plastic enclosure and buttons

[Casainho]

So, I were looking at Thingiverse to see what people had being done. I
found 2 interesting ideas:

- Print buttons directly: Tactile Switch Cap
http://www.thingiverse.com/thing:3146

- Form for silicone cast button pad: http://www.thingiverse.com/thing:1330

--
Cumprimentos,
Jorge Pinto

[Bob]

For one project I was roped in to to mechanical design and build
prototypes. Initially I did this with silicon keypad which I cut out
of remote control units, which had graphite inserts to contact traces
on the PCB. This was not very reliable. Later we had some plastic
keytops designed and made by a very expensive design consultancy,
which looked better on the outside but did not address the electrical
and mechanical design.

The modern trend seems to be for a plastic sheet with metallised domes
to provide electrical contact, so the keytops are just plastic bits.
Either way though, the design of a "simple keypad" is remarkably
intricate. It is technology developed expressly for mass production of
course and very difficult to adapt to low-volume or one-offs.

In order to get any sort of reliable operation I think it is really
best to go with tactile switches, so the key cap idea makes sense,
although they would need to be a lot smaller than the example. I
wonder if it is possible to get "professional looking" results with
the 3d printers? Getting a good fit between the case, keytops and
switches requires some precise tolerances.

I have seen hobby and low volume people get good results using printed
plastic sheets mounted above tactile switches. This combines the
benefit of high quality printing with the robustness of tactile
switches. It has the advantage of hiding the guts so precise machining
is not required.

On 9 June, 11:26, Casainho <casai...@gmail.com> wrote:
> So, I were looking at Thingiverse to see what people had being done. I
> found 2 interesting ideas:
>

> - Print buttons directly: Tactile Switch Caphttp://www.thingiverse.com/thing:3146

[Casainho]

> In order to get any sort of reliable operation I think it is really
> best to go with tactile switches, so the key cap idea makes sense,
> although they would need to be a lot smaller than the example. I
> wonder if it is possible to get "professional looking" results with
> the 3d printers? Getting a good fit between the case, keytops and
> switches requires some precise tolerances.

Well, Open Source DIY printers like RepRap are cheap and many users
have them, but they do not get the same tolerances "professional" and
costly 3D printers.

I think first we should try to print with RepRap, and even because I
have one (not working right now). We can also ask to buy the pieces
from community, because others have optimized their own RepRaps and
get better quality than me. Or maybe I can try to optimize my one.

Another option is to use commercial printing services like this one:
http://www.shapeways.com/ But will it be cheaper that buy the Hammond
plastic case + open holes + making buttons?

Anyway, we need the 3D printing technology if we want to produce a few
different devices from this board, IMO.

[Casainho]

> Well, Open Source DIY printers like RepRap are cheap and many users
> have them, but they do not get the same tolerances "professional" and
> costly 3D printers.

There are users printing layers of 0.35mm. The nozzles holes can be of
0.35mm or 0.3mm, this makes the sizes of the lines.

And I think with acetone, we can try to smooth the faces of the enclosure.

[Bob]

I was bouncing around some ideas for a custom case to fit the current
board + 2.4" LCD. I have uploaded a sketch here
http://rockboxplayer.googlegroups.com/web/nano-concept-custom.png This
design is about 20mm shorter than the Hammond 1553, and a similar
thickness if using a 5mm thick battery. The depth of the back shell
can easily be increased (or reduced) to accommodate different battery
thickness.

If the case was made longer, the battery could be located below the
keypad, and the case then made about 5mm thinner.

I got a bit stuck with how to mount the LCD when it is also soldered
to the PCB. I guess the LCD could snap-fit into the enclosure, then
the PCB is screwed or snapped in behind it. That might be doable
without straining the FFC too much. I don't think it would stand
repeated disassembly/assembly, but hopeful there would be no need to
do that. The battery and uSD which might need replacing can be
accessed from the back without removing the PCB.

On 9 June, 12:53, Casainho <casai...@gmail.com> wrote:
> > In order to get any sort of reliable operation I think it is really
> > best to go with tactile switches, so the key cap idea makes sense,
> > although they would need to be a lot smaller than the example. I
> > wonder if it is possible to get "professional looking" results with
> > the 3d printers? Getting a good fit between the case, keytops and
> > switches requires some precise tolerances.
>
> Well, Open Source DIY printers like RepRap are cheap and many users
> have them, but they do not get the same tolerances "professional" and
> costly 3D printers.
>
> I think first we should try to print with RepRap, and even because I
> have one (not working right now). We can also ask to buy the pieces
> from community, because others have optimized their own RepRaps and
> get better quality than me. Or maybe I can try to optimize my one.
>

> Another option is to use commercial printing services like this one:http://www.shapeways.com/But will it be cheaper that buy the Hammond

[Casainho]

So, the first 2 pictures shows real printed pieces on a "commercial"
3D printer. That detail is the maximum we can expect, I think I will
never be able to get on my printer.
Funny on that pictures is that the pieces are used to assembly that
RepRap 3D printer :-)
(all pictures for that set here:
http://www.flickr.com/photos/nicholasclewis/sets/72157624242649402/)

The 3rd picture is the same piece I printed last time on my printer
and so you can see the best quality I could get until now. But I were
printing with 0.5mm nozzle, while I have one of 0.35mm with me to test
next time :-)

And see how cool are the colors available to print on:
- http://www.flickr.com/photos/23124942@N03/4442251788/sizes/l/in/pool-1024769@N20/
- http://blog.makerbot.com/2010/04/14/1490/
- http://www.flickr.com/photos/matthewlaberge/4627989392/sizes/l/in/pool-1024769@N20/

[Casainho]

On Thu, Jun 10, 2010 at 6:49 PM, Bob <bobcou...@googlemail.com> wrote:
> I was bouncing around some ideas for a custom case to fit the current
> board + 2.4" LCD. I have uploaded a sketch here
> http://rockboxplayer.googlegroups.com/web/nano-concept-custom.png This
> design is about 20mm shorter than the Hammond 1553, and a similar
> thickness if using a 5mm thick battery. The depth of the back shell
> can easily be increased (or reduced) to accommodate different battery
> thickness.

Bob, I was thinking that maybe could be a good idea to have 2 crosses
of buttons on each sides. Advantages:
- having two directional, one for each left and right "big" finger of
each hand - can be good for some applications;
- right-hand and left-hand users would have a choice, and using
firmware config to change the buttons mapping;
- even we could decide to assembly just 2 buttons and we would have a
few options, like select the 2 buttons inline.

But even if a cross on right side is not good, then at least I think
right buttons should have some degree inclination, just like gameboy:
http://www.youtube.com/watch?v=FYY4T3DQExY&feature=related

I think it may be more ergonomic and less boring on design, like not
have having just square lines on buttons/enclosure :-)

> I got a bit stuck with how to mount the LCD when it is also soldered
> to the PCB. I guess the LCD could snap-fit into the enclosure, then
> the PCB is screwed or snapped in behind it. That might be doable
> without straining the FFC too much. I don't think it would stand
> repeated disassembly/assembly, but hopeful there would be no need to
> do that. The battery and uSD which might need replacing can be
> accessed from the back without removing the PCB.

