Thought I'd give a try at how a solid wood version would come out. Here is the result of about 45 minutes prototyping with a piece of 3/8" Cherry I had lying around.
A couple notes: The 5/64" (.078", 1.98mm) bit has some tendency to wander, as can be seen in the holes. I also goofed the measurement so these are 11/16" instead of 9/16" wide. I didn't notice that until I was nearly done, and figured that since they weren't actually going to be used it didn't matter.
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
Tor
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 7:44 AM, Tor, the Marqueteur Marqueteur@fineartmarquetry.com wrote:
Thought I'd give a try at how a solid wood version would come out. Here is the result of about 45 minutes prototyping with a piece of 3/8" Cherry I had lying around.
those look _great_.
A couple notes: The 5/64" (.078", 1.98mm) bit has some tendency to wander, as can be seen in the holes. I also goofed the measurement so these are 11/16" instead of 9/16" wide. I didn't notice that until I was nearly done, and figured that since they weren't actually going to be used it didn't matter.
one of the holes is space-saving (for 3d-printing), the other is essential (and is 3.5mm diameter) for the bolts to hold everything together.
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
hmm me too: a CNC lathe would do the trick. the slot is essential in one of the parts for access to the PCMCIA eject-button.
http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2_corner.stl
l.
I just hope you use extremely strong wood Luke.
;)
On 04/30/2017 07:01 AM, Luke Kenneth Casson Leighton wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 7:44 AM, Tor, the Marqueteur Marqueteur@fineartmarquetry.com wrote:
Thought I'd give a try at how a solid wood version would come out. Here is the result of about 45 minutes prototyping with a piece of 3/8" Cherry I had lying around.
those look _great_.
A couple notes: The 5/64" (.078", 1.98mm) bit has some tendency to wander, as can be seen in the holes. I also goofed the measurement so these are 11/16" instead of 9/16" wide. I didn't notice that until I was nearly done, and figured that since they weren't actually going to be used it didn't matter.
one of the holes is space-saving (for 3d-printing), the other is essential (and is 3.5mm diameter) for the bolts to hold everything together.
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
hmm me too: a CNC lathe would do the trick. the slot is essential in one of the parts for access to the PCMCIA eject-button.
http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2_corner.stl
l.
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 12:04 PM, zap zapper@openmailbox.org wrote:
I just hope you use extremely strong wood Luke.
;)
whole device is under 100g, but yes hard-grain wood will be needed as those are quite small grooves.
l.
On 30/04/17 01:38, Luke Kenneth Casson Leighton wrote:
On Sun, Apr 30, 2017 at 12:04 PM, zap zapper@openmailbox.org wrote:
I just hope you use extremely strong wood Luke.
;)
whole device is under 100g, but yes hard-grain wood will be needed as those are quite small grooves.
Given the overall sizes and weights, most woods will work. I wouldn't choose Oak by preference, but on the whole, the list of reasonably common woods that can't be used is probably smaller than the list that can. Maple or Birch would both be very good choices and match the plywood of the rest of the case.
Given Maple prices around here (on the more expensive side of Maple), the wood cost for 2000 of these should still come in under $50 US.
Tor
On 30/04/17 01:01, Luke Kenneth Casson Leighton wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 7:44 AM, Tor, the Marqueteur Marqueteur@fineartmarquetry.com wrote:
Thought I'd give a try at how a solid wood version would come out. Here is the result of about 45 minutes prototyping with a piece of 3/8" Cherry I had lying around.
those look _great_.
Thanks.
A couple notes: The 5/64" (.078", 1.98mm) bit has some tendency to wander, as can be seen in the holes. ...snip...
one of the holes is space-saving (for 3d-printing), the other is essential (and is 3.5mm diameter) for the bolts to hold everything together.
That makes things easier, and the 9/64" bit is more stable. Looks like I probably grabbed the numbers on that from an old version.
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
hmm me too: a CNC lathe would do the trick. the slot is essential in one of the parts for access to the PCMCIA eject-button.
http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2_corner.stl
Yeah. At the complexity of the inside you have there, I don't see anything short of CNC that that could reasonably do the job. Those sharp inside corners are hard, but it looks like a single cutter pass can probably run past far enough to work if that direction were taken.
