Some personal stuff - and prototypes building!

EvilDragon

Administrator
Staff member
Joined
Mar 4, 2003
Messages
29,277
Age
42
Location
Ingolstadt
Somewhere buried in the last news thread I've already made a small status update post, but here's a proper, more thorough information and a bunch of pictures, as I'll document the prototype building process very thoroughly.

Well, the current delay in prototypes building can be blamed on me - not any company or anyone else.
I've had a few rough weeks. First a colleague with whom I've worked for 20 years now has suddenly died due to a heart attack - then I catched a cold (maybe even the flu)... and while I'm usually recovering within 2 - 3 days, this time, it took more than 2 weeks until I fully recovered.
I tried to work on the prototypes - but literally fell asleep in front of them, which wasn't helpful :)

Maybe it was also a small burnout, who knows. But I guess it was all just a bit too much. A lot of work, death of a colleague which also leads to some demotivation, flu season... all that added up and took its toll.

Anyways, I gave myself some time to recover, as it won't help ANYONE if I continue working in Zombie mode, and it seems like I've fully recovered now, as I'm feeling fine and FINALLY started to work on the prototypes this weekend!

I hope you can understand that

Now, onto the important stuff: The news!

1. The cases

While the new company has already done some tweaks and design changes, they didn't have too much time as they had a huge project by one of their biggest customers which had to be done within a timeframe. They already told me that when I was visiting.
Well, the deadline for that ended last Friday - which means that they can now fully work on the Pyra case.
I can't wait to see the fixes they'll do :)


2. Small change for the keymats

The manufacturer of the keymats ran into an unexpected small issue while doing the mass production run: It was nigh impossible to prevent visible air bubbles inside the glue when attaching the keycaps.

That's why they're changing the design of the plastic caps so that they will stay affixed on top of the mat without glue.
Everything else stays the same. As soon as I know more, I'll let you know.


3. Prototype building

Finally, long overdue, the prototypes are being assembled by me.
And while I won't do a video (as such a setup is too complex and just hinders the assembly), I'll do a LOT of pictures for you to enjoy, basically a step-by-step guide on how the units are assembled!

I will post these pics here in this thread and will try to update it daily, as I continue to assemble the prototypes.

First some pictures which don't have anything to do directly with the assembling of the prototypes:
One resistor had a wrong value and needed to be replaced. Does anyone remember how MWeston had to replace over 400 resistors to fix Wifi on the first-built Pandoras...?

Yeah, we had something similar - though luckily, we noticed it before the mass production. This wrong resistor value limited the power input of the USB port to 1.5A - which caused a slower charging.

01 - SDCompare.jpg
The resistor is REALLY small. The second one from the left in front of the small chip needs to be replaced. And yes, that's a Micro-SD Card to show the size :)
Fun fact: The third resistor from the left is actually 90° rotated. That's from the pick and place machine :) It sometimes happens and is the case on multiple PCBs. That's not a problem though, just a fun fact.

02 - Desoldered.jpg 03 - Desoldered-2.jpg

Here are two pictures of the desoldered resistor.

04b - Resoldered.jpg 04 - IndexFinger.jpg

And here's the result: The new capacitor is in there. And my index finger once again shows how small those resistors are.

05 - BunchOfBoards.jpg 06 - BunchOfBoards-2.jpg

Well, these boards have all been modified - so let's start to turn these little ones into prototypes this week, right?

Now, let's start with the real assembly!

The Lid

The first thing that needs to be assembled is the lid.

07 - Bumper.jpg

Let's start with a sheet of bumpers and a bunch of LCD PCBs.

08 - BumperAttached.jpg 09 - BumperAtt-Closeup.jpg

These rubber bumpers are being sticked on the PCBs.
There are two reasons we're doing this with bumpers and not directly inside the plastic:

1. You would see marks on the outer side of the lid if this would be done inside the plastics, which we want to prevent.
2. When you push too hard with the stylus onto the touchscreen, the soft rubber will cushion that a bit and prevent damage.

BTW: The mass production LCD PCBs will look a bit simpler. The solder pads for the once-planned rotator chip will be gone as well as the ones for the other display we looked into. The 0R resistors in the middle of the rotator chip area will also be gone.

Next, we'll need the LCDs.

10 - LCDs.jpg
They come on trays - these are actually LCDs from the mass production batch, no prototypes anymore.

Now we'll connect the LCDs to the LCD PCBs:

11 - DisplayAttached.jpg 12 - DisplayAttached-Closeup.jpg

Simply plugging in the small connector...

13 - DisplayAttached-2.jpg 14 - DisplayAttached-3.jpg

... and simply wrapping the flex cable around the PCB, so that the LCD is in front of the LCD PCB.

