Type of case for futur docking station?

[mclien]

I did a little talk with ED at "lange Nacht der Computerspiele" about the future docking station.
As everybody learned so far, injection molds are quite expensive, so the question is: Do we need a injectin molded case for a docking station?
My point of view is, we don't. I 'd say the ideal version for the dock would be a modular 2 PCB dock, where 1 is the one for the "minimal dock" (like port replicator) and by adding the second PCB will expand to an "desktop replacement" with additioinal USB hubs build in, converters to wired ethernet, connectors for HDD/SSD and optical drive, maybe even a HDMI splitter which gives you two 720p Displays instead of only one fullDH and so on.

For the case I would simply design a version that is 3D printable and one that assembles from laser cut parts, so it could be done in 95% of the hackerspaces near you or is easily don from someone of the community.

 

[Grench]

I did a little talk with ED at "lange Nacht der Computerspiele" about the future docking station.
As everybody learned so far, injection molds are quite expensive, so the question is: Do we need a injectin molded case for a docking station?
My point of view is, we don't. I 'd say the ideal version for the dock would be a modular 2 PCB dock, where 1 is the one for the "minimal dock" (like port replicator) and by adding the second PCB will expand to an "desktop replacement" with additioinal USB hubs build in, converters to wired ethernet, connectors for HDD/SSD and optical drive, maybe even a HDMI splitter which gives you two 720p Displays instead of only one fullDH and so on.

For the case I would simply design a version that is 3D printable and one that assembles from laser cut parts, so it could be done in 95% of the hackerspaces near you or is easily don from someone of the community.

I would suggest simplifying that a step or 5 more.

Find a standard project box. Something like this only cheaper?:

Rip/cut 1/3 or whatever works right off of the front and rear panels.

Build ONE board with port framing to match what was ripped/cut from the front/rear panels. Yes, this means ONE version of the dock/replicator, not 2.

It should connect to all of the Pyra's back ports (which may be a challenge to orient).

It should have:
4 port USB 3.0 hub to produce 4 fully powered USB 3.0 type A host ports (No OTG or 'device' mode when docked).
2 USB 2.0 ports (straight through).
1 SATA port (broken out and built-in adapter)
1 full size/normal HDMI port
1 power connector capable of supporting the Pyra at 100% usage, charge the Pyra battery at max rate, supply full power to all 6 USB ports
1 head phone out port
1 microphone in port

Things like wired Ethernet, an HDMI splitter, additional USB hubs for 'desktop' usage can live as additional adapters plugged in at the desktop at the user's discretion based on what they need.

I wish it were easier to add stereo mic inputs and line level audio outputs - but that would require a USB sound card to accomplish.

 

[ible]

for the longest time i haven't been able to figure out what images you've been attaching to your posts, @Grench.

then I realized my adblocker was blocking them. are you an ad?

 

[Dimacus]

This is something I very much intend to build myself, but if it can be bought from ED's shop in kit form or whatnot, I would much prefer to buy parts like these from there, if nothing else to support the shop.

for the longest time i haven't been able to figure out what images you've been attaching to your posts, @Grench.

then I realized my adblocker was blocking them. are you an ad?

Thank you! I've been wondering the same thing (on these forums and others).

 

[___]

I'd take the best from both worlds. a lasercut case kit, but only one pcb like grench suggested. maybe use the dimensions of an easily available case so people can decide to buy that and do minor modifications instead of lasercut plastics/cnc cut wood or a kit.

 

[Agricola]

Why do people suggest only one PCB? The one connecting the Pyra mainboard to the docking station probably wouldn't contain any expensive parts, and if there will ever be a new mainboard, it (the docking station board) could be easily replaced for only a few bucks. Also, the weird USB3.0 port that's really a eSATA and USB2.0 in one is very specific to the OMAP5, so putting the parts for compatibility for this on a seperate board inbetween the desktop connections and the Pyra mainboard makes sense to me. To have a modular design not only in the Pyra itself, but even in it's docking station would be damn cool. And damn expensive in the beginning, but it should hopefully pay off with future upgrades.

 

[Grench]

for the longest time i haven't been able to figure out what images you've been attaching to your posts, @Grench.

then I realized my adblocker was blocking them. are you an ad?

This is something I very much intend to build myself, but if it can be bought from ED's shop in kit form or whatnot, I would much prefer to buy parts like these from there, if nothing else to support the shop.



Thank you! I've been wondering the same thing (on these forums and others).

Nope - not an ad. I do occasionally put in links to products on Amazon as convenient examples since anyone who was truly interested could then follow the link to understand the related specs. Buy things or not wherever you want, but linking to Amazon gives a nice shortcut to be able to describe an item.
[doublepost=1462212933,1462212020][/doublepost]

Why do people suggest only one PCB? The one connecting the Pyra mainboard to the docking station probably wouldn't contain any expensive parts, and if there will ever be a new mainboard, it (the docking station board) could be easily replaced for only a few bucks. Also, the weird USB3.0 port that's really a eSATA and USB2.0 in one is very specific to the OMAP5, so putting the parts for compatibility for this on a seperate board inbetween the desktop connections and the Pyra mainboard makes sense to me. To have a modular design not only in the Pyra itself, but even in it's docking station would be damn cool. And damn expensive in the beginning, but it should hopefully pay off with future upgrades.

Several reasons why I/we suggest ONE board for the docking station.

#1 This will be a niche portion of a niche population for the niche Pyra device. With luck the dock -might- sell 100s of units. That means a VERY low population to spread costs over which means a VERY simplified design and production process.

