Cloning the GameCube component cable

[Unseen]

Should've checked again before mentioning it, my bad. $15 isn't bad for a flasher at all, though I get the feeling it might only work for their dev boards?

Yes, they built a neat hack that converts from the (asynchronous) RS232 signal to one of the (synchronous) serial configuration modes the Spartan supports. If you prefer an adapter that works with all Xilinx chips, there are lots of cheap clones of the Xilinx Platform Cable USB available on eBay which works with Xilinx' own Impact tool in all of its glorious flakiness. In some cases you can also use it to program chips from other manufacturers if you can create an SVF file that Impact accepts.

[novenary]

Hmmm, that TXDI chip looks like it's just a buffer or something, am I right ? What's the logic level needed to flash the board ?

[Unseen]

Hmmm, that TXDI chip looks like it's just a buffer or something, am I right ? What's the logic level needed to flash the board ?

I have successfully used a generic FT232RL board with 5V output level to flash the board using KNJN's software, although I had to use FT_PROG first to select the "Invert TX" option. There is some kind of level-shifting on the RX line on the board, according to the PDF manual you can even connect it directly to a regular RS232 if you just want to upload a bitstream to the FPGA. Initially I tried to use an FT232RL board with just 3.3V output, but that seems to be too low to successfully pass through the circuit on there.

If you want to program the SPI flash (and for using the board with a Cube you very likely want to), you need a two-way connection, so hooking the board directly to an RS232 port isn't an option for that.

I didn't mention any third-party USB-RS232 interfaces in the programming section of the GCVideo DVI section because I suspect that the required chip configuration (inverted TX, noninverted RX) and voltage level (5V works, 3.3 doesn't) would cause a too high chance of confusion and programming failures "in the field" - but feel free to prove me wrong. =)

[the_tim]

Just registered to say that this is awesome.

Also: Regarding the wire porn of connecting to the GC, has anyone considered modeling a connector that uses some standard pin (e.g. Digi-Key PN WM2624-ND) and using a 3D printer to fab it? I have a decent 3D printer but don't have good enough 3D modeling skills yet to pull off something like this...

[bobrocks95]

Can you 3D print with conductive materials these days? And would the average Joe be capable of making a mold of the connector, or do you have to get a factory or some other place to do that (who would then probably demand a minimum order of thousands of units)?

[Unseen]

the_tim: Interesting idea!

Can you 3D print with conductive materials these days?

You don't have to - you just design a piece of plastic with a number of holes in the correct positions to insert pre-made contacts. The part number on Digikey he referenced is an example of such a contact, although it would be too wide for this purpose. ATX power plugs are constructed in a similar way - the plastic part is made without any contacts in it and the contacts are plugged into the housing later, after they have been seperately crimped to the cables. Using a 3D printer it may be possible to create a housing that would fit the digital video port, although I wonder how fine the print would need to be.

The other common way connectors are made is by putting the contacts in a mold and injecting plastic around it - I have never seen the plug on an original component cable, but I suspect they are built this way.

And would the average Joe be capable of making a mold of the connector, or do you have to get a factory or some other place to do that (who would then probably demand a minimum order of thousands of units)?

Getting a complete connector made would most likely require a rather high minimum order, but if it's 3D-printable small-scale production would be feasible.

Actually someone has contacted me before in private, asking me to exclusively license my project to them because then they believed they would be able to finance a custom-made connector. I refused - I prefer it if my work is available to the public(*) and it would've been unfair to those people who were already looking into producing the boards themselves.

(*) Don't you hate it if you find out that some years ago someone built exactly the thing that you would like to have, but then stopped distribution for whatever reason and never published the information needed to build more? (hey, that even matches the original component cable =) ) There are probably quite a few projects fitting that description and I'm annoyed every time I find one, so my stuff tends to end up as open source if I can be bothered to document it enough for a release.

[the_tim]

Yes, as Unseen described, I'm talking about printing a housing that you would slide pins into. I'm in electrical engineering and have dealt with many of these types of connectors over the years and with a 3D sitting in the room next to me, I just put two and two together.