15 - Touchscreen-Attached.jpg 16 - DisplaysFullyAttached.jpg

Connecting the touchscreen - and then we're done :)

17 - Magnets.jpg

Now, here's a bunch of magnets. I've got two different sizes, as I need to test which ones are strong enough to reliably trigger the lid switch. These are some neat NeoDym-Magnets.

18 - Magnet-1-Attached.jpg 19 - Magnet-2-Attached.jpg

Here they are, glued to the case.

20 - MagnetsInCase.jpg
And a bunch of transparent cases with magnets - hard to see, due to the transparency, eh?

21 - FoilRemove1.jpg 22 - FoilRemove2.jpg 23 - FoilRemove3.jpg

Next step: Putting the pre-assembled LCDs with the PCB into the case. We need to lift the protective foil a bit and put it in front of the case - otherwise, you could have a hard time removing it once the unit is assembled (as the foil would be below the bezel...)

24 - LCDCable-Closeup.jpg 25 - LCDCable.jpg

Now it's time to connect the LCD cable.

26 - DisplaysWithLCDCables.jpg

And here we are - 10 almost-finished lids!
Now we only need to roll the cable and put the backside of the lid on.

I couldn't do this today, as I noticed that I don't have the metal logo plates here (they're in the office, and I*m at home). While these will be glued in by the factory for the mass production, I need to do this myself for the prototypes.

I'll continue the assembly and picture posting tomorrow - so stay tuned for updates here at this thread :)

And sorry again for the delay - I really needed a bit of time to recover from all the stuff that happened. I'm only a dragon like everyone else :D

UPDATE
2019-03-12:

As promised, here is the continuation of our assembly journey :D

As mentioned, this next step won't be necessary for the mass production (as the metal pieces will be sticky on one side, so no glue is needed), but for the prototype production, it is necessary.

26-Lid and Logo.jpg

Here is the lid and the metal part of the logo.

27-Glued Lid.jpg

Let's put in some glue.

28-Logo in Lid.jpg

And push the logo firmly on top of that. First step done!

29-Glue on Metal.jpg

Then we need to put some glue on the metal logo.

30-Logo glued in.jpg

Put in the logo (which will be milky-white transparent in the final unit).

31-FinishedLogos.jpg32-FinishedLogos2.jpg

And here they are! The lids and the logo!

As the glue needs to dry first, I couldn't do anymore right now. But the journey will continue!

Aaaaand on we go.


UPDATE
2019-03-20:

I was a bit stupid - as you can see, I connected the LCD cables first and put the screen and PCB into the lid frame.
That's how it worked for the Pandora - but not with the Pyra.

With the Pyra, assembly works better like this:

27-Rolled Cable.jpg

First we roll the cable.

28-SlideThrough.jpg

Then we slide it through a tiny slit which is in the case - it's invisible unless you know it's there.

29-InsideLid.jpg

That's how it looks like once it's through.

30-ConnectedDisplayboard.jpg

THEN we connect it to the display board!

31-LidBackside.jpg

And put it inside the backside of the lid.

32-Applying Frontcover.jpg

And after that, the front cover is being put on.

33-AssembledLid.jpg

Done! It clicks in on the right and top...

34-Screws.jpg

... and needs screws on the bottom and left.

35-HingeOutside.jpg

Next, we put the hinge into the keyboard-case-part.
This is where it needs to be pushed all the way in...

36-HingeInside.jpg

... which looks like that.

37-SlidingIntoBase.jpg38-SlidThrough.jpg

Then we slide the LCD Cable into another slit which is in the keybard part.
39-PreAssembled.jpg40-CloseUp Roll.jpg

And here's the finished part!

41-Locking Hinge.jpg 42-Locked Hinge.jpg

Now the hinge is being pushed back from the inside into the lid.

43-FullyAssembledTop.jpg

Done! Here is the fully assembled lid with the basement.

44-MultipleAssembledTops-1.jpg 45-MultipleAssembledTops-2.jpg
46-MultipleAssembledTops-3.jpg 47-MultipleAssembledTops-4.jpg

And, of course, I needed to do more than just one :)

That's it for today... hopefully, I can continue soon. 47 Pictures already, whew :D

UPDATE
2019-03-26:

Found some time to move forward :)

48 - Rubber.jpg

Now, we need these rubber parts. They are used for sealing the speakers.

49 - Rubber in Speaker.jpg

First some rubber for padding...

50 - Speaker on Top.jpg

Then the speaker on top.

51 - Putting in the Keymat.jpg

Now we simply put in the keymat.

52 - Nubs.jpg
And the discs for the Nubs!