#2 Every board type that gets made is another trip through the pick & place machine with set up and production costs. If you put in two boards into the dock, it DOUBLES that cost.

#3 Unless someone overdoes it, there are no particularly expensive parts in the dock. A replacement 'interface board' with setup and testing costs would likely cost nearly as much to make as an entire dock board.

#4 As long as the main board (keyboard traces / ports / 4G) stays the same, the dock's interface would remain the same. I.e. even if the Pyra were updated with a new SoC, unless the main board changes too - that connection pattern is constant.

#5 I have no idea what the future SoC would or could be for a future that might dictate a change to the port pattern or types on the back of the Pyra. At the time that it does do that, though, that also means a replacement main board AND modification to or new case molds. Since those are new ports with new capabilities and new ports on the "in side" of the dock 100% means new ports on the "out side" of the dock, a new main board 100% dictates an entirely new dock.

To your specific point about the USB3.0 port that is really a USB 2.0 & eSATA port, that is specific to the main board in the Pyra - and requires that the SoC, whichever one it may be in the future, would need to support eSATA for it to be active under that SoC. If not, then it is no longer active - but that does not prevent it from still working as a USB 2.0.

The fact is, there is no advantage and some loss to take if it -isn't- a single board docking station.

 

[Agricola]

To your specific point about the USB3.0 port that is really a USB 2.0 & eSATA port, that is specific to the main board in the Pyra - and requires that the SoC, whichever one it may be in the future, would need to support eSATA for it to be active under that SoC. If not, then it is no longer active - but that does not prevent it from still working as a USB 2.0.

Last time I checked, ED planned this port to become a regular USB3.0 port if the future SoC supports this - that's why the extra parts for eSATA are on a seperate dongle. The other parts of your post remain valid though, but the only one who exactly knows how much more effort it would be to use 2 PCBs is ED, so I'm just going to wait for a statement from his side, which could take a while considering all of the problems of the preorder that have to get fixed. Whichever way it goes, I'm hyped fo a docking station!

 

[Djoga'Ro]

#2 Every board type that gets made is another trip through the pick & place machine with set up and production costs. If you put in two boards into the dock, it DOUBLES that cost.

How expensive is it to cut a board?

 

[rSl]

maybe it would be convenient to have a standalone battery chargeing-circuit built in the dock to load a second spare battery?
so one has always enough 'juice' using the pyra in handheld mode for a looong session. battery empty? just put in the fully charged one.

 

[mclien]

Some quick thoughts. he pick and place can easily be solved, since you put allways a frame with several boards into p&p (so could be 3 connector and 1 extensive board. basicly an organization thing). I already did organize a small run of the senoko boards (the powermanagement board of the novena project), so i know it's quite doable with a small production number (senoko was 30 boards).

designing the lasercut parts is very easy for me and much less work then cutting a standard box by hand, at least to my point of view and I would do that anyway. After that it's just take the file to tha lasercutter and glue it afterwards...
Maybe we should get a bit of @hns Opinion for this, too. (I remember ED mentioned something that he is already working on something)

 

[rygD]

A quick off-topic...

designing the lasercut parts is very easy for me and much less work then cutting a standard box by hand, at least to my point of view and I would do that anyway. After that it's just take the file to tha lasercutter and glue it afterwards...

This made me think of the logo plate. Do you have the resources and connections (for materials or whatever) to do small runs of affordable generic replacement logo plates (or maybe totally custom, if it isn't too much work or trouble)? This assumes that the original is easy to remove, and we can find something to do about the translucent insert thingy.

 

[mclien]

Yeah, I definitely have have a coworking/fablab/open workplace thing in the build here in my hometown, where I'm involved. And the intermediate version already running (in a small rented place).
I have some CAD skills, my son has really good blender knowledge. So something should happen..
But first I need to know if hns could involve the LCD PCB modification, which could make the exchangeable logo easier. If that works, I'll try talk ED into deliver the Pyra without the standard logo in place, so the customer can decide to either glue the original Pyra logo in or do the modification to make custom one. (If ED thinks it's affordable in the light of the "what could possibly go wrong?" section, with all customers having to do the logo gluing themselves.
For the custom logo there might be drilling involved, because the changing of the mod for that might not be possible (sorry for being too late for that)

 

[rygD]

What would need to happen with the display PCB to help this happen? I haven't looked closely at what is going on back there, and I don't know what changes are planned.

I still think it would be better to have tabs that are part of the logo plate hold it in place. If it is glued I don't want to make any more trouble for ED or most other buyers by having the plate not glued on, or at least not glued on for the few of us that want something custom.

What drilling do you think will be needed? I am probably up for some drilling.

 

[mclien]

Last time I talked to ED ist was planned to glue the logo plate in. Then it will be difficult to change it. My idea is a logo plate which is screwed from the backside, which will require holes in the lit of the LCD, which is a problem to build into the mold. But drilling some holes into the lit isn't a very difficult task for a sligly talanted hacker.
The heads of that screws will face, of course inward the lit, where the LCD PCB is placed. Talking to ED, he mentioned, that large parts of the PCB will be cutted away in the final version to reduce weight. i somply want to ask hns, if it is possible to get the space for the screw heads in that modification, too.

(I'll provide scetches to clarify that, since hns need them, too)

EDIT:
BTW: I keep losing the link to the blender files of teh case. Are there any of the current version around (including the versin with the loogo?