I wrote a message a few minutes talking about this, the LCD mounting.

For battery, do you have any idea? -- I bought this ones from DealExtreme:
http://www.dealextreme.com/details.dx/sku.19771

And here the result:
http://www.flickr.com/photos/43558168@N00/4662568863/sizes/l/

But are you thinking in some other, maybe from DealExtreme? could you
send the link? I would like to buy it because it take some time to arrive.

[Bob]

On 11 June, 12:16, Casainho <casai...@gmail.com> wrote:

> Bob, I was thinking that maybe could be a good idea to have 2 crosses
> of buttons on each sides. Advantages:

Sure we can do that. The button layout is quite arbitrary, they just
have to fit in the approx volume 60x30x5 mm. I haven't checked but I
think are quite a few GPIOs available and probably some analogs. We
could also have right angle switches, so the sides and bottom of the
case could also have switches. We can also use jogwheels etc,
basically anything surface or through hole mounted that fits in the
space. It would help to have a standard number and type of functions
(i.e. digital/analogue) so that the software is uniform but that still
provides plenty of scope.

> I wrote a message a few minutes talking about this, the LCD mounting.

We could print a custom plastic frame to hold the LCD on the PCB. I
have a sketch ;)

>
> For battery, do you have any idea? -- I bought this ones from DealExtreme:http://www.dealextreme.com/details.dx/sku.19771
>
> And here the result:http://www.flickr.com/photos/43558168@N00/4662568863/sizes/l/
>
> But are you thinking in some other, maybe from DealExtreme? could you
> send the link? I would like to buy it because it take some time to arrive.

Yeah, the funny thing there is a suggested the AA size Li-ion and got
a lot of skepticism that such a beast existed, and if it did it would
be hard to get hold of. So I came back with an alternative which is
widely available (iPod type battery) I assumed we were happy with
that.. so anyway we have the option. Here is a typical example
http://www.mdsbattery.co.uk/shop/productprofile.asp?ProductGroupID=1953
but they are available in a lot of places. Note there are some
versions which are li-poly, or have slighter bigger dimensions which
wont fit the 2xAA compartment.

I believe the Apple part number is 616-0159, but it seems to come in
li-ion and li-poly versions. Maybe it doesn't matter... Dimensions
should be about 31mm x 47mm x 6.5mm.

However, if not using AA then there is no reason to use the 1553-BAT,
and the non-battery case could be used, with a slighter wider battery.
Two of the mount points don't line up but I would glue in a spacer to
give support. The support does not need to be particularly strong at
that end of the PCB.

[Casainho]

On Fri, Jun 11, 2010 at 3:48 PM, Bob <bobcou...@googlemail.com> wrote:
> On 11 June, 12:16, Casainho <casai...@gmail.com> wrote:
>
>> Bob, I was thinking that maybe could be a good idea to have 2 crosses
>> of buttons on each sides. Advantages:
>
> Sure we can do that. The button layout is quite arbitrary, they just
> have to fit in the approx volume 60x30x5 mm. I haven't checked but I
> think are quite a few GPIOs available and probably some analogs. We
> could also have right angle switches, so the sides and bottom of the
> case could also have switches. We can also use jogwheels etc,
> basically anything surface or through hole mounted that fits in the
> space. It would help to have a standard number and type of functions
> (i.e. digital/analogue) so that the software is uniform but that still
> provides plenty of scope.

I would go with that buttons Matt told before, from Digikey. We need
to choose the actuation force, so maybe we could buy a few and try. I
already saw/used that buttons on Micropendous boards :-) -- they are
very small in Z axis.

Could you choose 2 or 3 different buttons from that ones?

>> I wrote a message a few minutes talking about this, the LCD mounting.
>
> We could print a custom plastic frame to hold the LCD on the PCB. I
> have a sketch ;)

Hmm, won't the full enclosure be able to do the frame already?

Today I were measuring the thickness of walls of latest enclosure for
Lyre I did design and print (http://www.thingiverse.com/image:9994),
and it have 2.5mm. The Hammond enclosure also have walls of 2.5mm. So,
the printed enclosure will not need to add more space/volume to
printed enclosure, when compared to this Hammond. Also I am sure I can
make it a bit small, like 2mm.

Also I were trying acetone on that printed enclosure and it gives a
smooth finish to the printed plastic :-) -- I will test again with
"strong" acetone, because I tested with the one used by woman's for
clean the finger nails.

>> For battery, do you have any idea? -- I bought this ones from DealExtreme:http://www.dealextreme.com/details.dx/sku.19771
>>
>> And here the result:http://www.flickr.com/photos/43558168@N00/4662568863/sizes/l/
>>
>> But are you thinking in some other, maybe from DealExtreme? could you
>> send the link? I would like to buy it because it take some time to arrive.
>
> Yeah, the funny thing there is a suggested the AA size Li-ion and got
> a lot of skepticism that such a beast existed, and if it did it would
> be hard to get hold of. So I came back with an alternative which is
> widely available (iPod type battery) I assumed we were happy with
> that.. so anyway we have the option.

The AA size have the same x size of actual board. Y size of 16mm. The
Z size maybe equal to the one of PCB1 + PCB2 + LCD. I am happy with AA
battery, even I have external charger for it, cheap from also
DealExtreme.

Maybe we can even stay with this AA, because I have one more which I
can offer you and send next time I will send more components. We could
now stay with this battery and later work again on this question,
maybe that one of Ipod is a better choice, even because the connector
and store more power (but have more volume and weight). I think that
when we have the PCB1 + PCB2 of buttons and LCD, then we can choose
better the battery/shape of our player - at the same time we will
design the enclosure.

Here is a typical example
> http://www.mdsbattery.co.uk/shop/productprofile.asp?ProductGroupID=1953
> but they are available in a lot of places. Note there are some
> versions which are li-poly, or have slighter bigger dimensions which
> wont fit the 2xAA compartment.
>
> I believe the Apple part number is 616-0159, but it seems to come in
> li-ion and li-poly versions. Maybe it doesn't matter... Dimensions
> should be about 31mm x 47mm x 6.5mm.
>
> However, if not using AA then there is no reason to use the 1553-BAT,
> and the non-battery case could be used, with a slighter wider battery.
> Two of the mount points don't line up but I would glue in a spacer to
> give support. The support does not need to be particularly strong at
> that end of the PCB.

I bought that enclosure that do not have battery case, just because
it's cheap and I had AA battery but just wanted to use one (not sure
on how to put them in parallel and safety).

[Casainho]

Bob, since you were making the sketch for the
boards/enclosure/position, would you mind to give-me the sizes for the
enclosure to design/print?

I am reading my printer, and I have motors working, I am not far from
working machine (I hope, mechanical problems may be worst for me to
resolve). I hope to be able to start printing if 5 days or less.