54 - Second Speaker.jpg 53 - Speaker sealing.jpg

Then both speakers will get the sealing put on top.
Please note: This will probably change for the mass production. It doesn't really make sense having extra rubber parts for sealing the speakers when we already have a silicon keymat.

The keymat will most probably change so that they will have this sealing included - but that's currently being discussed. I will let you know more about it soon.

55 - Put in the PCB.jpg

Now it's time to put in the PCB!
When it's inserted, the nub discs need to be rotated until they fall onto the nubs. Works like a charm :D

56 - PCB in the case.jpg

Here's the PCB inside the case, on top of the keymat and speaker sealing :)
57 - LCD Cable 1.jpg 58 - LCD Cable 2.jpg
Next: Clip in the LCD Cable :)

59 - Almost finished.jpg

From the front, the unit looks like it's finished already :)

Next steps (hopefully coming soon): Put in the heatsink and heatpipe together with the CPU-Board - and close everything!

I realized I need to order some more heatsinks to build the prototypes, but that's not a big deal, as I can put them in after I ran the tester software on them.

See you next time!

UPDATE
2019-04-15:


Okay, since I received the remaining stuff I was still in need of, I can finish with the prototype assembly.
So here's the next set of pictures :D

60 - Multiple Pyras.jpg

So, here we are. Eight prepared prototypes. Why eight? Well, I had two of the ten lying at work for some tests (so I couldn't include them), and I'm saving the remaining PCBs to use in the revised cases :)

61 - HeatpipesSimple.jpg

Now it's time to move the heat away from the CPU.
These are standard heatpipes that can be easily bought in exact that size.
Remember the custom cutout I posted somewhere? Well, this is the same material, but to save some time (and to check whether our custom cutout and the readily available stripes make a big difference), I used these this times.

62 - Exposing the Copper.jpg

Below the tape, they expose copper (and some self-adhesive glue).

63 - Putting it on the OMAP5.jpg

We simply need to put these on the OMAP...

64 - Applied.jpg

... which looks like this.

65 - CPU Board put in.jpg

Then put the CPU board into the mainboard. As you can see, the stripe is ALMOST fitting perfectly without any modification. Our custom cutout part will - of course - fit better, but who knows if we need it :)

66 - Heatsinks.jpg

Next, we got the heatsinks. The copper one is bought as is, whereas the aluminium ones have been bought as stick and manually cut by me (yeah, I know I did a bad job with that one :D)

Buying them as sticks and cutting them with a machine is simple and accurate, so if that works better than the available ones, I'm happy to do that. But that's what we need to test :)

67 - Heatsink applied.jpg

So here is the heatsink on top of the copper strip coming from the OMAP.
It sits directly above the USB port.
As said, the standard strip doesn't fit perfectly, as you can see here :)

68 - Thermal Insulators.jpg

Next, we have something I'd like to try out: Thermal Insulators.
This thin self-adhesive material is said to isolate heat - in theory, you should be able to put this on a 100°C hot stove plate and easily touch it with your fingers!

As we need to keep the CPU heat away from the battery, this might be a solution!

71 - Alternative Metal Plate.jpg

So far, as an alternative, we used lasercut 0.4mm thick metal plates - but of course, these weigh more than the other ones and they heat up over time, so maybe that thermal insulator is working better? I'll find that out :D

69 - Applied 1.jpg 70 - Applied 2.jpg

This is how the case looks when the material is applied.

72 - Stylus Holder.jpg73 - Plugged in.jpg

Next step is top put the stylus holder into the case. It's simply plugged in. Done.

74 - Stylus fits.jpg

Then the stylus snaps in nicely :)

75 - Shoulder Buttons.jpg

Time for the shoulder buttons. As you know, these are currently being worked on, so this part might change a bit.

76 - Lower Shoulder Button.jpg 77 - Upper Shoulder Button.jpg
Then we put in both shoulder buttons on both sides.

78 - Putting it together.jpg 79 - Still unscrewed.jpg
And it's time to assemble both sides!

80 - Screwing together.jpg

Screw it together...

81 - Heatsink inside the case.jpg

Done! Here you can see the heatsink sitting on top of the USB port.
We can also add a second one on the second port (using our custom-cut material), but according to our simulations, a second heatsink won't help much as there's only so much heat that can move through the heat pipe.
But I'll test that, of course :)

82 - Full Unit.jpg

And here's the unit fully assembled :)
What a journey :)

Now I need to assemble the remaining ones as well, and then it's time to do the test-run!

UPDATE:

Almost forgot! I assembled more of them!

