Finalizing the PCBs

Discussion in 'Pyra News' started by EvilDragon, Nov 27, 2017.

  1. SneHebNor

    SneHebNor Still Fresh

    Joined:
    May 15, 2016
    Messages:
    57
    Great news! Btw. I would buy a transparent case and leave the original one (black?) as backup if it breaks. Can I order here and now? :~)
     
  2. second exodous

    second exodous Well-Known Member

    Joined:
    Sep 27, 2005
    Messages:
    2,705
    Location:
    Utah, USA
    SNES game grey. . . I think that would be a color I'd go for. I really want a red one like on the site first though. So my color preferences in order would be

    1. Pyra Red
    2. SNES Game grey
    3. Classic Black

    Like what was said before clear would be awesome until I get sick of the light bleeding. I'd be curious how dying clear cases would go though. Could a clear case be dyed a solid color or would it remain semi-transparent?
     
    Gremblo, Djhg2000, Djoga'Ro and 3 others like this.
  3. levi

    levi Still fresh, damnit!

    Joined:
    Oct 6, 2008
    Messages:
    8,143
    Location:
    Somewhere off the coast of the EU
    A dyed case would likely remain translucent if it's possible to dye polycarbonate (I've never tried). However, if it does you could probably paint the inside with opaque white paint to fix that. I recently had success painting the inside of a white plastic box which contained a distracting LED backlight with black emulsion to block unwanted light bleed. Or, you could just paint the inside with whatever colour you wanted to get a very similar effect more simply.

    I'd advise against people spray painting their Pyra though. Many spray paints contain acetone as a solvent, and acetone weakens, discolours and dissolves plastic in sufficient quantity. I don't know how long it takes versus how long it takes for the acetone to evaporate as the paint dries, but either way plain brush-on emulsions are usually water-based so you're not taking any risk with those. With either paint solutions you'll have trouble not letting the paint pool in corners, which is an issue with my blacked out white plastic box in that there are lighter spots visible along the faces when the ceiling light is off at night, but should be less of an issue with an externally lit subject.
     
  4. Tanasinn

    Tanasinn AESTHETICS

    Joined:
    Oct 28, 2015
    Messages:
    36
    Well yes and no. Yes if it is left in the solution for the right amount of time it'll come out pretty great. No if it is left in for far too long. But by the time the case has been in the solution for that to happen, the acetone used to bind the dye to plastic would have destroyed the case well before it reaches this point. The trick is the ratio of acetone to the water-dye mix on top of the time needed to get the shade desired. Even with re-reading and looking up the sheet we have for props, the case from Form Action might need a prewash of sorts before the actual dying. Due to the limited amount of transparent cases available. The plan is to just get one [maybe two] , make molds out of that, then make new cases with the materials we know work well with our dyes instead of taking an extreme gamble.
     
  5. Silent-Hunter

    Silent-Hunter Advanced Member

    Joined:
    May 29, 2010
    Messages:
    2,474
    So we're waiting on another prototype CPU board, and then if that works is it time for mass production? Have I got that right?
     
  6. sebt3

    sebt3 homebrew player (P. & C.)

    Joined:
    Sep 9, 2008
    Messages:
    4,682
    Location:
    France
    The lastest mainboard revision have to be tested too. But otherwise that's my feeling too
     
  7. EvilDragon

    EvilDragon Administrator Staff Member

    Joined:
    Mar 4, 2003
    Messages:
    19,064
    Location:
    Ingolstadt
    Yep. Latest mainboard revision is already being tested, CPU-Board is being finalized right now.

    Sent from my XT1650-03 using Tapatalk
     
    Djhg2000, Alberto, Cralex and 8 others like this.
  8. ClockworkCoder

    ClockworkCoder Active Member

    Joined:
    Jan 21, 2016
    Messages:
    345
    Location:
    UK
    It's starting to look like a fantastic way to end the year; all the pieces finally slotting into place :)

    2018, year of the Pyra! (forget dogs, it's a dragon year again) And hopefully loads of profit, and rest, for ED ;)

    Edit: Agree that ED needs rest
     
    Last edited: Dec 5, 2017
    mecksg1 likes this.
  9. Kev2442

    Kev2442 Still French

    Joined:
    Nov 4, 2014
    Messages:
    453
    Location:
    42 (Loire) is the answer.
    Don't ever say that, you fool ! If given the choice, he'll pick profit over rest and die of exhaustion ! :confused:
     
  10. fusion_power

    fusion_power Advanced Member

    Joined:
    Dec 11, 2006
    Messages:
    6,858
    Location:
    germany
    Lol, the CPU heatsink solution, ED showed recently on Youtube is basicly the same idea that we had years ago... xD Interesting that it has now urgency to be built in. Guiding the heat through the PCB alone was not enough at the end. But nice to see that if works with the heatsinks (so much space in that brick-case).
    I would suggest some additional holes in the case where the copper heatsink sits, just to prevent heat accumulating there.

    Future SoC options with less TDP should be even more happy with such a generous cooling solution. Everything is fine as long as it prevents the so common CPU throtteling in modern mobile devices.
     
  11. EvilDragon

    EvilDragon Administrator Staff Member

    Joined:
    Mar 4, 2003
    Messages:
    19,064
    Location:
    Ingolstadt
    Urgency...? That solution has been tried in August at the Gamescom, and planned for a lot longer (I had to source parts, etc.) :)

    Now we're just improving an already working solition :)
     
    elwing and ClockworkCoder like this.
  12. Grench

    Grench Forum Addict!

