Case Improvements and the keymat


EvilDragon

Administrator
Staff member
Joined
Mar 4, 2003
Messages
29,986
Age
47
Location
Ingolstadt
Some quick update from me, this time with a new thread (as I don't want to hide that in the prototype assembly thread).

1. Prototype Assembly

I have ordered more material for cutting the heatpipes as well as more heatsinks but am still waiting for them.
They should hopefully arrive next week and then the assembly should be finished pretty quickly.


2. Other stuff I'm working on

In the meantime, I have started to look for companies that can produce custom heatsinks. While the standard one we used so far fits well, we still have 1.5mm more space for the width, so a custom made could improve things a bit.
If the price for doing a custom one is reasonable, then I'll do that. If not, I can always buy the standard ones :)
If you know of a company who does stuff like that (would be best here in Europe) and you know they are doing a good job and have reasonable prices, let me know :)

We're also looking inf different ways to properly apply the logo and metal plate. Glueing them manually is tedious. It would be best if the logos had some self-adhesive 3M glue on them, as it holds pretty well and it would only take seconds to apply them. The problem is: Right now they're laser-cut - but that would melt the glue :)
So we might need to punch them out, this would probably be the best way :)


3. Status of the case improvements

I am happy to report that the main fixes on the cases have been done!
Mostly production related fixes, except for adding springs to the shoulder buttons, which has already been done in the mould:

Mould1.jpg

A set of cases will be produced today and shipped to me.

There are still some smaller things I've noted down during assembly:
Even if the case itself work fine, there are some things you notice when you're assembling multiple units at once. Small stuff that slows you down (for example, putting ont he speaker sealing correctly, sliding the LCD cable through the hinge, etc.) and would cost more in production.

I've checked these and tried to find better solutions, and often it's just a tiny little thing that needs to be changed to fix that. I'm recording all this properly on video, so the guys working on the case can see what the issue is. I also present my idea how to fix it, but they also think about it and might find some better ideas (they have experience with stuff like that).
Recording that properly and showing exactly what the problem is takes a bit of time, so it's a slow process.
But it will REALLY speed up things when it comes to mass assembly!

Once I receive the case and the team sees my videos, I'll do a video conference with them to discuss any final tweaks that still remains. If all major issues have been fixed, the case is basically a finished Rev 1.0 which is fully usable :)

It's really refreshing to see how things are now moving again, after the case hasn't really been worked on for the last 1.5 years... :)


4. The keymat


As previously mentioned, glueing the keys didnt't work without bubbles.
I'm happy to report that the new keycaps which don't need to be glued have been finished :)

They look like this:

KeyCap2.jpgKeyCap1.jpg

As you can see, they now have small pegs inside so they will tightly sit on the rubber mat.
Additionally, they have some frame around them, so they will sit properly inside the case. They can't come loose and won't fall out, but in theory, it's possible for you to remove the caps and paint onto your keys to change the layout (though that changed key won't be backlit of course).

So, done and fixed.


The main slowness right now is probably still myself, for various reasons.
I am still doing things at a bit slower pace, as I really overdid it in December / January, which lead to a lack of motivation - and that's the least thing we need right now.
Secondly, the Pyra cost A LOT MORE MONEY than planned, so money is tight and I need to make sure the shop pulls in enough money to keep the company going. So I cannot simply abandon the shop for a week to fulltime work on the Pyra. Would be great, but it wouldn't work out.

And thirdly, my main profession is not designing and developing such products. When I need something like a custom heatsink, I first need to go and search for a proper companies. WIth some stuff, I don't even know the correct search term to find a company for that (example: I am thinking of maybe use some pre-cut self-adhesive tape for sealing the speakers, you know, the 0.2 - 0.3mm thin one that you simply put on and that's it. Probably the fastest and most secure way when mass assembling the units. A thin self-adhesive sticker, punched based on our own design. Something that can be seen in tons of devices, so pretty easy and common to get, right? Well... try to find a company for that. I still don't know what's the best search term to use to find one...).
It's tiny stuff like that - things you never imagined would steal your time - that slows you down.
But yeah, if you know of such a company... let me know :D I found one in Germany so far, but I'm happy for more options :)

Well, and hockey season is over for our team, so I have a bit more time available from now on.
But yeah, blame it on me now :D

5. Current Roadmap

As a summary (or if you didn't want to read through the whole text), here's my current roadmap (with no time given, as usual).