I hope to use Blender for design the enclosure, however if I could
find a easier one for this task....

http://www.blender.org/

[Bob]

I call them sketches because they are an approximate scale, I can
specify the size precisely in mm but positioning the components is
quite tedious. I tend to ignore clearance and case thickness for
simplicity.

Rather than spend more time on the drawings, when I have the actual
components, I can build a physical model and see how it goes together,
and also try fitting it into a box like the Hammond.

At that point I can write down dimensions, I might have a stab at
Blender again.

[Casainho]

> At that point I can write down dimensions, I might have a stab at
> Blender again.

If you could go and draw it in Blender, would be perfect!

And for start, I would make a "square" box, without holes. The idea is
for I go print many revisions and giving feedback.

And yes, I was thinking in internal dimensions. Walls should be of
2.5mm for now, later we can try to go for 2mm (or maybe a bit less).

[Bob]

On 13 June, 12:08, Casainho <casai...@gmail.com> wrote:

> And for start, I would make a "square" box, without holes. The idea is
> for I go print many revisions and giving feedback.

Oh in that case I could do it quicker. I am looking at internal
dimensions of 104x62x22mm, so that could be two half shells at
104x62x11.

Then we can add cutouts, mounting pillars, and maybe a place for the
battery?

I really should write that program to print out folding box designs on
paper.

[Casainho]

> Oh in that case I could do it quicker. I am looking at internal
> dimensions of 104x62x22mm, so that could be two half shells at
> 104x62x11.

Yes, just that, two half shells.

> Then we can add cutouts, mounting pillars, and maybe a place for the
> battery?

Not yet right now, because it's very difficult to go back if we need IMO.

Did we already decided for a battery? I would prefer AA since looks to
be smaller (but less energy comparing that one for Ipod), meaning a
smaller device.

> I really should write that program to print out folding box designs on
> paper.

You know, others like Domonoky (yes, the other hacker from Rockbox
that even coded for Mini2440) are coding is 3D designs in OpenSCAD:
http://openscad.org/

Here an example made by Domonoky "Sansa e200 series Dock":
http://www.thingiverse.com/thing:1555

[Opendous Support]

>I really should write that program to print
>out folding box designs on paper.

A long time ago I remember reading somewhere about some Open Source
box creation software. Input dimensions and out comes a printable
page with folding and cutting instructions. All I have been able to
find is the following:

http://www.cpforbes.net/tuckbox/

Later,
Matt,
Opendous Inc.

> --
> You received this message because you are subscribed to the Google Groups "rockboxplayer" group.
> To post to this group, send email to rockbo...@googlegroups.com.
> To unsubscribe from this group, send email to rockboxplaye...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/rockboxplayer?hl=en.
>
>

[Bob]

I have made a Thing! http://www.thingiverse.com/thing:3363

OpenSCAD is brilliant, thanks for the tip!

On 13 June, 12:08, Casainho <casai...@gmail.com> wrote:

[Casainho]

On Thu, Jun 17, 2010 at 7:40 PM, Bob <bobcou...@googlemail.com> wrote:
> I have made a Thing! http://www.thingiverse.com/thing:3363

Nice!

We should think if we can make the enclosure without using screws. I
went to a local hardware shop and they don't have small screws... :-(

> OpenSCAD is brilliant, thanks for the tip!

Well, maybe I will look your sources and see if I understand how it works ;-)

The axis of my RepRap Mendel works nicely. I can print and it looks
like to works. The extruder, I ended it just today and the motor and
the mosfet for heater seems to work.
Now I need to make the heater and test it - tomorrow. I will need also
to assembly the limit opto switches for each 3 axis. Finally I will
need to print a lot to fine tune the printer.

If you keep update the thing:3363, I will use it to fine tune the
printer and soon I have something good, I will upload pictures to that
thing :-)

[Casainho]

My extruder extrudes now :-) I tried with ABS and PLA plastic :-)

With the 0.35mm nozzle I am getting extruded wires of 0.8mm. I am not
sure, but I hope to get 1mm lines when extruding and printing. Maybe I
will be able to get walls of 2mm (minimum of 2 wires, 1mm each wire).

I am sending pictures, and on one we can see plastic extruded, that I
extruded for testing the extruder. Also the electronics are based on 3
Arduino boards!! (One board just for providing 5volts power from USB
and USB<->serial).

[Bob]

On 17 June, 23:48, Casainho <casai...@gmail.com> wrote:
> We should think if we can make the enclosure without using screws. I
> went to a local hardware shop and they don't have small screws... :-(

I prefer screws over little plastic tabs that break off any day, but
you have come up with a terrible excuse ;) You can buy small screws
lots of places, e.g. on ebay!

I would think the low resolution of the 3D printer, plus unpredictable
properties of the plastic, means that the tabs will have to be quite
large, which may negate the goal. They are also more intricate to
design.

I would be more willing to give it a go if there was a more convincing
reason not to use screws ;)

[Casainho]

It was an idea. Who knows if later we can design a strong ones?

I would prefer to buy screws on local shop. Ok, I don't have any idea
for type of screws yet, maybe best is to print the enclosure and send
to you, on that time we can discuss this idea again.

[Bob]

And another thing... http://www.thingiverse.com/thing:3372

This is a two part box in the style of the Hammond box. I have
included pillars but I have not yet placed them in the correct
position for our board.

[Casainho]

On Fri, Jun 18, 2010 at 10:23 PM, Bob <bobcou...@googlemail.com> wrote:
> And another thing... http://www.thingiverse.com/thing:3372
>
> This is a two part box in the style of the Hammond box. I have
> included pillars but I have not yet placed them in the correct
> position for our board.

Nice!

"Designed for Lyre audio player project." -> can you put a link to
project page? this project needs all possible references ;-)

Maybe put a link like I did here: http://www.thingiverse.com/thing:1795

[Bob]

I have updated the design with the correct layout of mounting pillars
for the Lyre v1 PCB. In this version it is basically a reproduction of
the Hammond 1553 case.

At some point I will add apertures for connectors etc. The mounting
bosses in the upper part may change, for now they are the same as the
lower part.

I haven't checked the design with paper prints so use with caution,
but I believe the dimensions are correct. We can of course play around
with sizing but don't ask for any curved surfaces ;)

To view the STL files I am using a "free" STL viewer from these guys
http://www.modelpress.com/. It works well but I wonder if there any
easy to use open source viewers?

A way to convert the 3D model to accurate 2d prints would be nice too!

Regards

On 19 June, 00:46, Casainho <casai...@gmail.com> wrote:
> On Fri, Jun 18, 2010 at 10:23 PM, Bob <bobcousin...@googlemail.com> wrote:
> > And another thing...http://www.thingiverse.com/thing:3372

[Casainho]

> To view the STL files I am using a "free" STL viewer from these guys
> http://www.modelpress.com/. It works well but I wonder if there any
> easy to use open source viewers?

Blender! it can import and edit STL files ;-)

> A way to convert the 3D model to accurate 2d prints would be nice too!