The specific Digi-Key part number I mentioned was just meant to be an example of the type of pins I'm talking about. I figured it wouldn't be the right size, but I bet there is one that would work. If someone with the original cable took some detailed dimensions and photos of the OEM connector I might use this as an excuse to finally hone my 3D modeling skills.

If I really had lots of time on my hands I'd take this and run with it. I bet with a few tweaks to the PCB, a 3D-printed mold, a large PCB order and a reasonably-priced PCB assembly house we could probably produce a run of 300-500 of these for $50 a piece. Throw up a Kickstarter to make sure you get enough committed orders before placing the expensive order for the full PCB run and bulk buy of the components, and you're good to go.

I'll be keeping my eye on here, because if anyone else does it before time magically appears in my schedule, I'm totally in.

[MockyLock]

AS far i can remember, someone is doing some connector mold at ASSEMbler forum. Maybe he could give some hints.
I can't check for now (children crying for starve), but i'll try to do so in the afternoon.

[megalomaniac]

one problem with 3d printing will be the limitations of strength
the pins sit within a channel that has about a 0.5mm wall to the edge

that is hardly enough material to provide a suitable strength and could lead to quick failure of the connector
also to mention, a 3d printer fab should require at least a 1mm wall thickness and even then 1mm is still weak


how accurate are home based 3d printers to be able to provide a detailed resolution at 0.5mm ?
i know some allow 0.3, but even then the issue at hand still comes down to overall strength of the connector...

[the_tim]

how accurate are home based 3d printers to be able to provide a detailed resolution at 0.5mm ?
i know some allow 0.3, but even then the issue at hand still comes down to overall strength of the connector...

Well I haven't exactly put this to the test, but my 3D printer (Form 1) is one of the ones that can do 0.3mm.

There's one way to find out whether it would be do-able...

[megalomaniac]

if fabs will usually not allow less than 1mm walls, you think a 0.5mm will have the proper strength...
there are other areas of the connector that use less that 0.5mm....that area is where the middle where the digital port board inserts to the connector...
if i had to guess, and looking by my best guess, its easily less that 0.5mm
that area itself should be considered the most crucial and will have the weakest walls if 3d printed

[MockyLock]

I found the link
http://www.assemblergames.com/forums/sh ... A-V-Socket
Do'nt know if it useful, but at least, i've brought it to you.

[novenary]

That's not the digital port but still a nice thing.

[bobrocks95]

He was citing it as an example of a community-made, 3D printed connector.

[novenary]

Yeah I know, the problem is this is a female connector, the male side is a bit trickier.

[bobrocks95]

Unseen, did you have any success in fitting the Pluto II board into the Gamecube? Does it seem feasible?

[Unseen]

Unseen, did you have any success in fitting the Pluto II board into the Gamecube? Does it seem feasible?

I haven't mounted my board inside the Cube at this time because I'd have to desolder all of the pin headers and it would make further development harder, but I believe it should fit in the space above the original connectors. Don't some existing mods use that area for VGA or Component connectors?

[novenary]

There should be a little bit of space between the drive and the cage where the board might be able to fit.

[RMwashere]

*Bump*. Guys you don't know how long I have been following this board. I would really be stoked if this thing became a reality.

[thepiet]

*Bump*. Guys you don't know how long I have been following this board. I would really be stoked if this thing became a reality.

Yeah me too, I am really looking forward to seeing Luigis Mansion in 480p finally

[megalomaniac]

.....have patience

[RMwashere]

Just checking to see if any progress has been made. Still looks pretty rad.

[dacide]

Hey Unseen, check it out (Audio over HDMI ):
https://github.com/charcole/NeoGeoHDMI

[Unseen]

Hey Unseen, check it out (Audio over HDMI ):
https://github.com/charcole/NeoGeoHDMI

Yes, I've seen that. I'll look into it after the OSD is completed.

[pr0ton]

Amazing to see this work, very impressive. Also great that you shared it with everyone the same time you had it working. Other members around here also were busy on it but rather kept it for themselves.

From now on, would it be very hard to include DisplayPort as well? I wonder because since I have this board and an eDP dispay, that would make it for me the best semi-portable solution .

Also, does the Wii have digital video signals that can be tapped before entering the DAC?