  • I'll cut the heatsink material as soon as it arrives and finish assembling the prototypes.
  • I'll test the prototypes and finally ship them out
  • Case Company finishes the first set of cases today and will send it to me (with all major fixes done and an improved shoulder button which uses a spring)
  • I'll discuss final possible non-mandatory tweaks with the case company.
  • If functionality of the prototypes is confirmed, mass production of PCBs is beind ordered (this will take around 6 - 8 weeks)
  • Heatsink material and maybe proper heatsinks are being produced
  • Cases are being produced
  • Keymats arrive
  • Mass production happens

As usual, any questions, remarks, blaming, shouting and screaming, insulting, etc.: Post here at the thread :D
 
great to see the ongoing steady progress, how you do it all and do how many jobs at the same time amazes me!
 
I'll blame it on the boogie
static.gif


But this is refreshing to see from a company, not putting the blame on others. But be sure to keep your (mental) health in mind, I for one think you were and are doing an awesome job and even doing it better than some industry standards in many respects. Especially when it comes to communication. So cheers to you, the other plastics company and to the pyra
static.gif
 
"So little time. So mutch to do."

I think it is amazing how everything comes together. Rearly good work.

I can't imagin how twitchy the pre-pre-orderes must be right now:D
 
So I guess about 4 months may pass before start of mass productions, and around 6 month before I get my pyra.

Next 1/2 week: arrival of heat sink and assembly of pyra , new case might also arrive during the same time
3/4 week: prototypes received by devs, pre-preorderers
at about 6/7 weeks prototypes tested and if all is well PCB goes in mass production
in about 4 months: PCB produced, during same time, cases and keymats and heatsink shall be ready too
2 more months for mass production and shipping of about 1100 pyras.
 
I know, but I hope it is better this time.
And I said 2 more months after mass production starts ( total about 6 months from today ) with bold if in between.
 
"Slow and steady wins the race" as the proverb goes. And also yields results one can be happier with at the finish line.

One question @EvilDragon: Is the test suite for the prototypes already written and ready to run, or is that still on the To Do list?

On producing custom heat sinks there is: "EEVblog #1196 - Custom Heat Sink Design" [1]. Only about 4% relevant, but there might be something useful in there. (I find this fellow a little annoying -- far too much energy! -- but he certainly knows some stuff.)
Code:
[1] https://www.youtube.com/watch?v=RTmzVs6LHRo

A web search on "laser cut self-adhesive" returns results that suggest that aluminium backed with self-adhesive tape _can_ be laser cut. It also turns up firms that laser cut adhesive tape which I think is maybe what you were looking for for sealing the speakers.
 
Last edited:
So I guess about 4 months may pass before start of mass productions, and around 6 month before I get my pyra.

Next 1/2 week: arrival of heat sink and assembly of pyra , new case might also arrive during the same time
3/4 week: prototypes received by devs, pre-preorderers
at about 6/7 weeks prototypes tested and if all is well PCB goes in mass production
in about 4 months: PCB produced, during same time, cases and keymats and heatsink shall be ready too
2 more months for mass production and shipping of about 1100 pyras.
You forgot to factor in the volcano eruption
 
Another sterling update, thanks. It's really starting to feel like all the outstanding items are starting to gel together into a complete and tangible product. Things are indeed starting to become a lot more exciting again :)
 
Last edited:
Another sterling update, thanks. It's really starting to feel like all the outstanding items are starting to gel together into a complete and tangible product. Things are indeed starting to become a lot more exciting again :)

Yes. I wish I would've been able to do everything faster, but even small steps are steps - and we're moving in the right direction.
 
Great update once more! Very happy to read it!!! One minor point you intended to extend the keymat with speaker sealing. Is that still a plan? My feeling is that a custom heatsink might be another obstacle.

