Cloning the GameCube component cable

[tueidj]

And it looks like shit thats like you saying that "I would like to install my wiikey fusion with wires then a megadrive because there is less of a chance of failure" that's crazy

Doing something right the first time so you never have to look at it again instead of using a fragile, unreliable, difficult-to-replace part that will break down in the future (unless it's perfectly straight and uses rounded silver plated connectors) is not crazy at all.

[Unseen]

I'd prefer wireporn over FFC any day, that shit is designed to break down and be non-repairable.

What!? Are you high you would rater have 22 wires then one FFC cable?

May I suggest a compromise?

ribboncable.jpg

(178.04 KiB) Not downloaded yet

[iamdablasta]

I'd prefer wireporn over FFC any day, that shit is designed to break down and be non-repairable.

What!? Are you high you would rater have 22 wires then one FFC cable?

May I suggest a compromise?

ribboncable.jpg

Very nice! Sturdy and not ugly.

[Unseen]

Very nice! Sturdy and not ugly.

Actually I don't like the blob of hot glue that much, but it was the easiest way to isolate the connections and provide mechanical stability.

[andre104623]

I'd prefer wireporn over FFC any day, that shit is designed to break down and be non-repairable.

What!? Are you high you would rater have 22 wires then one FFC cable?

May I suggest a compromise?

ribboncable.jpg

Nicely done, But in the future people without the skill of building and soldering these boards (or whatever comes of this) will like a plug-and-play or QSB instead of the mess of wires. I can deal with the wires for now but I would like a plug-and-play like a clone of the component connecter. I was thinking of ripping apart my non-working component cable for the connecter and making a plastic box which would house the GCvideo and on the end of the box would be component cinch connecters but this is all in my head and I didn't even get started building yet.

[tueidj]

May I suggest a compromise?

That's why I still have a box full of MFM HDD and floppy ribbon cables, even though the drives are long gone.

[LOCtronicz]

Alright so I have compiled my first board. However I have zero experience in programing a FPGA board. When I use the regular programming mode with the .jed file I get this error:

Device#1 LCMXO2-640HC: Failed to verify the ID
(Expected: 0x012B9043 Read: 0x00000000).

ERROR - Check configuration setup: Unsuccessful.

ERROR: pgr_program failed.

ERROR - Programming failed.

With Fast RAM and .bit file I get the same. When I scan for devices it also says it can failed to scan the board. My status light on the porgrammer is green so it it working as far as I know.

Anyone have a little insight on what I'm doing wrong?? Can't wait to test it out

Here are some pics of my assembly:

[Unseen]

Alright so I have compiled my first board. However I have zero experience in programing a FPGA board. When I use the regular programming mode with the .jed file I get this error:

Device#1 LCMXO2-640HC: Failed to verify the ID
(Expected: 0x012B9043 Read: 0x00000000).

Did you connect the board to a power source, e.g. the Gamecube? Generally the VCC line of a JTAG programmer is only used to set the correct I/O voltage for its drivers, it does not power the circuit.

[LOCtronicz]

no I did not hahaha. Just powered it of an external 3.3v source and both jed and bit programming was a succes. Thanks gonna hook it up to my cube and see if all is working

[andre104623]

Alright so I have compiled my first board. However I have zero experience in programing a FPGA board. When I use the regular programming mode with the .jed file I get this error:

Device#1 LCMXO2-640HC: Failed to verify the ID
(Expected: 0x012B9043 Read: 0x00000000).

ERROR - Check configuration setup: Unsuccessful.

ERROR: pgr_program failed.

ERROR - Programming failed.

With Fast RAM and .bit file I get the same. When I scan for devices it also says it can failed to scan the board. My status light on the porgrammer is green so it it working as far as I know.

Anyone have a little insight on what I'm doing wrong?? Can't wait to test it out

Here are some pics of my assembly:

Very nice good job what kind of programmer are you using? I have a xilinx dlc95st and a nand-x from my Xbox JTAG days I'm hoping I won't need to buy a programmer just for this project

[Duhasst0]

Do you have the nand x to coolrunner programming attachment? Cause if not you probably won't be able to use that unless you solder wires from the serial port on it.

[Ashen]

This is so exciting. Zen that looks awesome man.

[LOCtronicz]

Yes I can confirm that it is working identical to the original MX chip. On PAL and NTSC console. I have tested almost all my games, homebrew etc.. and it works the same. Quality is the same as far as I can see with my bare eyes and it will also works with only 1 ground line. I have it connected now with only 13 wires. 3 Audio lines, 12v, 27mhz are not needed. I would advise to use all ground lines as stated before. This will provide better stability. For portable use however (which is what I'm planning to use it for) 1 thicker wire is good enough. As Mega said only when we have many boards tested and working with no complaints can we really say goodbye to that ridiculously expensive component cable!! YES!! UNSEEN YOU ARE THE MAN!!!

For all the others who want to attempt this if you have decent soldering skills this is very doable. Here is my Digikey order list. They have all the parts in stock.