Tor
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
hmm me too: a CNC lathe would do the trick. the slot is essential in one of the parts for access to the PCMCIA eject-button.
http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2_corner.stl
Yeah. At the complexity of the inside you have there, I don't see anything short of CNC that that could reasonably do the job. Those sharp inside corners are hard, but it looks like a single cutter pass can probably run past far enough to work if that direction were taken.
Not sure if you are looking for one but this reasonably-priced CNC kit machine seems to get good reviews on Amazon: https://www.bobscnc.com/collections/cnc-routers-engravers/products/e3-cnc-en... https://www.bobscnc.com/collections/cnc-routers-engravers/products/e3-cnc-engraving-kit
- Bluey
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 3:04 PM, Bluey bluey@smallfootprint.info wrote:
Not sure if you are looking for one but this reasonably-priced CNC kit machine seems to get good reviews on Amazon: https://www.bobscnc.com/collections/cnc-routers-engravers/products/e3-cnc-en...
that's more than twice the cost of ones here in china :)
my friend found one that's not half bad - absolutely no endstops and using threaded bar not even lead screws of course - but it works surprisingly well and would do the job.
my only concern is: it's gonna take forever to machine 2000 parts, which would need to be turned twice through 90 degrees *in different axes*. that's a 5-axis CNC which starts to get a bit hairy.
l.
On 1 May 2017, at 12:19 AM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 3:04 PM, Bluey bluey@smallfootprint.info wrote:
Not sure if you are looking for one but this reasonably-priced CNC kit machine seems to get good reviews on Amazon: https://www.bobscnc.com/collections/cnc-routers-engravers/products/e3-cnc-en...
that's more than twice the cost of ones here in china :)
my friend found one that's not half bad - absolutely no endstops and using threaded bar not even lead screws of course - but it works surprisingly well and would do the job.
my only concern is: it's gonna take forever to machine 2000 parts, which would need to be turned twice through 90 degrees *in different axes*. that's a 5-axis CNC which starts to get a bit hairy.
l.
Oh, yeah, I saw plenty of ones that were cheaper but the reviews suggested that it would be prudent to give them a wide berth!
I was more thinking of the CNC for cutting any flat-ish timber components; although a 5-axis machine with sufficient precision (e.g., via a specialist milling service) could potentially be used to make moulds for plastic injection manufacture. How many component variations are there to make?
On 1 May 2017, at 1:03 AM, Bluey bluey@smallfootprint.info wrote:
On 1 May 2017, at 12:19 AM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Sun, Apr 30, 2017 at 3:04 PM, Bluey bluey@smallfootprint.info wrote:
Not sure if you are looking for one but this reasonably-priced CNC kit machine seems to get good reviews on Amazon: https://www.bobscnc.com/collections/cnc-routers-engravers/products/e3-cnc-en...
that's more than twice the cost of ones here in china :)
my friend found one that's not half bad - absolutely no endstops and using threaded bar not even lead screws of course - but it works surprisingly well and would do the job.
my only concern is: it's gonna take forever to machine 2000 parts, which would need to be turned twice through 90 degrees *in different axes*. that's a 5-axis CNC which starts to get a bit hairy.
l.
Oh, yeah, I saw plenty of ones that were cheaper but the reviews suggested that it would be prudent to give them a wide berth!
I was more thinking of the CNC for cutting any flat-ish timber components; although a 5-axis machine with sufficient precision (e.g., via a specialist milling service) could potentially be used to make moulds for plastic injection manufacture. How many component variations are there to make?
Just came across this on hackaday: https://hackaday.com/2016/02/09/learn-resin-casting-techniques-duplicating-p... https://hackaday.com/2016/02/09/learn-resin-casting-techniques-duplicating-plastic-parts/
Perhaps it could be a quick, effective, and affordable option once you’ve got a few sets of high-quality components printed?
Cheers, B.
On 30/04/17 04:19, Luke Kenneth Casson Leighton wrote:
my only concern is: it's gonna take forever to machine 2000 parts, which would need to be turned twice through 90 degrees *in different axes*. that's a 5-axis CNC which starts to get a bit hairy.
The other direction to avoid 5-axis would be to make strips about 3' long of appropriate size. Mill the dados with a suitable blade in a mini table saw. A router in table can handle the corners for radius consistency, or P150 in a palm sander (as I used) can do pretty nicely. Then do a sanding of that in 3' lengths.
At this point, except for one lip by the dado that's nearly a rabbet the entire thing becomes a 3-axis flat milling operation. Even that one I suspect could be done by letting the cutter making the rabbet run just past the PCB slot and tolerating the radius.