All in all it seems that the things are moving along nicely. Side note I've made another bet with a friend that Pyra will be done this year (after losing one couple of years ago or so). ;-)

Last and not least, please take good care of yourself. Pyra will be no doubt an amazing device but health is much more important.
 
While the standard one we used so far fits well, we still have 1.5mm more space for the width, so a custom made could improve things a bit.
I can't imagine that 1.5mm are worth a custom heasink solution and all the additional time this needs.
If you know of a company who does stuff like that (would be best here in Europe) and you know they are doing a good job and have reasonable prices, let me know
No clue about Europe but like usual, if you don't find what you need in Shenzen, China, it just doesn't exist. ;)
We're also looking inf different ways to properly apply the logo and metal plate.
The entire logo construction seems heavily over-complicated and expensive. And it does not even look that great imho. I whould just had the logo molded into the lid, done. If you really want to light it up (why we wanted that actualy?), just use less case material on that logo spot so it is thinner than the surrounding material and let some light through. Just an Idea.
I'm happy to report that the new keycaps which don't need to be glued have been finished
I like that solution. Really clever idea and it makes the keyboard sorta modular. And everything is better when it works without any glue. :)
the Pyra cost A LOT MORE MONEY than planned,
The SoC alone is so stone-old now, you should get it fro free actualy. :D
 
I can't imagine that 1.5mm are worth a custom heasink solution and all the additional time this needs.

Yeah. And even if they are, since the development costs seem to be a problem by now, you could sell it as an upgrade later …

The entire logo construction seems heavily over-complicated and expensive. And it does not even look that great imho. I whould just had the logo molded into the lid, done. If you really want to light it up (why we wanted that actualy?), just use less case material on that logo spot so it is thinner than the surrounding material and let some light through. Just an Idea.

It's a bit late in the development cycle to make those kinds of changes, I fear.

The SoC alone is so stone-old now, you should get it fro free actualy. :D

Good thing the Pyra design is modular so that a SoC upgrade could be available half or one year after release.
 
The entire logo construction seems heavily over-complicated and expensive. And it does not even look that great imho. I whould just had the logo molded into the lid, done. If you really want to light it up (why we wanted that actualy?), just use less case material on that logo spot so it is thinner than the surrounding material and let some light through. Just an Idea.

The LEDs in the lid are somewhat critical for being able to show status & alerts by type with the lid closed. By flashing in different colors for different lengths of time we can have the Pyra differentiate between calendar, call, new email, messaging, etc. That ability is pretty core to using the Pyra as a communications device.

I suggested this idea of thinning out the spots where the LEDs shine through back when the plate/plastic/glue was put forward. Trying to do that via case material thickness could be a very tricky business - it gets down to how many thousandths of a mm thick can it be to let light through while still being thick enough not to just 'punch out' the moment any pressure is applied to it. There is no guarantee of even finding that 'happy medium' - it may not exist.

The shiny plate & translucent backing plastic & glue solution is a bit of a vanity/aesthetic thing, but it takes the place of what would otherwise be light pipes that have proven to be cumbersome in the Pandora. That said, these need to be self-adhesive with a pretty serious glue to bond between the aluminum and the plastics - and the adhesive likely needs to be applied to the aluminum prior to oxidation of the aluminum surface to be able to hold well. Aluminum oxidizes very quickly.

@EvilDragon Regarding cutting the aluminum with an adhesive layer pre-applied. Laser cutting, stamped - but there IS a 3rd option that I am aware of. Water jet cutting. No significant heating and it can be nearly as precise as laser cutting. These machines are CNC based and can be found or contracted through most medium to large size machine shops.

@EvilDragon Regarding specific custom sized heat sinks. I'm assuming you are pursuing copper or aluminum for these. The same water jet cutting facilities should be able to cut these out of copper or aluminum sheets. The process works for stock from 0.02mm to 300+mm thickness. The catch is that these cut in 2D, so one of your dimensions needs to match the chosen stock dimension and the result will have squared edges to that dimension. So, maybe one visit with the right machine shop could solve both of these issues through a single supplier?
 
Back
Top