Bringing the case and PCB together

[rygD]

I'm back with another stupid idea.  Looking at Asmo's retro Pandora mock-ups got me thinking about a few things, and most can be done after production, but the keyboard must be done before, unless we can print our own high quality key caps, and maybe mats.  Would anyone find dark red, or any color other than black, for all or some of the keys acceptable?

For reference: http://pandoria.org/gallery/asmo-cases/

 

[Damon]

3. The LCDs Nikolaus had a small error in the wiring for the Solomon chip, but already fixed it: two traces were switched. A trivial fix - except for that everything is pretty small. Here is a microscopic view of the fix. The wires you see are about 1,5 times as thick as a normal human hair... so yep, it's pretty small! solder.jpg

Hi,

Thanks for the update, I'm already in line to buy my Pyra and i'll sleep in front of of your house if I've too

But I'm just curious about this point, what does it mean? this will be the 'final fix' or the PCB will be printed again for the final release with the correct 'track'?
 

 

[Granitehead]

Thanks for the update, I'm already in line to buy my Pyra and i'll sleep in front of of your house if I've too

But I'm just curious about this point, what does it mean? this will be the 'final fix' or the PCB will be printed again for the final release with the correct 'track'?

Of course it will be produced again (it has probably only be produced a couple of times, how do you produce thousands of Pyras containing the part without producing the same amount of parts? ;-) ), probably there will be at least a few additional iterations of /everything/ before it's final.

 

[matzesu]

According to Evildragons Twitter, Nikolaus managed to finish to debug the Main PCB, and there are only 3 Bugs left yet..


I cant bring a link, because im now on my Smarthphone..

 

[EvilDragon]

Not bugs, but issues. For example, traces too close together, etc.


It's really complicated to design such complex PCBs avoiding Micro-Vias and staying within the technical boundaries.


That's quite a bit of a puzzle

 

[matzesu]

Whit "Vias" you mean "Short Cuts" (Kurzschlüsse) because the traces are to close togehter because the limited space??


And this is just the Main PCB, which is, when i remember corektly the Big PCB whit the Key Contacs, and the Conektors for Wifi, USB, etc on it..


So the CPU PCB could be lot more divicult to make..

 

[Granitehead]

Vias are connections from one layer to another, matzesu. If I remember right, the main PCB is just two-layered: that is, there is one layer on the top and one on the bottom. Accordingly vias are connections from the top to the bottom. Traditionally they are visible as a kind of hole with a metal inlay. I assume micro-vias are a more modern variant of this, which takes less room and isn't visible like that.

 

[matzesu]

Vielen Dank T4b jetzt bin ich etwas schlauer/ thangs very much T4b, now i know what ED meant


And if these VIAS arent on the right Place, there could be issues whit the PCB just as Short Cuts, or lost Conektions..


Im glad when the Main PCB is finishd,


But then they have to make the CPU PCB, and the Cooling System, and all the little Things whe need for a Pandora Like Device just as Batterie Comparment, Speakers, Etc

 

[fusion_power]

Not bugs, but issues. For example, traces too close together, etc.


It's really complicated to design such complex PCBs avoiding Micro-Vias and staying within the technical boundaries.


That's quite a bit of a puzzle

How many layers will the final SoC PCB have? 

 

[EvilDragon]

The SOC one? Probably 6 - 10.

 

[TeDaDeS]

Will Nikolaus also be at the Gamescom? (guessing 'no' but nice to meet another Pyra team member)

I was wondering how you choose the pull-ups/downs for the Pyra, and if so how to decide what the best trade-off would be (price, space, averages, reaction-time, power draw, etc).

 

[EvilDragon]

Nope, Nikolaus won't be there.

Well, deciding which LCD to use depends on refresh time, support, availability, pricing, etc.

A lot to take care of

 

[rygD]

Nope, Nikolaus won't be there.

Well, deciding which LCD to use depends on refresh time, support, availability, pricing, etc.

A lot to take care of

Can we get a short video of the LCD testing when there is time?

 

[fusion_power]

Nope, Nikolaus won't be there.

Well, deciding which LCD to use depends on refresh time, support, availability, pricing, etc.

A lot to take care of

I guess at first you have to choose a screen that actualy WORKS, then comes the other factors.  It seems that the choosen Pyra configuration with extra rotation chip takes longer to get LCD's work than just use the "easy" way unlike todays smartphones do.

So or so it will be a complex task to find a supported screen that matches all the needed requirements, including the obligatory "Sonic the Hedgehog" refresh test.

 

[EvilDragon]

Easy way smartphones do?

What are you talking about - the reason we don't have the displays running yet it because the driver has not been written and Nikolaus didn't have time to do that, as he was finishing the main layout PCB.

These drivers have nothing to do with the Solomon chip (which is already working, btw).

Any smartphone developed these days has to get drivers written as well, there is no "easy" way.

 

[fusion_power]

Ah, ok then. I thought the tough part is to make the screens working with the rotation chip and that's the hard part where the drivers needs to be written for (to rotate the screen). Didn't checked the part about the actual display driver for the Pyra Firmware itself. ^^" I just had in mind that Smartphones mostly use their OS to rotate the screen and do not use special dedicated chips.

 

[EvilDragon]

The chip basically only needs to be enabled and does all the magic. That's a lot easier than the display driver.

 

[TeDaDeS]

Nope, Nikolaus won't be there.

Well, deciding which LCD to use depends on refresh time, support, availability, pricing, etc.

A lot to take care of

Would have been cool to meet him, but meeting you again and the Pyra will be fine by me.

My question was actually regarding the resistors used in the Pyra. Although I think the same thoughts are applied for all parts used.

As the pull-up/down resistor will influence how the controls handle (a tiny bit) and how little power is used by them.

It is generally agreed that certain ranges work, but wondering if you decided to use specific ones and if so why you choose these.

 

[fusion_power]

The chip basically only needs to be enabled and does all the magic. That's a lot easier than the display driver.

Sounds good, so at least this part is already solved. And for sure it will be an big advantage over any rotation solution via software or even via SoC. 

For the display driver, I wonder if this can alredy be finished before an actual "Pyra OS" exists or doesn't that matter? Is the current (standard) distro you use already enough to make it all work?

 

[EvilDragon]

The chip basically only needs to be enabled and does all the magic. That's a lot easier than the display driver.

 Sounds good, so at least this part is already solved. And for sure it will be an big advantage over any rotation solution via software or even via SoC. 

For the display driver, I wonder if this can alredy be finished before an actual "Pyra OS" exists or doesn't that matter? Is the current (standard) distro you use already enough to make it all work?

The driver is completely OS independent, it's part of the kernel.

You can use the kernel to boot any Linux OS (Ubuntu, Debian, Arch, etc.)