Maybe in Blender, however I never did such thing. In Blender, one unit
== 1mm, so you can go and re-scale your model to 1 unit == 1 mm. Then
I am sure you can export images for each axis or something like that
:-)

[Bob]

Ah, the key phrase there was "easy to use". Blender certainly is not.
YMMV

[Casainho]

On Sat, Jun 19, 2010 at 9:02 PM, Bob <bobcou...@googlemail.com> wrote:
> Ah, the key phrase there was "easy to use". Blender certainly is not.
> YMMV

I am not sure if there are anyone better and Open Source. Thingiverse
have a recent blog message with links for video tutorials on Blender:
http://blog.thingiverse.com/2010/06/22/blendercookie-tutorials/ -- I
think they also like Blender for 3D design/print.

My printer now prints :-) -- here one video when printing your
enclosure: http://blip.tv/file/get/Jpcasainho-my_first_reprap_mendel_printjpcasainho20100622ogm407.ogv

However I went to short it since I have one AA battery and I do not
need all the space available - since you use the same dimensions from
Hammond enclosure.

I do not have yet a print of your enclosure, because the software I am
using now to STL -> GCode is not good. The one I was using before were
very good but it's now giving me some problems. I will try to figure
it.

Also I do not have the 0.35mm nozzle working, instead the 0.5mm one,
because I had some problems for putting the 0.35mm working - I will
try again later. For now I will try to get best quality possible with
this 0.5mm.

--
Cumprimentos,
Jorge Pinto

[Bob]

Nice! Now I am into the 3D design I may have to get a 3D printer ;)

I realised that the mounting bosses may have too thin walls, the
screws might break them. I will try to do a version with 2mm wall
thickness.

My internet access is pretty poor so I won't be online so much this
week.


On 22 June, 15:21, Casainho <casai...@gmail.com> wrote:

> On Sat, Jun 19, 2010 at 9:02 PM, Bob <bobcousin...@googlemail.com> wrote:
> > Ah, the key phrase there was "easy to use". Blender certainly is not.
> > YMMV
>
> I am not sure if there are anyone better and Open Source. Thingiverse

> have a recent blog message with links for video tutorials on Blender:http://blog.thingiverse.com/2010/06/22/blendercookie-tutorials/-- I

> think they also like Blender for 3D design/print.
>
> My printer now prints :-) -- here one video when printing your

> enclosure:http://blip.tv/file/get/Jpcasainho-my_first_reprap_mendel_printjpcasa...

[Casainho]

On Tue, Jun 22, 2010 at 7:37 PM, Bob <bobcou...@googlemail.com> wrote:
> Nice! Now I am into the 3D design I may have to get a 3D printer ;)
>
> I realised that the mounting bosses may have too thin walls, the
> screws might break them. I will try to do a version with 2mm wall
> thickness.

I printed the enclosure, and seems that both parts are equal, no?

Even after I shorted 3mm from actual Bob design, there are more 4mm to
remove from the enclosure (for using with size AA battery).

We still have a few mm to gain on Z, from actual enclosure. The big
cilinders at center of box for use screws, are 2 or 3mm displaced from
top (USB connector) board.

My print is a bit bad, is not optimized and there are a gap between
walls. And I think we should go for 2mm walls.

I don't know what to do now, because I can keep changing my STL files
and that will not help Bob with OpenScade. I don't have free time to
learn OpenScad and I already invested on Blender.

[Bob]

On 23 June, 18:09, Casainho <casai...@gmail.com> wrote:

> I printed the enclosure, and seems that both parts are equal, no?

That looks pretty good!

They may look similar, but the lower part has a hole through the case
and a countersink for the screw, the top part has no screw holes. In
future when I add apertures for the connectors, LCD and buttons, the
top/bottom will be significantly different.

> Even after I shorted 3mm from actual Bob design, there are more 4mm to
> remove from the enclosure (for using with size AA battery).

I feel like a broken record... you can't just go round arbitrarily
chopping bits off. It may look like I threw this together in 5
minutes, but I have actually spent quite a while thinking about it and
designing it.

I guess if you don't want to leave any room for buttons, then go
ahead, hack away!

> We still have a few mm to gain on Z, from actual enclosure. The big
> cilinders at center of box for use screws, are 2 or 3mm displaced from
> top (USB connector) board.

Maybe... I am thinking of changing the design a lot and mounting the
PCB to the top half so that it clamps the LCD to the case. The back
can have a slot for the battery.

> My print is a bit bad, is not optimized and there are a gap between
> walls. And I think we should go for 2mm walls.

No problem...

> I don't know what to do now, because I can keep changing my STL files
> and that will not help Bob with OpenScade. I don't have free time to
> learn OpenScad and I already invested on Blender.

Well, now it is my turn to say that OpenSCAD is so easy you don't even
need to watch any training videos to use it! I don't think you should
be changing the design anyway ;)

I have put the SCAD source and STL files in the Lyre SVN under
"box_design", btw.

[Bob]

Ok, I have done a new concept for a custom case. See
http://groups.google.co.uk/group/rockboxplayer/web/nano-concept-custom-2.png?hl=en

- The internal volume is 100x60x22mm.
- the design allows for either AA or iPod type battery. A couple of mm
can be shaved off the height by using the iPod battery, but not much
in it. A long battery with a thin profile would be good if we can find
one. Alternatively the case can be extended in the X direction.
- button board can be strip board or PCB
- I have no solution currently for key caps, so the plungers simply
protrude slightly through the case

Assembly :
- the button board is fixed in.
- the LCD is soldered is the main PCB
- the connection from the main board to the button board is with
soldered leads
- the LCD is friction fit into case housing.
- main board is then screwed to top half of case
- the battery is mounted to lower half
- case halves screwed together

Issues:
- no power switch yet! Maybe have a right angle mounted switch
somewhere

Cheers

[Bob]

Yesterday I got the 2.4" LCD etc from Digikey. The Flex connector on
the LCD is a lot less flexible than I anticipated, in fact I would
call it slightly flexible. It really needs to lie flat against the
PCB, especially if it is glued to the PCB. There are really only two
position where the LCD can sit without excessive strain on the
connector, one overhangs the case and the other is in the middle of
the case, not leaving much room for buttons. In addition the LCD
footprint on the PCB is offset from the centre.

Therefore I can't really see the concept I posted being a workable
solution, sadly. I guess it depends on how far one is willing to bend
the connector.

I will have another look at using a secondary board to carry the LCD
and buttons.

On Jun 25, 9:09 pm, Bob <bobcousin...@googlemail.com> wrote:
> Ok, I have done a new concept for a custom case. Seehttp://groups.google.co.uk/group/rockboxplayer/web/nano-concept-custo...

