DOL-101 Gamecube and XenoGC clone

I'm trying to install a XenoGC in a DOL-101 platinum Gamecube, but I'm having issues. Basically, no matter what I do, I'm only getting a red status LED (no green or amber/yellow) flashing in sync with the disc drive. Also, the console only manages to read original discs about 10% of the time, while it reads them 100% of the time without the chip. I haven't managed to launch any burned discs either, while they work fine with my Wii.

The chip I have is an obvious clone because "XEN8 GC" is written on it, instead of "XENO GC", but given it's just a PCB with an Atmel microcontroller and a few SMD components what could go wrong? I mean people make their own homebrew clones, so another clone should be just as good, and at least it's honest about being a clone.

I started with a wireless install because I was confident I could pull it off, but seeing the issues I got, I told myself I possibly did not succeed. So, I carefully desoldered the chip with desoldering braid and a solder sucker in order to proceed with a wired install, as it would allow to test my connections.

While I had the chip off, I used the occasion to test the console without it, and the drive worked well with original games, so at least I knew I didn't damage it.

Once I was done with the wired install, I tested my connections with a multimeter and I concluded that all my connections were correct, that they were all in the right spot, and that there was no solder bridge. I then tested again, and I got the exact same symptoms.

Looking for solutions, I tried tweaking the laser potentiometer. The original resistance was 356 ohms which seems high to me, for a DOL-101. Still, I patiently lowered it 10 ohms at a time, down to 150 ohms, but I always got the same problems every time (all 20 of them!). On a few occasions, maybe around 230 ohms and 180 ohms, the console could launch an original game on first try, but it wasn't really reproducible. 2 in 20 is the same success rate i had with the potentiometer at its original value. Once I reached 150 ohms, I didn't dare going any lower, by fear of frying the laser. So I put the potentiometer back to its original value and desoldered the chip, yet again.

And once more, with the chip off, I tested the console and it could read original games just fine. So I didn't fry the laser with the potentiometer tweak. While I had the chip in hand, I took the time to verify the PCB itself and make sure that the traces were proper. With a multimeter, I made sure that it followed the schematics presented on the gc-forever wiki, and it did, with the exceptions of the two resistors for the LEDs being lower-valued (100 ohms vs 330 ohms) and there being a second capacitor (for AVCC) which follows Atmel's recommendations better.

For my last attempt at finding a solution, I used my Raspberry Pi as an ISP programmer and reflashed the chip. I had already done that in the past to succesfully flash an Arduino Mini. It wasn't too hard to wire the chip to the Raspberry Pi because most of the pins used for ISP programming are the same ones used for accessing the Gamecube drive. The only one missing was the reset pin, but the PCB trace with the chip's name on it leads right to it. So I carefully scraped the solder mask off the two holes of the "8" in "XEN8", hereby turning it into an "O" and a usable solder pad.

Once I dumped the contents of the flash memory with avrdude, I discovered that the microcontroller had its locking bits set. I also discovered that its fuse bits weren't set for brown out detection and 0 ms boot, which apparently isn't optimal. (However, the locking bits might have prevented me from seeing their true state.) Since I couldn't get a proper dump to compare against, I took a shot in the dark and erased the chip so as to reflash it with the hex file in the XenoGC source code and I also set the lfuse to 0x84.

I then put the chip back, once again. And... I still have the same issues. From a bit of googling, it seems that there are some Xeno GC clones that only use red LEDs (how intuitive!), but looking at the reflection of the chip in the metal plate, I'm pretty sure that's it's always the same LED that's flashing red which would indicate an improper install. However, the connections are all good, the firmware is freshly and properly flashed, the fuse bits are well set, and it still doesn't work.

Furthermore, when I do manage to luckily launch a game and hold the start button, I get a screen saying "XenoShell v1.05" in garbled text, but apparently that's because it's programmed to run in PAL instead of NTSC. Shouldn't this show that the chip is installed correctly? Or is the garbled text indicative of a problem?

I'm at loss. As far as I know, the XenoGC should work with any Gamecube revision, from any region. Therefore, I doubt that there may be any compatibility issue. If I buy another, more expansive XenoGC (clone or otherwise) from a more reputable resaler such as EurAsia, how am I to know it would work any better?

The two choices I'm faced with right now are to either get my hands on a different Gamecube (preferably a DOL-001 instead) and install the chip on it, or try my luck with a XenoGC from a different supplier.

 

Yes try another xeno gc. I have a bunch. If you need one pm me. I have installed 2 of them so far out of this batch and they both worked just fine.

 

i got a 10 pack of xenoGC's from Robert at Eurasia , all work great so if your looking for a reliable seller grab some from him or turfster

 

Sounds like you are not using media/burner that works well with the cube.

The fact you get the xeno menu means the install is fine, as is the chip.

You want to get ritekg04 media ideally and a burner that works well with it.

The cube is EXTREMELY fussy with media.

 

Yeah, I know about the Gamecube being very picky, and the choice of media is another problem I'll have to solve, but I don't want to tackle it right now because the fact that I can't even reliably launch original games with my apparently faulty XenoGC is an even bigger problem. I mean, I might already have good media on hand, but I wouldn't even be able to know.

 

Put said media into your burner and load up img burn. You will get the media id on the right.

