Dreamcast Internal VGA modification

I don't see anything wrong with what your list has in it except the capacitors are supposed to be 220uF not 2200uF.

 

By the way would there be any issues with using internal composite and s-video mods with those VGA mod

Yeah kinda figured the capacitor wasn't right thanks for confirming that

 

I recently did this mod and it works great. The only changes I made were the placement of the switch and I added connectors to everything so I can swap the circuit into a different DC more easily if I need to in the future.

I'd also like to be able to output 15kHz RGB from the DE-15 connector so I can run my DC on my Sony PVM-2530. Can I do this by only grounding pin 7 (as opposed to both pins 6 & 7 for VGA) from the DC PCB and combining the H-sync and V-sync wires?

 

I just finished this mod and it works great! This was my first ever soldering project. It is awesome being able to relive some DC classics without the craptastic composite video. A few tips I found helpful.

1) Use flux on the DC solder points. The difference is night and day. I used http://www.amazon.com/gp/product/B008OC0E5M/ref=oh_details_o08_s00_i01?ie=UTF8&psc=1
2) Use thin (<=0.022") solder as it makes adding solder to the DC pads much easier. Also I found the 63/36/2 silver bearing solder to be better than 60/40.
3) If you're in the US, Radioshack has overpriced components. I found Tayda Electronics (http://www.taydaelectronics.com/) had great prices. Shipping was a bit slow, but well worth the savings.

 

I designed 2 different internal variation of the schematic and I can't for the life of me get this to work. Do pins 6 & 7 need to both be grounded in order to go into VGA? I seem to look at the examples and think one is grounded while the other isn't and I assume this is one of the problems so I soldered those pins together and had both grounded when the switch was in the right spot. I'm following mmmonkey's guide and have made a board to solder onto the test points on the bottom of the video socket and tested for continuity from those points to the board with the VGA socket. On the following pictures the ground pins are connected correctly to the ground plane of the boards so just ignore the no connection of the GND points, I simply removed them so it will be easier to follow the traces.

Component Board:



Adapter board:


and Pin out of Test Points on mmmonkey's site:


can you guys see anything out of the ordinary? or is it the 2 computer monitors I tried that might be the culprits.

 

This is the correct schematic if you use a bufffer IC (recommended for compatibility with CRT) :

If you don't use the IC you can simply bypass it and connect the H/V sync directly to the VGA connector. Note that DC AV port pins 6 and 7 and connected together and put to ground to enable VGA-out.

Note that the schematics with diodes are resistors are wrong. It's simply a design that passed from people to people. Those components aren't necessary. It's better to keep the buffer though, to insure maximum compatibility with screens.

FG

 

God damn! Back to the drawing board, so just to be sure the way I have it setup like the MMMonkey setup won't work? Thanks for the info FG.

EDIT: Ah shit I just realized that the capacitors are backwards on my board and might be the problem I'm having.

 

The + should point towards the DC, this probably is your problem. Though as I know you're probably planning on selling those, you could really remove the R/Diodes and add a buffer to avoid any hassle. There's also sufficient space on your PCB to add a scanline generator on the back. I've worked on that a bit last months but never found the time to finish the design.

 

My initial design didn't have space for the buffer or scan line generator but this definitely can and it's still a WIP till it works how I'd like but I will follow your schematic for the next revision.

 

Why do you say that? The original Sega VGA box had them, and they seem to make perfect sense to me. When you disable VGA mode, the buffer is switched to tristate mode and the V and H sync lines float - the resistors provide a fairly weak pullup that will hold the lines high when the buffer is disabled but won't source enough current to interfere with the output signal when the buffer is enabled.

The diodes also make sense because dragging the output of a CMOS chip below the negative rail (say through cable ringing or ESD) has a risk of inducing latchup in the output stage and this can damage it. Maybe not actually necessary, but certainly not a bad idea - and because they are always reverse biased during normal operation they are not going to interfere with anything.

 

Well, maybe wrong was a bit harsh, but it's clearly not necessary, especially for a device that'll be on vga mode 99% of the time after the mod.

 

Anyone able to confirm/deny this not working using the caps on USA/Can DC? Not sure why it would make a difference.

 

I've yet to get mine working but that could be alot of other things and not sure the Caps are the culprit but I will post here if I get this working or not.

 

Well, I got a cable that's based exactly on this design and it works quite well in my NTSC|U console (a Canadian console, but that doesn't matter at all here).