1x SN74LVC14ADR / IC HEX INVERTER / 296-1699-1-ND
1x LCMX02-640HC-4TG100C / IC CPLD / 220-1492-ND
1x CDK3404ATQ48 / DAC 8bit Triple 180MHZ / 1016-1935-ND
9x C0805C104K5RACTU / Cap Cer 0.1uF / 399-1170-1-ND
2x RC0805FR-074K7L / Res 4.7k Ohm 1/8W / 311-4.70KCRCT-ND
2x F920J106MPA / Cap Tant 10uF / 478-8115-1-ND
1x CIB21P110NE / Ferrite Chip 11 Ohm / 1276-6349-1-ND
4x RC0805JR-0775RL / Res 75 Ohm 1/8W / 311-75ARCT-ND
1x RC0805FR-07348RL / Res 348 Ohm 1/8W / 311-348CRCT-ND
1x RC0805FR-07100RL / Res 100 Ohm 1/8W / 311-100CRCT-ND
11x C0805C103K1RACTU / Cap Cer 10000PF / 399-1159-1-ND

Also if you don't have a compatibile programmer get this one:

http://www.ebay.com/itm/111284428437?_t ... EBIDX%3AIT

This is the clone Unseen was talking about. It works like a charm in the Lattice programming software. Note you have to flash 2 times. One time the bit file and one time the jed file. At least that's what I had to do to make it work.


Here a final pic of it wired up:



Now gonna make the rest I have and test it with the n64

[andre104623]

Time to sell those component cables before price drops down. It will be some time till die heart Nintendo fans find out about the clone but when they do price will really drop since supply will increase and demand will drop. I will hold on to mine I like having a complete GameCube collection

[Unseen]

I have a xilinx dlc95st and a nand-x from my Xbox JTAG days I'm hoping I won't need to buy a programmer just for this project

As usual the Lattice software only works with Lattice programmers (or their clones), but if your Xilinx cable is compatible with Xilinx' Impact tool it may be possible to use it with an SVF file to program the Lattice chip. I'll try to test that later and add an SFV file to the repo.

This is the clone Unseen was talking about. It works like a charm in the Lattice programming software. Note you have to flash 2 times. One time the bit file and one time the jed file. At least that's what I had to do to make it work.

For me flashing the internal config memory appears to work fine using just the .jed file while direct SRAM upload (much faster during development, but obviously not power-down-resistant) only works with the .bit file.

Time to sell those component cables before price drops down.

I would love to see a graph of eBay prices for these cables against time =)

[andre104623]

I have a xilinx dlc95st and a nand-x from my Xbox JTAG days I'm hoping I won't need to buy a programmer just for this project

As usual the Lattice software only works with Lattice programmers (or their clones), but if your Xilinx cable is compatible with Xilinx' Impact tool it may be possible to use it with an SVF file to program the Lattice chip. I'll try to test that later and add an SFV file to the repo.

This is the clone Unseen was talking about. It works like a charm in the Lattice programming software. Note you have to flash 2 times. One time the bit file and one time the jed file. At least that's what I had to do to make it work.

For me flashing the internal config memory appears to work fine using just the .jed file while direct SRAM upload (much faster during development, but obviously not power-down-resistant) only works with the .bit file.

Time to sell those component cables before price drops down.

I would love to see a graph of eBay prices for these cables against time =)

That would be cool to use my xilinx programmer. I saw on ebay they have lattice/xilinx programmers I really don't want to spend more money on a programmer again if it helps this is what I bought http://www.ebay.com/itm/Compatible-with ... 20e97c5b8a

[Unseen]

That would be cool to use my xilinx programmer. I saw on ebay they have lattice/xilinx programmers I really don't want to spend more money on a programmer again if it helps this is what I bought http://www.ebay.com/itm/Compatible-with ... 20e97c5b8a

Looks a lot like mine, which I just successfully used to flash my prototype to N64 mode and back to Gamecube with Xilinx Impact. =) The SVF files for that are now on Github, they write to the internal config flash of the XO2 as this is probably what most people want.

Getting Impact to write a non-Xilinx chip is a bit annoying, especially as the quality of that program is on a similarly high level as most of the other GUI programs included with ISE. First, you need the BSDL file which is available here - in the documentation table click on "BSDL model", then download the "[BSDL] LCMXO2-640HC TQFP 100" entry. Renaming the file from *.bsm to *.bsd may be helpful, but is probably not required. Next, connect your JTAG programmer to the board and the PC and start Impact.

In Impact, choose to create a new project, select the option to configure devices using boundary scan (JTAG) and in the drop-down select the automatic connection method (should be the default). Impact should now scan the JTAG chain and find one unknown device, so it asks about a BSDL or BIT file. You just downloaded the BSDL file, so choose yes and select it in the dialog box - do not select the Lattice .bit or .jed files, Impact will throw a lot of errors and may need to be restarted to return to a sane state.

