Release Postman's "PIIza TIme" Custom Wii Regulator Boards + Preorders

[Postman]

Hey everyone!

Sorry for being MIA for so long, but I return with tasty memes.

I present to you all: PIIza TIme. A custom Wii regulator board built off of 2x TPS54395 switching voltage regulators with around ~90%-95% on the 5.0v and 3.3v lines and ~80%-85% efficiency on the 1.15v and 1.0v lines. Also has a pad that can brought high or low to turn on/off the regulator board with a 2 position switch.



I started this project after MGC last year as my first ever PCB design project. After two boards that were garbage, some help from @Gman ( u baber), and many butt tears, I finally have a design that kicks ass.

The name is broken into two parts "PIIza" because it's shaped like a slice o' pizza and fits perfectly in with an "OMGWTF" trim to complete the rectangle it almost makes. It's also a shout out to @Aurelio because he is a badass piiza man, thanks for all your work man! "TIme" because it's built off of a TI regulator.

PIIza TIme does get pretty hot, but unfortunately I don't have a FLIR or even a thermocouple to test exactly how hot. In my testing it did not seem to affect performance of the regulators but I would recommend putting some type of small heatsink on top of the chips. Doesn't have to be fancy, something small would do with a little thermal pad would do. Raspberry pi heatsinks would do nicely.

Spoiler: Voltages measured under load (Double Dash)

Spoiler: PIIza TIme in action

There are two options on how to get your filthy mitts on this board:

1. If I can get people to commit to at least 10 boards all together, I will be selling them at $30 a piece + shipping (around $3 in the US, $10 international). Just gotta hit the 10 board mark to make it worth the time required for assembly.

2. Make your own, the design is now open source for BitBuilt to do as they please. Files + BOM in my github here:

https://github.com/PostmanMods/PIIza-TIme

I got the design practice I wanted out of this project so I release it into the world for people to do as they please with it. All I ask is if you use it just throw me the credit for the design if you use it in your design.

I hope to see this in portables ASAP! Let me know below if you are interested, just leave your username, the desired quantity, and whether you are US or international:

 

[Aurelio]

I have a concern. You say you get 90-95% efficiency, but also that the regs get really hot. How is that possible?

 

[Postman]

I have a concern. You say you get 90-95% efficiency, but also that the regs get really hot. How is that possible?

A very good point, I should probably not write this at 2am lol

Two distinctions to make here:
1. Only the 1.0v/1.15v regulator chip gets hot to the touch, that is because:
2. The efficiency of those voltages drop to about ~80% percent at max current draw of 1.84A on the 1.15v line and ~85% on the 1.0v line with a max current of 1.54A

These numbers I am getting from the data sheet on page 11: http://www.ti.com/lit/ds/symlink/tps54395.pdf

Two additional reasons the chips may be getting toasty:
1. I only had a chance to test a single board so far, it may be the connection to the thermal pad underneath the TPS chip itself.
2. The thermal pad connects to the ground plane which is fairly small since the board itself is small. Considering the data sheet expects you are using this in a larger application, the ground plane may be too small of a heatsink and may need an additional one.

A little bit more testing needs to be done on these boards to figure the exact cause of that, but I think I have it nailed down pretty what it could be and it should be not too hard to figure it out from here. I just have to make another board using solder paste to solder the chips to the thermal pads to ensure my thermal pad connection is 100% connected, which is what I am currently thinking is a contributor to the issue. If that doesn't seem to at least help then it is probably a lack of available ground plane to use as a heatsink.

Thanks for pointing that out, I will update my OP.

 

[MikaMaslak]

Interested to try this out!
Does the efficiency you assessed it at deteriorate over a long operation time?
Also do you have a pic of the back and edge of an assembled one?

 

[Postman]

Interested to try this out!
Does the efficiency you assessed it at deteriorate over a long operation time?
Also do you have a pic of the back and edge of an assembled one?

AFAIK the efficiency is directly tied to how much current is drawn from the regulator not the time it has been used.

I can update with more pictures when I am back in the modstation!

 

[Postman]

Hello everyone, back with an update!

I revised the PCB to improve the chip thermals. The board now has exposed ground plane with vias around the chips to help keep the TPS regulators cool.

I also added an attiny10 for pulsing the U10 point at startup. The program is designed to run at startup and pulse the U10 line for a second or so and then remains high @ a stable 3.3v. You can also attach a switch to the u10 pad that connects to ground for a makeshift reset button if you so choose.

PCB files are up on my github and updated BOM are up on my github. Can't seem to find the attiny10 code right now but Ill upload it once I find it.

Pics to follow once I receive the boards from the fab house!

 

[GingerOfOz]

You can also attach a switch to the u10 pad that connects to ground for a makeshift reset button if you so choose.

How thoroughly have you tested this feature? I feel like messing with the U10 to engage a reset isn't a great idea, espescially when a switch between a pin and ground... is how reset normally works anyways, but in a less jank way.

 

[Postman]

How thoroughly have you tested this feature? I feel like messing with the U10 to engage a reset isn't a great idea, espescially when a switch between a pin and ground... is how reset normally works anyways, but in a less jank way.

I tested it probably a dozen times with no instability but YMMV, that's why I said if you choose. I remember vaguely you explained why this isn't a good idea but I can't remember your exact answer. Could you elaborate?

 

[Nold]

I really like the idea of this PCB!


... but whyy??

Also it seems like i've to host a soldering bootcamp when i'm at nobble

 

[cheese]

I tested it probably a dozen times with no instability but YMMV, that's why I said if you choose. I remember vaguely you explained why this isn't a good idea but I can't remember your exact answer. Could you elaborate?

The U10 line isn't designed to reset the system. On top of the fact that is is possible to corrupt the NAND doing this, it doesn't always reset the system.

 

[Postman]

I really like the idea of this PCB!


... but whyy??
View attachment 8276

Also it seems like i've to host a soldering bootcamp when i'm at nobble

oof ouchie ow my soldering pride.

But in all seriousness YES! I will take all the advice and help I can get. Ain't nobody perfect.

The U10 line isn't designed to reset the system. On top of the fact that is is possible to corrupt the NAND doing this, it doesn't always reset the system.

Noted. No shorting U10 to ground for reset. Not even if I REAAAAALLY want to.

 

[Nold]

oof ouchie ow my soldering pride.

But in all seriousness YES! I will take all the advice and help I can get. Ain't nobody perfect..

no offence dude - i was mostly wondering why the wires are crossed. would be dope if the vout pads just swapped places.