furon

Destroy All Humans! followed closely by Stubbs the Zombie. I love super stylized games like those two. The desire to play DAH on a PC is what got me interested in emulator development in the first place.

This may be a cardinal sin, but I just cut a hole in them. I use a hobby knife, start in the middle, cut out toward the side until I hit the edge of the screw hole, then follow it around. Same with the stickers.

I rebuilt the LPC port on my Xbox v1.6 today and had kind of a tough time with it, so I'm curious what everyone else uses to mod their consoles; type of wire, flux, solder, soldering iron/station, soldering tip, soldering temperature, cleaning solution, etc. I used some 30AWG wire-wrapping wire from Amazon, an Aoyue 937+ with the fine tip that came with it, RadioShack rosin flux paste, and RadioShack 62/36/2 rosin-core solder. I've soldered quite a few 0.1" headers on microcontroller and FPGA development kits with great success, but their pads were like twice as thick as those on the Xbox's LPC port. I had a hell of a time getting the pads to heat so that the solder would flow nicely. I also ended up burning the solder mask in a couple of places where I left the tip on too long trying to get the pads heated. Fortunately it still seems to work, but the joints look pretty bad. I think part of the problem was the fine solder tip, but I didn't have much luck with a chisel tip either.

Looking at the LPC spec again, I realized it doesn't matter anyway, because LFRAME# is absolutely required in order to recognize I/O read cycles, which put 0b000 on LAD during the CYCTYPE+DIR phase. A modchip would skip both START and CYCTYPE+DIR, then recognize the first non-zero I/O port adress nibble as the start of a transaction.

As we know, v1.2-1.4 Xboxes don't have an easily-accessible LFRAME# signal, so modchips work on the assumption that the southbridge will never abort a cycle early by asserting LFRAME# and driving 0b1111 onto LAD[3:0]. That may be true for memory read cycles, but what about memory writes and I/O read/write cycles? Besides a timeout (which shouldn't happen anyway if long wait states are used), is there ever a case where the MCPX might abort an LPC cycle? EDIT: I may have posted this in the wrong section. I need this info for a hardware mod, but it might be better suited in the Modchip or General forum. Sorry about that.

30?! Got any photos of your collection?

Additionally, on v1.6 consoles, pin 4 should read 3.3V while the Xbox is plugged in and show continuity with the left side of resistor R3G4 (to the left of the front panel connector, J2G1). Grounding it momentarily should toggle the Xbox on and off just like the power button does.

I've been contributing to the Xbox Dev Wiki for a while, but just found out about this site. Good to see a bunch of people still interested in the classic Xbox! My interests mostly lie in emulator development, but I've been doing a lot of hardware work lately. Currently working on a replacement for the Super I/O debug board with a USB interface, CDC serial port, on-board LPC flash programmer, and SMBus controller.