Here's the picture:

DSC03620.JPG

That's about half the prototypes I have, so I can assemble at least 9 more.
Time to continue work on the tester SD Card and test these little babies :)
 
Last edited:

dfkman

One day I'll be writing code in the car...
Joined
Nov 14, 2017
Messages
15
Don't feel bad ED. Just keep up the good work!
 

levi

Still fresh, damnit!
Joined
Oct 6, 2008
Messages
11,527
Location
Somewhere off the coast of the EU
Yeah, there some bad bugs doing the rounds these past few weeks. Good to hear your over it and feeling better. My best wishes to family and friends of the deceased also.

Those are some cute neodymium magnets! Looking forward now to the blow by blow updates!
 

docbroke

Banned
Joined
Feb 21, 2019
Messages
251
Age
38
Location
India
I am so glad that you are feeling alright. Please keep posting, it is great to have some update.

So as far as I understand, once cases, keymats and some cooling parts are finalized Pyra will be ready for mass production ?
 

Djoga'Ro

moonstruck
Joined
Apr 3, 2016
Messages
996
Thanks for the update, ED. It's good to hear you're better.

The manufacturer of the keymats ran into an unexpected small issue while doing the mass production run: It was nigh impossible to prevent visible air bubbles inside the glue when attaching the keycaps.
Ain't vacuum chambers solving issues like that?
 

ingoreis

Advanced Member
Staff member
Joined
Jan 18, 2010
Messages
8,089
Age
38
Location
49.491276,8.423518
I had this Flu too..it took this Year up to 3 Weeks to Recover.
Normaly i do have a good Immune System but this Year it was realy resistent.

Good to see the Progress and good to see that you do not forgett your Healthness.
 

ClockworkCoder

Chaotic Neutral
Joined
Jan 21, 2016
Messages
1,217
Location
Menzoberranzan
Thanks for the update. Looking forward to seeing progress of the prototypes and hopefully speedy progress from now on ;)

... and simply wrapping the flex cable around the PCB, so that the LCD is in front of the LCD PCB
I presume this isn't likely to be caught tightly between the PCB and the case enough to cause any problem?

Everything else looks good. I don't envy you having to replace those resistors. Almost need a microscope to see them!
 

matzesu

Hardcore Member
Staff member
Joined
Oct 24, 2008
Messages
9,221
Age
35
Location
Germany,, Saarland, at home
It isnt frapped Multiple Times like a Headphone Cable around a Walkman, , if i remember from previous Prototyphe Assemble Videos, its just around the Corner, and will not harm during use, because its stays in place and dont move,
Hopefully the Keyboard Change dosnt mean another 8 Weeks leadtime..

Glad your feeling better now, "flue is a bitch"

Your Health is a bit more important than the Pyra, and these few weeks later ditnt make much diverent..
 

ClockworkCoder

Chaotic Neutral
Joined
Jan 21, 2016
Messages
1,217
Location
Menzoberranzan
if i remember from previous Prototyphe Assemble Videos, its just around the Corner, and will not harm during use, because its stays in place and dont move
I assume this to be correct. The it's just a detail I don't recall seeing before; most of the focus was on the other ribbon cable for which I really like the solution and attention to detail.
 

Bernd

Very Active Member
Joined
May 4, 2016
Messages
189
2. Small change for the keymats

The manufacturer of the keymats ran into an unexpected small issue while doing the mass production run: It was nigh impossible to prevent visible air bubbles inside the glue when attaching the keycaps.

That's why they're changing the design of the plastic caps so that they will stay affixed on top of the mat without glue.
Everything else stays the same. As soon as I know more, I'll let you know.
Does that mean its possible to pop of the caps and use small inlays to change the keyboard-layout?
(Sorry for the resurfacing of the keyboard discussion)

Take your time and stay healthy.

[EDIT]
If the resistor get changed to increase the charging current to >1,5A, is it possible to reduce the charging-current to reduce the battery wear with software?
[/EDIT]
 
Last edited:

levi

Still fresh, damnit!
Joined
Oct 6, 2008
Messages
11,527
Location
Somewhere off the coast of the EU
I assume this to be correct. The it's just a detail I don't recall seeing before; most of the focus was on the other ribbon cable for which I really like the solution and attention to detail.
It can't be crushed front to back because it doesn't go in front of the screen on the one side, and on the reverse, those bumpers stop it being crushed against the lid top. I expect there's something keeping the LCD in place probably on the screen frame insert (so that the screen stays in place), so as matzeus says, it shouldn't go left-right very much at all either.
 

ClockworkCoder

Chaotic Neutral
Joined
Jan 21, 2016
Messages
1,217
Location
Menzoberranzan
Is it just me, or does it seem a little unfair / unrepresentative to still call them prototypes at this point? More like alpha-release or release candidates if anything.
 
Top