After selecting the BSDL file, a "Device Programming Properties" dialog pops up which will most likely be almost completely empty. Just hit OK which should give you access to the main workspace showing the JTAG chain with just one device in it. Right-click it and select "Assign new configuration file...". In the following dialog, you need to select the .svf file from my Github repository - even though the .bit and .jed file will be listed there, they're merely a great way to get Impact to spew a lot of errors until it is restarted, so make sure to select the .svf file. Another great example of the high code quality will show up, a Yes/No dialog titled "Assign File" with no text in it. "Yes" is the correct choice here. If succesful, the device icon in the workspace will change from a chip to an SVF icon with multiple small chips inside it (it's symbolic, the file actually expects exactly one chip).

Finally, you can right-clock the SFV icon on the workspace and select "Execute XSVF/SVF". After a minute or so the FPGA should be programmed.

(Yes, this sequence could be improved a lot by some screenshots along the way - I'm too lazy to do that right now, even though it would probably be nice to add it to the project wiki on Github some day)

[andre104623]

That would be cool to use my xilinx programmer. I saw on ebay they have lattice/xilinx programmers I really don't want to spend more money on a programmer again if it helps this is what I bought http://www.ebay.com/itm/Compatible-with ... 20e97c5b8a

Looks a lot like mine, which I just successfully used to flash my prototype to N64 mode and back to Gamecube with Xilinx Impact. =) The SVF files for that are now on Github, they write to the internal config flash of the XO2 as this is probably what most people want.

Getting Impact to write a non-Xilinx chip is a bit annoying, especially as the quality of that program is on a similarly high level as most of the other GUI programs included with ISE. First, you need the BSDL file which is available here - in the documentation table click on "BSDL model", then download the "[BSDL] LCMXO2-640HC TQFP 100" entry. Renaming the file from *.bsm to *.bsd may be helpful, but is probably not required. Next, connect your JTAG programmer to the board and the PC and start Impact.

In Impact, choose to create a new project, select the option to configure devices using boundary scan (JTAG) and in the drop-down select the automatic connection method (should be the default). Impact should now scan the JTAG chain and find one unknown device, so it asks about a BSDL or BIT file. You just downloaded the BSDL file, so choose yes and select it in the dialog box - do not select the Lattice .bit or .jed files, Impact will throw a lot of errors and may need to be restarted to return to a sane state.

After selecting the BSDL file, a "Device Programming Properties" dialog pops up which will most likely be almost completely empty. Just hit OK which should give you access to the main workspace showing the JTAG chain with just one device in it. Right-click it and select "Assign new configuration file...". In the following dialog, you need to select the .svf file from my Github repository - even though the .bit and .jed file will be listed there, they're merely a great way to get Impact to spew a lot of errors until it is restarted, so make sure to select the .svf file. Another great example of the high code quality will show up, a Yes/No dialog titled "Assign File" with no text in it. "Yes" is the correct choice here. If succesful, the device icon in the workspace will change from a chip to an SVF icon with multiple small chips inside it (it's symbolic, the file actually expects exactly one chip).

Finally, you can right-clock the SFV icon on the workspace and select "Execute XSVF/SVF". After a minute or so the FPGA should be programmed.

(Yes, this sequence could be improved a lot by some screenshots along the way - I'm too lazy to do that right now, even though it would probably be nice to add it to the project wiki on Github some day)

Your the man. Thank you unseen you saved me some money. There's a lot steps but I will figure it out

[webhdx]

Awesome! Now we need someone to clone BroadBand Adapter

[MockyLock]

And we need someone to start a little batch...

[iamdablasta]

And we need someone to start a little batch... :D

I think we should wait for the testing before we order more batches.

[MockyLock]

My post was in the same "irony" than the one from webhdx.
Obviously it needs more testing. Specially after the fireworks of andre104623.
Humm, speaking of that, i wonder if we shouldn't ask megalomaniac or Ashen, or even another specialist like LOCtrnics, to think about a fire/water proof shell for the GC :p

(Please andre104623, don't take that bad, it's just for joking. If you feel this as rude, let me know).

[looney bin jim]

Awesome stuff man!! Thanks for this!

[LOCtronicz]

Anyone tested the n64 flash file yet? I have flashed my board and wired it to the n64 as mentioned but on my 15khz compatible screen I get a white screen. When I load donkey kong I get a white screen then it boots up with a video signal and then after the Rare logo it goes back to a white screen again. Composite does send a signal. So I assume it has to do with the flash file.

@Unseenhow is it working for you? Also in your notes you state a D7 pinout. However all the n64 has are D0 to D6. Am I missing something here?

[Unseen]

I have flashed my board and wired it to the n64 as mentioned but on my 15khz compatible screen I get a white screen.

But your TV/monitor does detect a signal? That would mean that at least the sync bits are correctly picked up from the N64 and sent to the output. You don't happen to have an oscilloscope or logic analyzer available for remote diagnostics? =)

how is it working for you? Also in your notes you state a D7 pinout. However all the n64 has are D0 to D6. Am I missing something here?

It's working fine for me, otherwise I wouldn't have released it. The D7 in the pinout was a mistake, that input on the GCVideo board should be left open.