From there, the possibilities diverge. One option is to cut to corner
length pieces and find a way to secure them in a CNC. This has the advantage of a consistent point to index from, but requires filling a jig with lots of little pieces.
The other option is to load lengths into the CNC, but then either you need the CNC to cut it for accuracy or you have to reference a CNC edge for later cutting, because a small error in length will accumulate fast and throw the whole thing out of tolerance.
Last would be drilling the holes. I just made a mark, lined several up in the vise on my milling machine, and ran across them. Not hard, but it would get a little tedious for 2000.
Looking another way, if the extrusion mold mentioned isn't so difficult, then you just need about 170' of custom extrusion to load into the CNC or cut into pieces and load.
Tor
l.
i was just thinking, if you had the appropriate cutters (like a router / plane with different width bits) it would be possible to just run the parts through some sort of jig.
l.
On 30/04/17 17:47, Luke Kenneth Casson Leighton wrote:
i was just thinking, if you had the appropriate cutters (like a router / plane with different width bits) it would be possible to just run the parts through some sort of jig.
This is definitely possible. With the corner design on my screen (apparently the blue ones in the photos), all the different levels and widths of cutout make it something I'd not want to think about doing that way.
Still, it looks like it could be simplified a bit. You seem to have the PCB slot both top and bottom, which is nice for interchangeability, but could be dispensed with. I'm thinking about other possibilities, but I can't yet visualise fully what the tolerances are relative to the populated PCB.
I know revisions of the PCB are generally hard (at least if they involve rerouting traces), but I have to ask what the implications of adding a sixteenth to each side edge of the PCB for mounting would be. The length front to back would remain the same. If that could be done and would allow the full rectangle of the corner piece to be used, then there are only two operations that would need to be done on the individual corner piece, namely the stopped dado for the PCB (only one per corner, two different corners) and the hole for assembly bolt.
Tor
l.
I never got a reply to this message, and seeing you ready to send the microdesktop PCB to production, I thought I ought to make sure you saw it, Luke.
On 30/04/17 18:36, Tor, the Marqueteur wrote:
On 30/04/17 17:47, Luke Kenneth Casson Leighton wrote:
i was just thinking, if you had the appropriate cutters (like a router / plane with different width bits) it would be possible to just run the parts through some sort of jig.
This is definitely possible. With the corner design on my screen (apparently the blue ones in the photos), all the different levels and widths of cutout make it something I'd not want to think about doing that way.
Still, it looks like it could be simplified a bit. You seem to have the PCB slot both top and bottom, which is nice for interchangeability, but could be dispensed with. I'm thinking about other possibilities, but I can't yet visualise fully what the tolerances are relative to the populated PCB.
I know revisions of the PCB are generally hard (at least if they involve rerouting traces), but I have to ask what the implications of adding a sixteenth to each side edge of the PCB for mounting would be. The length front to back would remain the same. If that could be done and would allow the full rectangle of the corner piece to be used, then there are only two operations that would need to be done on the individual corner piece, namely the stopped dado for the PCB (only one per corner, two different corners) and the hole for assembly bolt.
Tor
l.
On 30 April 2017 at 08:44, Tor, the Marqueteur < Marqueteur@fineartmarquetry.com> wrote:
Thought I'd give a try at how a solid wood version would come out. Here is the result of about 45 minutes prototyping with a piece of 3/8" Cherry I had lying around.
A couple notes: The 5/64" (.078", 1.98mm) bit has some tendency to wander, as can be seen in the holes. I also goofed the measurement so these are 11/16" instead of 9/16" wide. I didn't notice that until I was nearly done, and figured that since they weren't actually going to be used it didn't matter.
I also didn't add the slot for the PCB, mostly because I wasn't sure of the size, and I didn't have the appropriate cutter lying close by. I have a couple thoughts on that.
ok, so I see this part is extruded profile in fact. One alternative is to make it from aluminum. cost of tooling should not be more then couple of hundred USD and you can order some qty for `trial`, prototyping etc. producer will cut you this part as per your size. Basically you pay around 4 usd/kg and some more on cutting, all together not much. just idea.
arm-netbook@lists.phcomp.co.uk
-
Bluey -
Hrvoje Lasic -
Luke Kenneth Casson Leighton -
Tor, the Marqueteur -
zap