[Casainho]

On Sat, Jun 26, 2010 at 3:05 PM, Bob <bobcou...@googlemail.com> wrote:
> Yesterday I got the 2.4" LCD etc from Digikey. The Flex connector on
> the LCD is a lot less flexible than I anticipated, in fact I would
> call it slightly flexible. It really needs to lie flat against the
> PCB, especially if it is glued to the PCB. There are really only two
> position where the LCD can sit without excessive strain on the
> connector, one overhangs the case and the other is in the middle of
> the case, not leaving much room for buttons. In addition the LCD
> footprint on the PCB is offset from the centre.
>
> Therefore I can't really see the concept I posted being a workable
> solution, sadly. I guess it depends on how far one is willing to bend
> the connector.
>
> I will have another look at using a secondary board to carry the LCD
> and buttons.

When I receive the LCD and components you sent me, I will take some
pictures to show it on the enclosure. Also show the "ipod" battery.

My 3D printer is giving me some problems (on extruder I guess) and
because of that I weren't able yet to full print an enclosure.

Does anyone have ideas for the design, I mean, with some rounded curves?

[Casainho]

Here some nice pictures of an LCD "flexible" cable... and a small enclosure:

http://forums.benheck.com/viewtopic.php?f=1&t=38001&start=30

[Casainho]

On Sat, Jun 26, 2010 at 3:05 PM, Bob <bobcou...@googlemail.com> wrote:

> Yesterday I got the 2.4" LCD etc from Digikey. The Flex connector on
> the LCD is a lot less flexible than I anticipated, in fact I would
> call it slightly flexible. It really needs to lie flat against the
> PCB, especially if it is glued to the PCB. There are really only two
> position where the LCD can sit without excessive strain on the
> connector, one overhangs the case and the other is in the middle of
> the case, not leaving much room for buttons. In addition the LCD
> footprint on the PCB is offset from the centre.
>
> Therefore I can't really see the concept I posted being a workable
> solution, sadly. I guess it depends on how far one is willing to bend
> the connector.
>
> I will have another look at using a secondary board to carry the LCD
> and buttons.

And I got my LCD, battery and more components, sent by Bob :-) - thanks!

I also think the only/better way for the LCD is to have a secondary
board and solder it there. I am sending 2 pictures of it.

I think the connector should be soldered under the LCD, just like I
show on 2nd picture - that way the can solder the LCD on that board
and the buttons.

I can't see how this battery can be placed under the boards... maybe
the best it to get done the secondary PCB and then we will see.

As for enclosure, I were stuck many days because of a bug on RepRap
Mendel firmware - there are a problem on it and I found a way to avoid
it while losing a functionality (I wish there was an OpenSource and
cheap debugger for Arduino/AVR like for ARM).
I am printing right now and I hope to finally get the enclosure.

Bob, the screws you sent are BIG! I was expecting ones of half of
size. You also didn't sent the low profile connectors :-(

I am very busy on this days, as soon as I can, I will start looking
for the schematic Bob did the for secondary board and try to verify
it.

[Bob]

On 30 June, 17:07, Casainho <casai...@gmail.com> wrote:
> Bob, the screws you sent are BIG! I was expecting ones of half of
> size. You also didn't sent the low profile connectors :-(

They are the ones specified by Hammond for their case.

> I am very busy on this days, as soon as I can, I will start looking
> for the schematic Bob did the for secondary board and try to verify
> it.

I have made some changes to the IO board and uploaded a new version to
the Lyre SVN. I haven't updated PDFs yet.

I changed the GPIOs used per Matt suggestion, added a mini JTAG
connector and a power switch. The power switch will need some flying
wires to the main board.

I am hoping to get some low profile side actuated tactile switches so
that the power switch can be operated from a button at the bottom end
of the case. Our top electronics retailer has 2 versions and both are
out of stock ;) There are many variants at Digikey and Mouser but
these are proving expensive to order from. Our last order from the US
attracted a 30% import charge.

In the meantime a cheap solution may be to glue a regular tactile
switch to the side of the PCB, and solder on some wires.

[Casainho]

On Wed, Jun 30, 2010 at 11:37 PM, Casainho <casa...@gmail.com> wrote:
> On Wed, Jun 30, 2010 at 7:30 PM, Opendous Support
> <opendous...@gmail.com> wrote:
>> I am busy as well but will mount my LCD screen and create a buttons
>> test board in the next couple of days. I plan on etching the PCB
>> myself to save $. If you can manage to get your secondary board to be
>> single layer I can etch yours as well. I need to fill up an 8.5x11
>> size PCB and will need to fill it with adapters or something anyway.
>> Make your traces as wide as possible when routing the board.
>
> Matt, can't you re-use Bob secondary board and design one for us? you
> would have to make 3 boards, one for me, other for Bob and the last
> one for you.
>
>
>> Casainho, your headphones connector will not work as mounted because
>> there is no electrical connection between it and the i.MX233. You
>> need to mount the resistors, ferrites, and capacitors.
>
> Yeah, I know. That was just to mount the audio connector to see the
> maximum Z. I want first to test the LCD and buttons, I will leave
> audio for the end of Rockbox port.
>
> Matt, If you design and etch a secondary for us 3, I will print and
> offer one enclosure for both of us :-)
>
> Right now I don't know what battery we will use, because I can't see
> how the "Ipod" one can be used... Bob??
>
> I just printed 2 times the same enclosure, bottom side :-) -- I really
> like the results, with strong walls.
>
> So, next step is to have the buttons + LCD board done, and define the
> best mechanical space for all the system, including the battery place.
> We will quickly design the full enclosure, I will print and send to
> all of us :-)

I forgot to send this e-mail to group.

I just bought 4 AAA size rechargeable lithium batteries, 3.7V 300mA**:
https://www.dealextreme.com/details.dx/sku.36026
This ones are smaller, 46mm X 10mm X 10mm, they have also much less energy...

For Z axis, this batteries should be better, let us put the LCD +
buttons board on top and have even a lower Z then Hammond enclosure!

As for that images with AA battery size I am sending, this ones AAA
batteries should give us less 4mm in Y axis :-)

** https://www.dealextreme.com/forums/Default.dx/sku.36026~threadid.621727

[Casainho]

Ok, I just printed an enclosure for AAA size battery :-) I think the
battery can be glued and have wires soldered to contacts - example of
soldering wires:
- http://linux-7110.sourceforge.net/howtos/netbook_new/LIbattery/LIbattery.html
- http://terrorhurtz.com/a123/build.aspx

The LCD connector do not touch the expansion header, and so the board
can be done to have LCD at center.

The white buttons on picture 7, are the ones suggested by Matt on a
Micropendous board. That buttons have the same height as LCD.

This idea is different from Bob board. Also I think Matt already told
about this idea.

I think secondary board must have battery power input (if possible),
so we can solder battery wire to it. Also the board should be
rectangular but smaller as possible, and so we will have room to make
round lines on enclosure, for a better design.
The AAA battery adds 10 mm to Y axis of main board, and so are this
the interior dimensions.

[Bob]

The low profile switches are nice, they generally have a short
plunger. Either way some keycaps are needed. I have ordered some from
RS http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=174-3168
I am not sure exactly which switches they will fit.

I have designed some keycaps with SCAD, would you be interested in
printing some and see if they are usable?

