In my last update, I mentioned my plan to narrow the project’s scope and simplify my adapter. In some ways, I succeeded. In others, I went the opposite direction. Then came the month-long wait for PCBs and components to arrive so I could test my new creation.

What did I change?

First, I removed the mounting spot for the DC-DC converter, which made the board slightly smaller—a small but meaningful improvement.

Second, I addressed an old issue I’d been wanting to solve. As I mentioned in Update 5, the Xbox draws most of its current from the 5V Standby rail of the PSU, making it incompatible with standard ATX power supplies that cannot provide 10 amps on that rail.

Interestingly, the PicoPSU and its derivatives don’t have this limitation. They don’t care whether the power is drawn from the 5V or the 5V Standby rail. However, exploiting this feature always felt wrong. Why rely on a single pin to handle so much power while ignoring five others designed for the job? Doing things the right way could even extend the PSU’s lifespan.

A new approach: Dynamic power switching

I started searching for a solution and came across the LTC4416 controller. This chip can dynamically switch between two power sources while maintaining a stable voltage output.

Figuring out how to use it was a huge challenge. Reading the datasheet, calculating component values, and designing the connections pushed me into new territory in electronics. But it was also a rewarding learning experience.

One of the LTC4416 connection diagrams

My KiCad schematics are now looking much cleaner than before (though, as a warning: don’t copy them—they don’t work yet!):

The idea was simple: the chip would detect which power rail is active. If only the 5V Standby rail is powered, it means the Xbox is off, and the chip would supply power from that rail. When the 5V rail wakes up (indicating the Xbox is turning on), the controller would switch to the main rail to provide power.

What went wrong?

Unfortunately, this solution didn’t work as expected. For reasons I haven’t yet figured out, I couldn’t power on the Xbox using this setup. Worse, the chip always supplies power to the 5V rail, even when it shouldn’t—definitely not the behavior I wanted.

The LTC4416 controllers also turned out to be quite expensive, which would have negatively affected the price of a complete adapter unit.

Next steps

Now, I need to troubleshoot what went wrong and look for a better solution. I’ve already started experimenting with the simpler and cheaper LTC4412 controller, which should be easier to wire up and use.

This may work

I knew this journey would be long, but every failed attempt tests my patience. Still, I won’t give up. I’m determined to make it work. And when I do, I’ll finally take a break, fire up my Xbox 1.6 with the first-ever ATX PSU adapter, and play Halo 2.

If you want one of these adapters too, don’t forget to join the waitlist. Until the next update!