Creating a Development Kit (DVT-5)

[Bowlsnapper]

34 minutes ago, applesauceman said:

https://github.com/zatchbot/OGX_LPC_Breakout

SICK. I was wondering if this existed. Perhaps I should give it a try.  Thank you, Saucer!

 

@applesauceman What revision of the Serial Debugger is this? The UART header looks different and there is 5V power on the corner!

[Bowlsnapper]

https://www.amazon.com/RS232-Converter-interfacing-microcontroller-device/dp/B01M4PZA7R

I may choose to install this UART to Serial converter in order to have native serial output on the Devkit. It's just an idea, as the USB-C FTDI serial output should already do what I need, but It would be very DXT-4-ish.

[applesauceman]

5 hours ago, Bowlsnapper said:

SICK. I was wondering if this existed. Perhaps I should give it a try.  Thank you, Saucer!

 

@applesauceman What revision of the Serial Debugger is this? The UART header looks different and there is 5V power on the corner!

Rev C. I built a handful of these a few months before they released the Rev D. I'm planning on assembling another half dozen of this revision though. USB C is incredibly convenient. Do you have a parts list handy from your recent mouser order?

[Bowlsnapper]

On 4/8/2024 at 7:39 PM, applesauceman said:

Rev C. I built a handful of these a few months before they released the Rev D. I'm planning on assembling another half dozen of this revision though. USB C is incredibly convenient. Do you have a parts list handy from your recent mouser order?

@applesauceman You should be able to upload this into Mouser. I don't have time for a true BOM file right now (who knows; maybe this actually IS a BOM and it's simpler than I think it is. lol), but if you go to tools on their website you should be able to easily upload and translate this Excel spreadsheet to a "BOM". I tested it and it loads perfectly, creating a useable cart for checkout. I made sure to include proper quantities and Customer Reference numbers so you can follow the silkscreen part designations.
 

Edit: @applesauceman BTW, I had to substitute the 14.3181MHZ crystal for another one that was still being manufactured. The pinout is the same and it is compatible, according to both the datasheets. The footprint is slightly different, but the contact spacing and placement is more or less the same. If you must have the original crystals, they are available on ebay from China, although, I can see you've built a few recently so I'm sure you've settled on a solution at this point.

SuperIO BOM: https://mega.nz/file/JqNHkaKK#tw1_BUJcyYm6pITErpMtecFOayQmK3zlAkBKJWbTRsA

[KaosEngineer]

19 hours ago, Bowlsnapper said:

 It seems that the Debug Kit simply had 128MB of RAM and the only debugging came from the dashboard or bios and transferred through LAN. As far as I know nothing else is different from a stock/retail console. However...

IIRC, they also have a different MCPX chip v2 instead of the retail unit's v3 version.

[Bowlsnapper]

27 minutes ago, KaosEngineer said:

IIRC, they also have a different MCPX chip v2 instead of the retail unit's v3 version.

What does this mean in practical terms? Was serial debugging possible with the MCPX X2, but not used? I guess it may have just been what came before the X3 and that's all.

----------------------------------------------------------------------------------------------------------------------

 

The fuzz has since been removed.

 

[applesauceman]

10 hours ago, KaosEngineer said:

IIRC, they also have a different MCPX chip v2 instead of the retail unit's v3 version.

The v2 southbridge had no chain of trust protocols. In practice the difference between v2 and retail is you dont need hacked bios to bypass the security protocols durring the boot process. This is the reason using a recovery XDK disk on a retail console will brick it.

[Bowlsnapper]

So for one reason or another, I am not getting the dubugger to work properly. Even on a 1.0 where LFrame is guaranteed to be working. I think I need to build another one and try again.

This project is on hold I guess.

[Bowlsnapper]

I am taking this time to iron out the PC side of this. There are no real XP drivers for the USB FTDI serial host, so I think I'll have to add a serial (RS-232) port via PCMCIA and then add another RS-232 port to the xbox via the UART output of the SuperIO debugger.

https://www.ebay.com/itm/264727392495?chn=ps&norover=1&mkevt=1&mkrid=711-117182-37290-0&mkcid=2&mkscid=101&itemid=264727392495&targetid=1493661221511&device=c&mktype=pla&googleloc=9031634&poi=&campaignid=19851828444&mkgroupid=145880009014&rlsatarget=aud-1412318123216:pla-1493661221511&abcId=9307249&merchantid=6296724&gad_source=1&gclid=CjwKCAjw8diwBhAbEiwA7i_sJfxniuhIlZOnTDQvyuX6-ksX9fzFcHCejxDt1JzZSYbTnitUuku7YRoCuTIQAvD_BwE

https://www.amazon.com/RS232-Converter-interfacing-microcontroller-device/dp/B01M4PZA7R

[Bowlsnapper]

I got the debugger working. First, some pins were loose on the SMSC chip linking to the LPC pins. Now, it will not work with the 1.4 board I'm trying to use, because the LFrame via is borked. I already know, man. I gouged the shit out of it.