I have uploaded a screenie here http://groups.google.co.uk/group/rockboxplayer/web/keycap_round.JPG?hl=en

[Casainho]

>  I will etch three 1-layer copies of Bob's board when it is done.
> Sounds like a fair trade; PCB for Enclosure.

It think is done now, however it's 2 layers. Matt, could you design
that board for us? maybe if you have some that buttons from
Micropendous, you could use/source them to us (6 buttons for each one
of us).

I forgot to say, but I would like to offer the enclosure and the AAA
lithium battery. I also can offer AA size if later we want an AA size.

We need to decide a good size for buttons tap, I would like to be like
the original gameboy.

I want now to try print using 0.35mm nozzle, to achieve higher print quality.

Also I would like to design and share the 3D files for enclosure
bottom, top, cross button tap and round button taps -- with that files
we can try to make/discuss different designs, like I think that with
actual space/volume, LCD could be off the center. Unfortunately, for a
different LCD and buttons placement we need a different secondary
board (however it's much more simple to make than the imx233 base
board), if not, users could quickly create their own design!! :-) --
you know, the business of personalized drawing and colours on mobiles
phones and computers... :-)

>  I want to do my own very simple buttons board as well.  It will be
> much simpler than Bob's expansion board and I am hopeful it will be
> adequate.  Just 8 buttons and an I2C IO Expander connected at the
> bottom edge of the PROpendous using just 4 pins from the expansion
> header.  It will be just about the same as
> lyre_enclosure-05-20100701.jpg.  The LCD can then be connected to the
> main board.

But the LCD will not be at center of enclosure, I think. Don't you
want instead to go with the same solution as me and Bob? I believe
would be better for all of us, for the success of the project - we in
very small numbers yet, we should join our resources. Later anyone can
build his custom enclosure/board/etc.

>  As for updates to the PROpendous board, I have given up on the idea
> of an 0805-based version.  If a DIY'er can do 0.4mm QFP they can do
> 0603.  A far more important goal is getting all the ICs on one side
> (leaving the bottom only for decoupling capacitors and other
> components that can withstand 2 reflow cycles).  The next board
> revision I will produce will have the DDR on the top layer and only
> the 2.4" LCD connector (need room for DDR so no 1.8").  I will send it
> out for fabrication after testing the LCD.

I would remove the video out section, serial JTAG, boot EEPROM,
accelerometer and SDCard. Maybe one uSDCard for boot and user data are
enough? or maybe 2 uSDCard (one for user data)? -- for a device that
charges over USB, I believe users will use USB for put the data files
on uSDCard. Ok, SDCard have more space and are cheap, then one SDCard
could be added on expansion board.

I would like to see board a bit smaller if possible, but also more
rectangular and not almost square as now. For a player, I think
rectangular is more beauty/ leave more room/options for enclosure
design.

>  Bob, have you managed to test the DDR SDRAM on your board?

I did it on Bob board before sent to him, however just the 000 and 111 test.

>>I think secondary board must have
>>battery power input (if possible)
>

>  If you are soldering wires directly then it might as well be on the
> main board.  Let me know the best position for the battery connector
> and I will try to move it.

Yeah, would be nice to be able to solder on pads. We can save on the
connector and maybe we will need to always solder cables thanks to
many different batteries.

[john cooper]

Casainho wrote:
> It think is done now, however it's 2 layers. Matt, could you design
> that board for us? maybe if you have some that buttons from
> Micropendous, you could use/source them to us (6 buttons for each one
> of us).

:

> We need to decide a good size for buttons tap, I would like to be like
> the original gameboy.

I had substituted ALPS mounted dome bare tacts (3.7mm x 3.7mm)
in my interface board in favor of the larger conventional tacts
and directional nav -- mostly motivated by the significant cost
differential. Although when the samples arrived I had second
thoughts of hitting the pinhead-sized bare domes accurately with
a keycap bottom probe held loosely captive in the case. I don't
believe I've ever seen a smaller tact/snap dome in any form.

> Also I would like to design and share the 3D files for enclosure
> bottom, top, cross button tap and round button taps -- with that files
> we can try to make/discuss different designs, like I think that with
> actual space/volume, LCD could be off the center. Unfortunately, for a
> different LCD and buttons placement we need a different secondary
> board (however it's much more simple to make than the imx233 base
> board), if not, users could quickly create their own design!! :-) --
> you know, the business of personalized drawing and colours on mobiles
> phones and computers... :-)

That was my rationale months ago for a secondary board, namely
flexibility of user interface layout. Still for maximized
maneuverability you may want to use surface mount connectors on at
least the user board as through hole versions will non-negotiably
claim real estate on both surfaces. Even using surface mount
connectors I had enough of a headache in several cases straddling
the base board connector and LCD footprint given two layers.

> But the LCD will not be at center of enclosure, I think. Don't you
> want instead to go with the same solution as me and Bob? I believe
> would be better for all of us, for the success of the project - we in
> very small numbers yet, we should join our resources. Later anyone can
> build his custom enclosure/board/etc.

One thought I had when laying out my board was rotating the works
90* to create a landscape orientation (which personally I prefer).
That could have helped somewhat accommodating other LCD
configurations however doing so left no usable room for navigation
and tact switches. Something you might want to keep in mind as
you have more real estate with which to work.

>> As for updates to the PROpendous board, I have given up on the idea
>> of an 0805-based version. If a DIY'er can do 0.4mm QFP they can do
>> 0603.

Yea that was my reasoning albeit stretched to justify use
of 0402 packages. The lqfp-128 itself realistically sets the
basic fabrication difficulty here.

> A far more important goal is getting all the ICs on one side
>> (leaving the bottom only for decoupling capacitors and other
>> components that can withstand 2 reflow cycles).

While this was one of my initial motivations to keep large
packages on the component surface, in retrospect it may have
been more of a speculated need. But ultimately locating
bypass close to the imx233 (and DDR to a lesser extent)
pushed that decision.

It worked out ok so far. I did try to keep all solder side
discretes under 1mm height to fit in the available board-to-battery
void left by the direct-mate smd battery connector.

> I would remove the video out section, serial JTAG, boot EEPROM,
> accelerometer and SDCard. Maybe one uSDCard for boot and user data are
> enough? or maybe 2 uSDCard (one for user data)? -- for a device that
> charges over USB, I believe users will use USB for put the data files
> on uSDCard. Ok, SDCard have more space and are cheap, then one SDCard
> could be added on expansion board.

You're converging. :)

Still I'm not sure I'd lose the i2c flash just yet. If it
helps, I was able at one point to locate the flash in a tssop-8
footprint opposite the DDR which was largely unused. Though
I've since relocated the tssop-8 to the component side, the
old footprint still exists opposite the DDR as that space
thus far wasn't needed for any other use.

> I would like to see board a bit smaller if possible, but also more
> rectangular and not almost square as now. For a player, I think
> rectangular is more beauty/ leave more room/options for enclosure
> design.

Maybe. I think the user interface layout (LCD and input devices)
have a greater impact on the choice of footprint rectangle ratio
and size.