Any xenos you get now will be clones, but like you say - it's a chip and couple of passives, not much to get wrong.

 

I never got to the bottom of it myself, that perfect sweet spot. I make sure mine can at least read the swiss disc and from there I just load stuff from sd gecko as even when it does work with ritek mini dvds I still get DREs ocasionally.
On my Viper GC with a full case mod I noticed that full size ritek G05 (from memory) still work more reliably than minis
I would also try a different burner if you have one.
Not being able to read originals with the chip installed is a worry. First time I'm hear it would mess with reading the original discs sorry I can't help there

 

Ritekg04 work best in my experience (but, if your burner doesn't work with them well, then it's still not going to be good)

But he ritekg05 are also up there in cube compatibility.

 

btw use IMGburn for GC burning , it creates less C1/C2 errors than other burning software. Also i use Verbatim DVD-R

 

I've got some Taiyo Yuden TYG02's, Memorex RITEKF1's and Verbatim MCC-003-000's. All quality media that are said to work according to this list: http://club.myce.com/f33/gamecube-blank-dvd-media-229874/. So that's not the problem. I'm also waiting for some printable 8cm media I took a chance with to arrive. If they end up not being compatible, there's a seller near where I live that has Philips mini DVDs (but they're sadly not printable).

ImgBurn has been my burning software of choice for years . I always burn at 4x speed with verification, no matter what I burn.

 

I have ty and verbs. They don't consistently work on all cubes, I can assure you.

Also burning at slower speed doesn't improve burning quality. Some media/drive combinations work better faster.

 

there is a thread/sticky on ClubmyCE mentioning with todays burners burning at the slowest speed is worse. Sure in the early days slowest was best but that's not the case now , I burn GC at X6 if i burner faster or slower than this i get more C1/C2 errors meaning a poor burn

 

I had a cube that refused to play any burned media. I did a replacement no named laser from eBay and it fixed the issue

 

I thought about buying a new laser. Did you have problems reading original games as well, after you've installed a chip?

 

No original games always worked fine before and after the xeno.

 

This is the media I burn the backups to:
https://www.amazon.com/gp/aw/d/B0056BYHAM/ref=ox_sc_saved_image_10?ie=UTF8&psc=1&smid=A3ISABYPHHUJT8

They worked great and were cheap. Personally I'm looking to install a megadrive V4 when the next batch comes out. I chatted with badassconsoles and he told me it would be at least another month. Something for you to consider at least.

 

If anyone is interested in the history of xenoGC, the different versions (and how little it matters), and the clone situation (original xenoGC production ended in 2009), then you are welcome to read this story:
http://www.eurasia.nu/modules.php?name=News&file=article&sid=2740


The original source code and binary files to flash xenoGC to version 2.0 (shell v1.05) is here:
http://www.eurasia.nu/modules.php?n...etails&lid=7802&ttitle=xenoGC_2.0_source_code

NOTE: Flashing/programming the xenoGC is not easy, but can be done, some soldering required.


The official site for xenoGC can still be found over at archive.org:
http://web.archive.org/web/20061004053437/http://xenogc.com/

 

Yeah, it's not really the version of the firmware that worries me, it's mostly the hardware differences between the clones because I'm still trying to find a cause to my failed install. I'd also rather receive a chip that has LEDs of different colors, because two red LEDs doesn't tell me much at all other than the chip is powered.

However I may have found a likely cause and that would be my own fault. Typically, microcontrollers are designed to allow the voltage applied to their GPIO pins to stray away from GND and VCC only within a small margin. For the ATMEGA8A/L, that's VCC+0.5v for a logical high and GND-0.5v for a logical low. Outside of that, and the microcontroller can be damaged.

For a wireless install, the VCC pad of the XenoGC is a circular hole and it's possible for solder to merely form a meniscus in it without making contact with the solder pad below. Furthermore, the pad on the Gamecube DVD drive board is connected to a rather beefy trace, so it requires extra heat for it to accept solder. Without a connection to VCC, VCC is effectively 0v and a logical high on any of its GPIO pins at 3.3v becomes too much so the microcontroller can be damaged.

It's likely that I was in that situation during my initial wireless install. I didn't have the presence of mind to identify test points on the Gamecube DVD drive PCB that have continuity with the chip's solder points beforehand (likely because a few of them actually go to the other side through vias) so I didn't test my work with a multimeter afterwards. Looking at the PCB now, the capacitor that has a hole in the XenoGC meant for it (next to the LEDs) actually has continuity with VCC and GND on either of its sides, so that's something I'll keep in mind once I try again with a new chip.

That doesn't really explain why I could still successfully flash and dump the contents of the microcontroller's flash memory afterwards through ISP. Maybe the damage made it so that it still works well at 5V, but not at 3.3V.

 

Yes, quick soldering (wireless) sucks pretty bad in general.

It's never a good idea to have board (PCB) soldered directly against board, besides the risk of shorting something out between the boards during install, over time changes in heat will expand and contract the boards which stress the solder joints to the point of cracking. Always do install by wires, or in the case of xenoGC, use "wire snip" method when soldering (pogo style), so you get a few millimeters of gap between xenoGC and the drive board held by very short non-insulated wires.

 

By that reasoning, any surface mount component will fail the same way - and it doesnt.

Nothing wrong with the quick install, long as you do it properly.