin Hell

Those Vertical Lines are usually indicative of Sync not being right.

Each Colour has it's own -ve or return. Always best to use each one rather than a common. A fault on the common will fault on R,G & B & is harder to diagnose issues by looking at the screen.

I don't know enough about the board I am using and why the h-sync and v-sync are not to be used but all recommendations I have come across say to use the c sync with resistor in series. The resistor is to prevent too high a voltage from the Amiga damaging the board. The missing ground that I do not have, I already crocodile clipped it to the shield braid (which is that particular cables only other ground wire) and nothing happened.

I can either try another cable, or put it back to the way it was.

[robinsonb5]

Were you previously using the 5-pin header or the 8-pin header on the board?
Having done a bit of googling about the board, I think the 8-pin header is equivalent to the VGA-in, but the 5-pin header isn't - so if you were using the 5-pin header before that could explain the difference.

Either way, my next step would be to try wiring a VGA socket to whichever header you were previously using, so you can still use the cable you've just made, and see if that improves matters.

[rabidgerry]

Originally Posted by robinsonb5

Were you previously using the 5-pin header or the 8-pin header on the board?
Having done a bit of googling about the board, I think the 8-pin header is equivalent to the VGA-in, but the 5-pin header isn't - so if you were using the 5-pin header before that could explain the difference.

Either way, my next step would be to try wiring a VGA socket to whichever header you were previously using, so you can still use the cable you've just made, and see if that improves matters.

Oh I was using the 8 pin header.

Mind you only 6 pins have wires on the plug, and the yellow wire is not used.

Someone else has done something like that I found. They made an enclosure and wired a 9 pin vga to the 8pin header plug then used a RGB to VGA cable to connect to this new VGA port they installed.

Not sure if I can be bothered with that. In the meantime I'm going back to the original set up until I get another 680ohm resistor for the the csync. Then when I get that I can still have a separate cable for both connections (I've another RGB connector) until I figure out what the deal is. Perhaps the 8220 board just doesn't take Amiga 15khz.

[rabidgerry]

Ahhhh I guess this is what the issue is then

https://www.arcadeworlduk.com/conten...VGA-Manual.pdf

The manual it states on the second page VGA in only accepts 30.5k - 32.5k

Well there I go! Perhaps connecting my own VGA port to this isn't such a bad idea after all.

[Falcon]

I noticed something else in your second image - you seem to have a ground wire/braid connected to pin 15 on D23, which is _CCK_C in the schematics. Pins 16-20 are GND.

I don't know if this could be causing your problem, but it doesn't sound good regardless!

[rabidgerry]

Originally Posted by Falcon

I noticed something else in your second image - you seem to have a ground wire/braid connected to pin 15 on D23, which is _CCK_C in the schematics. Pins 16-20 are GND.

I don't know if this could be causing your problem, but it doesn't sound good regardless!

Jeeez how'd I miss that I probably thought it was on of those extra grounds. As far as I know 16 to 20 are all grounds on the rgb. I'm going to move that ground wire to a different pin and see if that affects anything.

I do vaguely re-call trying this type of lead before and it worked ok using a cheap cable, but I could be trippin

[Kin Hell]

Originally Posted by Kin Hell

<snip>
Each Colour has it's own -ve or return. Always best to use each one rather than a common. <snip>

Originally Posted by rabidgerry

Jeeez how'd I miss that I probably thought it was on of those extra grounds. As far as I know 16 to 20 are all grounds on the rgb. I'm going to move that ground wire to a different pin and see if that affects anything.

I do vaguely re-call trying this type of lead before and it worked ok using a cheap cable, but I could be trippin

[rabidgerry]

Originally Posted by Kin Hell

Each Colour has it's own -ve or return. Always best to use each one rather than a common

Your previous post has nothing to do with me using the wrong pin for an earth? Or at least it doesn't suggest that to me. I actually don't know what you are talking about here, I'm not THAT technical. Something to do with using all the grounds?

Also that pin 15 I used on the rgb, I used in the previous method of hooking up to the board and there was no issue. This is why I used it because it already had solder on it. Yes I should have payed more attention, but worth mentioning it didn't cause an issue before.

I will swap later and see what happens.

[rabidgerry]

Ok moved that ground wire from pin15 and soldered to pin16 and the lines are gone! Getting as good as I was originally. Thought perhaps there would be slight improvement over the 8pin header on the board (now I've a good shielded cable) but at least it's as good. Still very faint jail bars down the display which are not massively noticeable. Board runs great with the GBS Control custom firmware. It really gets rid of all the noise and artefacts that it suffered from before and the ghosting on fast moving sprites etc. Very sharp looking on my tv.

So now in my ignorance I ask myself, how is it working at all when the board spec clearly states VGA input takes 30.5k - 32.5k?

Is it perhaps a bad idea to keep using this way?