-john

--
john....@third-harmonic.com

[Bob]

On 1 July, 23:18, Casainho <casai...@gmail.com> wrote:
> But the LCD will not be at center of enclosure, I think. Don't you
> want instead to go with the same solution as me and Bob? I believe
> would be better for all of us, for the success of the project - we in
> very small numbers yet, we should join our resources. Later anyone can
> build his custom enclosure/board/etc.

I don't think we are working on the same solution, you seem to have
made several changes independently. That's fine, just don't say we are
on the same solution. I just don't understand the AAA/Gameboy
direction at all really. It hardly worth seems having an agreed goal,
when the design decisions seem to change randomly.

To be honest, I am not very interested in 3D printed enclosures. My
goal was to design for the Hammond box, and hopefully produce
something that you can also use comes out of that.

[Casainho]

On Fri, Jul 2, 2010 at 4:52 PM, Bob <bobcou...@googlemail.com> wrote:
> On 1 July, 23:18, Casainho <casai...@gmail.com> wrote:
>> But the LCD will not be at center of enclosure, I think. Don't you
>> want instead to go with the same solution as me and Bob? I believe
>> would be better for all of us, for the success of the project - we in
>> very small numbers yet, we should join our resources. Later anyone can
>> build his custom enclosure/board/etc.
>
> I don't think we are working on the same solution, you seem to have
> made several changes independently. That's fine, just don't say we are
> on the same solution. I just don't understand the AAA/Gameboy
> direction at all really. It hardly worth seems having an agreed goal,
> when the design decisions seem to change randomly.

Yeah, that's true. I just tested the IPod battery, AA and AAA and I
think AAA will give the smaller device. I even shared the pictures of
different enclosures with different sizes. I thought we all wanted a
smaller device as possible that why I decided for the AAA size.

> To be honest, I am not very interested in 3D printed enclosures. My
> goal was to design for the Hammond box, and hopefully produce
> something that you can also use comes out of that.

Ok. I prefer to try use the same board/hardware.

[Bob]

On 2 July, 20:31, Casainho <casai...@gmail.com> wrote:
> Yeah, that's true. I just tested the IPod battery, AA and AAA and I
> think AAA will give the smaller device. I even shared the pictures of
> different enclosures with different sizes. I thought we all wanted a
> smaller device as possible that why I decided for the AAA size.

So what happened to "just use a bigger battery"? I guess you will have
forgotten that discussion ;) Small as possible means not wasting
space, it doesn't mean removing useful features. Remember also that
you chose the large LCD?

Anyway, I have updated the Lyre SVN with a slightly improved IO board.
I have shaved a few mm off the ends. Further size reduction would
involve using smaller switches, and as John notes there are some
really small switches. The question there is what is the minimum
usable spacing, personally I struggle with keys spaced <10mm. I would
like to get this board made fairly soon, even though it's not quite
final since I am still waiting on a few components. I have attempted
to create dual footprints for SMT/through hole switch styles.

Obviously with a touch screen we shouldn't need so many keys, but we
need support in the software. The Rockbox touch support is a bit
primitive last time I tried it, however we should look at doing a
"Lyre Touch" in future with minimal or no keys.

I have also revised the concept sketches and will upload a PDF to the
files area. Basically we have two options, a short case which is
thick, or a longer case which is thin. In both variants, an AA size or
iPod size battery can be used, it doesn't make much difference. The
"long" design is more efficient on space, but requires low profile
connector/connection. The "short" case would allow "standard" height
connectors.

The "Long" design is almost compatible with the Hammond 1553B (non-
battery) case, which has internal dimension of about 115x18.5mm. It
will work with the iPod battery, but the AA requires taking a further
4mm or so off the IO board.

[Bob]

Some PDFs

http://groups.google.co.uk/group/rockboxplayer/web/lyre-concept-custom.pdf?hl=en

http://lyre.svn.sourceforge.net/viewvc/lyre/hardware_design/lyre_v1_io/lyre_v1_io_sch.pdf?revision=39
http://lyre.svn.sourceforge.net/viewvc/lyre/hardware_design/lyre_v1_io/lyre_v1_io_pcb.pdf?revision=39

On 4 July, 16:26, Bob <bobcousin...@googlemail.com> wrote:

[Casainho]

I am now reading about etching PCB and find if I can do it at home.

Bob, now with that images I can understand well the placement of all parts.

Why 2 buttons for power? why not use one of the 8 buttons?

[Bob]

There is really only one power button, but there are two positions.

As I understand, the power button is not reusable as a GPIO, so you
would have one less button for the user, but of course there are
several places where it can go. It would be simpler construction wise
if the power button was front facing like the others. It's partly a
question of aesthetics as to where to put it.

[Casainho]

On Sun, Jul 4, 2010 at 9:06 PM, Bob <bobcou...@googlemail.com> wrote:
> There is really only one power button, but there are two positions.
>
> As I understand, the power button is not reusable as a GPIO, so you
> would have one less button for the user, but of course there are
> several places where it can go. It would be simpler construction wise
> if the power button was front facing like the others. It's partly a
> question of aesthetics as to where to put it.

That power button is PSWITCH, right? -- must be used for power up the
imx233/system and can also be used as IO, like for play, pause, etc!
Maybe you could put a button the right cross to be the power up button.

And about PCB, I just printed today with my old Epson, a PCB. The ink
I have is water soluble and so it do not work as a mask for acid to
etch PCB. I will try find such ink on local stores. While then, I hope
to test the method for using a laser printer and hot iron.

A MID LEVEL PSWITCH state can be generated by connecting the VDDXTAL
output of the SOC to PSWITCH through a switch.

PSWITCH acts like a high impedance input (>300 kΩ) when the voltage
applied to it is less than 1.5V. However, above 1.5V it becomes lower
impedance. To simplify design, it is recommended that a 10 kΩ resistor
to VDDIO be applied to PSWITCH to set the HIGH LEVEL state.

32.4.1
Power On
When the PSWITCH pin voltage is higher than the minimum MID level (as
specified in Table 2-3., “Recommended Power Supply Operating
Conditions”) for >100 ms, the DC-DC converter begins its startup
routine. This is the primary method of starting the system via
PSWITCH. All products based on the i.MX23 must have a mechanism of
bringing PSWITCH high to power up via an always-present supply (e.g,
battery or VDDXTAL).

32.4.2
Power Down
If the PSWITCH pin voltage has a falling edge faster than 15 ns, then
this sends a power-down request to the DC-DC converter. The
fast-falling-edge power-down may be blocked by the
HW_POWER_RESET_PWD_OFF function. The fast-falling edge can also be
prevented by placing an
RC filter on the PSWITCH pin. Most i.MX23-based systems do not use the
PSWITCH fast-falling-edge power-down and include the RC filter to
prevent it from occurring accidentally.