 

Anyway, if I have to use the 1.0 that I used here to test the unit, then I will, but I'm gonna try to get the 1.4 Lframe working. Project is back on track and the BOM I posted is good.

[Bowlsnapper]

The relocated debugger RX and TX LEDs are working.

[Bowlsnapper]

Well, It's basically done. I didn't get the DVD "Ready" light to work (it kept fucking with my DVD ROM drive and making it act funny), but I did get the activity light going, separate from the HDD activity LED. LAN is there as well. The two on the left are the RX and TX LEDs for the Super IO debugger. I'll test more tomorrow with the XDK and debugging through LAN, but it looks like this project is good  

[Bowlsnapper]

The HDD and LAN activity LEDs have the cathode go to the point indicated. DVD points, however, must have anodes go to the point.

[Bowlsnapper]

 

Debugger wired up with relocated TX & RX LEDs

Quite the shitstorm of wires, I know. But as long as they're never seen... fuck it.

I'll hide the wires better

Here it is. SuperIO RX and TX lights working on the left. On the right side, from left to right: DVD activity, DVD Ready, HDD activity, LAN activity. 

This was a long project, but worth the work. I wanted to do a Development kit, but with a little more than the LAN debugging that came with the regular Debug kit. This is more like a mix between the DVT-2 (Debug Kit) and DVT-4 (Development kit). The DVT-4 had serial debugging at the kernel level, as opposed to just LAN (XBDM). 

I wanted to internalize the serial debugger, so I did, keeping things neat. The only thing is, I would not have view of the Transmission/Receiving LEDs if I did so, so I soldered molexes to the SMD LED pads on the debugger and relocated the LEDs up front. At first I had issues with the debugger, but I managed to finally find some loose pins on the SMSC chip on the debugger linking the chip to the LPC. Once that was fixed, I had to swap out the 1.4 board I was planning to use with a 1.0, since I determined that Lframe could not be made to work on the 1.4. I basically trashed the via on the edge of the MCPX and couldn't repair it. JESUS that thing is fragile. So I plugged it into the 1.0 and I got output through Windbg. Awesome. Then I saw the video of the status LED array and wanted to do the same thing, so I did. A very statusy and verbose console I have here. I call it the DVT-5. Lol. Obviously the console is upgraded to 128MB of RAM. The SuperIO debugger case is printed green (off color but still) as well as the fan shroud. Also printed some green feet since 3 were missing and the one remaining was falling off. The clips used for the LEDs are 3D printed as well. I went with PETG instead of resin, since I have shit luck with resin and everything I make with it turns into crap. lol.

The software setup is TSOP flashed Cerbios 2.3.2 Hybrid flagged as debug in the ini. I wanted to keep the LPC free for the SuperIO ribbon. You can do an LPC breakout and have both the debugger and a modchip, but it served no purpose for me and I wanted as much space as possible. I have a PBL out to retail (XBMC) and will create a shortcut in the future so that I can load XBMC straight from the XDK if I need to debug through it for whatever reason. The HDD structure is just like my other HDDs and consoles. 2TB with XBMC as the main and based on the Origins image. The debug flag in the ini is what gets me to boot into the XDK dash.  However, I still need to learn to code, so I will probably start with hello world, and then write my first app installing and then enabling or disabling the FPS overlay DXT extension.

The USB-C breakout I used is here: https://www.adafruit.com/product/4090 This is what fits into the Rear USB-C breakout board mount STL I included in the bundle. Buy one, it's only a few bucks, but get the 1 Ft cable, too. Forgive me for not doing a pinout, but a pinout can be done yourself with a multimeter and a cable with the wires exposed on one end and the other plugged into the jack. That's what I did. You have to put the cable through the RF shield and then solder the wires to the board... unless you wanna just cut some of the grille off the RF shield. However, you do need to fit it into one of the bottom vents (the bottom left, looking at it from the inside of the case) and you will see a channel in the bottom of the fan shroud for the USB cable to go through. Line it up with that. Keep in mind that you will need a 60MM fan. I used a Noctua.

I have included in this package all the 3D prints, wiring diagrams and info (Gerber, BOM) for the SuperIO debugger that I used to make this development kit. If you need anything, ask me.

Debug Kit Goodie Bundle: https://mega.nz/file/JqdiFTqb#Ca4eF9tDnwTdIUDDyewTpFyj7p-Jqa73LLHCZmvV5jU