    Joined:
    Oct 3, 2008
    Messages:
    5,219
    Let me start with this: Having a CPU that can throttle is a GOOD thing.

    Consider two systems, A&B. Both have the SAME thermal envelope. I.e. they can dissipate the same energy at the same rate.

    System A has a CPU that when running at it's maximum capability produces exactly as much heat as the system can deal with. I.e. they're a perfect match and the CPU can run at 100% all the time. That is the model that a lot of desktop computer cooling systems are built on.

    System B has a CPU that when running at it's maximum capability produces 500% as much heat as the system can deal with. I.e. if it ran continuously at full tilt it would either melt down or burst into flames. This is the model that a lot of handheld cooling systems are built on.

    If your preference is for consistent performance over extended time periods, you would have a preference for System A -OR- for System B with systematic underclocking applied - to make it run the same consistent speed as System A.

    If your preference is for the spot/moment/current task to get done as quickly as possible, but the before & after moments have nearly nothing going on (majority of communications & office type tasks), then your preference is system B.

    Note that System B has the ability to cover BOTH use cases. If the CPU is fast enough that it needs to throttle to stay within the device thermal envelope, it means that the CPU has the capability to run quicker for moments. It can also be clocked down to run consistently within the device's envelope.

    Within a handheld environment, there is always a thermal limit. I.e. how quickly the device can dissipate the heat being generated. The heatsink that ED is adding assists with giving the heat a place to go to during peak loads for it to dissipate over time. It will also make that spot on the lower-back of the Pyra warmer and that should assist some with dissipation. But, nothing is going to give enough dissipation for the OMAP or any likely future SoC to run balls out continuously forever in a pocket-sized passive thermal envelope. They will all have to have some form of throttling applied - either by the SoC when things get too hot or by the user purposely underclocking for the session.
     
    Djhg2000, pocak, bennr01 and 4 others like this.
  13. Eight Bit

    Eight Bit Advanced Member

    Joined:
    Nov 16, 2008
    Messages:
    1,262
    Location:
    Amsterdam, Netherlands
    @EvilDragon As there is no airflow in the case, I was wondering if the copper mass/heat buffer and it's surface are big enough to absorb and radiate all the excessive heat or does it over saturate after a while?
    Would having a bunch of small holes in the case, or some jail bar construction running over the back and bottom at that heatsink spot be viable? How hot does it get? Does the case get very hot there too?
     
    rygD likes this.
  14. PCXT

    PCXT Active Member

    Joined:
    Sep 14, 2016
    Messages:
    146
    It would be interesting to compare solid block vs the block used here, with cuts. These cuts are for better removing heat when there is an air flow, and without them, inside the casing, the "heat buffer" would be larger and dissipation may be a bit better (contact area with casing will be larger).
     
  15. Eight Bit

    Eight Bit Advanced Member

    Joined:
    Nov 16, 2008
    Messages:
    1,262
    Location:
    Amsterdam, Netherlands
    It would be sweet to have an adapter in the 2nd space, just some copper joint or something that could connect to a docking station with a bigger more capable cooler, or a copper cable running through some mittens for warm hands when on the go :D
     
  16. Zwerg01

    Zwerg01 Still Fresh

    Joined:
    Feb 15, 2017
    Messages:
    34
    There is no convection, the radiation at this temperature is minimal as described in stefan-boltzmann law and the surounding plastic has a high thermal resistance. So, this copper block is not an heatsink. Its acts like a thermal capacitor. Once it is charged, it has no cooling effect.
     
  17. Linux-SWAT

    Linux-SWAT Advanced Member

    Joined:
    Feb 13, 2010
    Messages:
    7,950
    Yes it's likely. The copper block should be exposed to the outside.
     
    rygD likes this.
  18. WizardStan

    WizardStan Mega GP Mania

    Joined:
    May 24, 2008
    Messages:
    16,355
    That doesn't sound right. Actually sounds very wrong. Even if the thermal resistance is high it's not infinite: some heat will necessarily bleed through it, the case will become warm and that heat will radiate away from the device. Once the heatsink is "charged" its cooling effect will be limited by the surface area of the plastic: it may be low but it won't be zero. If you want to argue that it's lower than the output of the CPU then I don't know enough to argue against that, but you can't say it has *no* cooling effect, especially since the primary heat exhaust is open air via the USB, not enclosed in plastic. I can envision a USB powered fan which cools itself off if absolutely necessary for extended play.
     
    Djoga'Ro and Silent-Hunter like this.
  19. EvilDragon

    EvilDragon Administrator Staff Member

    Joined:
    Mar 4, 2003
    Messages:
    19,064
    Location:
    Ingolstadt
    Well, at GamesCom, we ran the system for about 30 minutes with 100% CPU load at Dual-Core 1.0GHz, and it never overheated. It reached maximum temperature after about 10 minutes.
    Sure, it's not as effective as with some airflow like with a fan, but it seems to exchange heat enough to stay at a sane temperature.
     
    Djhg2000, remjey, Alberto and 4 others like this.
  20. Silent-Hunter

    Silent-Hunter Advanced Member

    Joined:
    May 29, 2010
    Messages:
    2,474
    That's really good to hear, ED!
     

Share This Page

Loading...