32.4.3
Software Functions/Recovery Mode
When the PSWITCH pin voltage is pulled up to the minimum MID level (as
specified in Table 2-3., “Recommended Power Supply Operating
Conditions”) the lower HW_POWER_STS_PSWITCH bit is set. Software can
poll this bit and perform a function as desired by the product
designer. Example functions include a play/pause/power-down button,
delay for startup, etc.
When the PSWITCH pin is connected to VDDIO through a current limiting
resistor, the upper HW_POWER_STS_PSWITCH bit is also set. If this bit
is set for more than five seconds during ROM boot, the system executes
the Freescale USB Firmware Recovery function, as described in Chapter
35,
“Boot Modes.” If the product designer does not wish to use Freescale
USB Firmware Recovery, the product can be designed to not assert a
voltage higher than the maximum MID level (as specified in Table 2-3.,
“Recommended Power Supply Operating Conditions”) on the PSWITCH pin.
Refer to the Freescale i.MX23 reference schematics for example
configurations of the PSWITCH pin.

[Bob]

Are we trying to save 20p?

There are a couple of technical problems. The VDDXTAL is not on the
expansion header, so I would need some way of connecting that, but
more importantly I don't really understand enough electronics to do
it ;) If there was an example somewhere, I could copy it. I also
wanted to avoid duplicated components on the IO board, since there are
already the PSWITCH stuff on the main board, and seeing as I need some
extra wires anyway, the simplest option is to solder wires to each
side of the PSWITCH button on the main board.

There is also a practical issue which is avoiding the power button
being pressed accidentally. The way to do that is to have the power
button recessed or on a firmer tactile switch than is likely to be
operated accidentally, both of which mean the button is less usable as
a regular key.

What I could do is to make one switch selectable between PSWITCH and
GPIO, and let the user decide...

On 5 July, 19:47, Casainho <casai...@gmail.com> wrote:

[Opendous Inc.]

>I just printed today with my old Epson, a PCB.
>The ink I have is water soluble and so it do
>not work as a mask for acid to etch PCB

Did you bake the PCB after printing? Thanks for that link by the
way, the process seems promising and it may be possible to print
directly onto flexible laminates such as DuPont Pyralux (http://
www2.dupont.com/Pyralux/en_US/products/laminate/index.html) with any
cheap printer.

As for VDDXTAL and PSWITCH, should it be made available for
expansion?

[Casainho]

On Wed, Jul 7, 2010 at 6:28 AM, Opendous Inc.
<opendous...@gmail.com> wrote:
>>I just printed today with my old Epson, a PCB.
>>The ink I have is water soluble and so it do
>>not work as a mask for acid to etch PCB
>
>  Did you bake the PCB after printing?  Thanks for that link by the
> way, the process seems promising and it may be possible to print
> directly onto flexible laminates such as DuPont Pyralux (http://
> www2.dupont.com/Pyralux/en_US/products/laminate/index.html) with any
> cheap printer.

Ok, I got a print on PCB where the ink didn't erased with water or my
fingers press movement. It was Epson UltrBrite ink which is made of
some resin. I had to bake the PCB after print.

I had to buy a new Epson printer for 50€, because for my old one there
was no such UltraBrite ink.

This new printer is more complex, because it have scanner :-( -- and I
just printed one PCB as test but I couldn't print again because PCB
skips on roller for the Y axis, the one that pulls the paper.

One guy builded a table for Y that moves with a stepper motor, and
takes input from opto sensor of original Y motor. I wish I could do
that. That guy tried to make a 3D printer... I wish I could control
and put the Epson inkjet head on my RepRap Mendel...
http://www.indoor.flyer.co.uk/3dprint.htm

>  As for VDDXTAL and PSWITCH, should it be made available for
> expansion?

I don't know which one should be available. I think PSWITCH is
available on JTAG (for failed JTAG RESET). I think expansion board
must have a button for power on the system (the PSWITCH).

[Casainho]

>  Did you bake the PCB after printing?  Thanks for that link by the
> way, the process seems promising and it may be possible to print
> directly onto flexible laminates such as DuPont Pyralux (http://
> www2.dupont.com/Pyralux/en_US/products/laminate/index.html) with any
> cheap printer.

I were looking at ebay for such flexible PCBs... I couldn't found any
and I believe they are expensive...

Here some info of a guy that could control his Epson inkjet print
head: http://techref.massmind.org/techref/pcb/etch/custom-vs.htm

[Casainho]

On Mon, Jul 5, 2010 at 8:22 PM, Bob <bobcou...@googlemail.com> wrote:
> Are we trying to save 20p?
>
> There are a couple of technical problems. The VDDXTAL is not on the
> expansion header, so I would need some way of connecting that, but
> more importantly I don't really understand enough electronics to do
> it ;) If there was an example somewhere, I could copy it. I also
> wanted to avoid duplicated components on the IO board, since there are
> already the PSWITCH stuff on the main board, and seeing as I need some
> extra wires anyway, the simplest option is to solder wires to each
> side of the PSWITCH button on the main board.
>
> There is also a practical issue which is avoiding the power button
> being pressed accidentally. The way to do that is to have the power
> button recessed or on a firmer tactile switch than is likely to be
> operated accidentally, both of which mean the button is less usable as
> a regular key.
>
> What I could do is to make one switch selectable between PSWITCH and
> GPIO, and let the user decide...

My girlfriend cheap MP3 player have a central button that turns player
ON/PLAY/PAUSE e turn OFF. We could do the same with the same PSWITCH,
LCD, memory VDD, etc, would not power if user press the PSWITCH for at
least 1 second. The same for turn off.

I would like to simplify the number of buttons/board. I think many
buttons may get more confuse. Yeah, I asked for 2 crosses (4 + 4
buttons), but for now I would not assembly 2 buttons from the cross on
right side (the vertical ones). I would prefer to have PSWITCH on
first right side button (counting from left to right).

As for PCB, I quitted for now to try print using my actual Epson
printer, just because it can't feed correctly the PCB and so print
ends very bad. The ink really is strong against water or my finger
press movement, I didn't tested it with Ferric Chloride.
Someday I would like to build a PCB carriage moved by a stepper motor
and in sync with DC motor/encoder from the paper feeder - I need to
buy a slider from a drawer and try to build it...

Today I hope to get a borrow laser printer and try the "toner
transfer" technique - it will be my first time!

[Bob]

On 7 July, 06:28, "Opendous Inc." <opendous.supp...@gmail.com> wrote:

>   As for VDDXTAL and PSWITCH, should it be made available for
> expansion?

I think it would be a good idea.

When I tried etching PCBs, I got some reasonable results using photo
resist. I printed the design on transparent plastic using a postscript
laser printer. I found that the accuracy of ink jet printers
inadequate then, they may be better now. An error of only a few
percent was enough to mean a 40 way connector no longer fitted.
Getting the exposure and chemical treatment requires a bit of
practice, I never got good enough to achieve consistent results.

Then of course there is the problem of drilling, aligning two sided
boards, through hole plating, solder mask, silk screen, chemical
stains everywhere... I concluded that the cost of the equipment and
time was not worth the results. I guess it is different if you do a
lot of boards and can fine tune the process.

I wonder if using photo resist with a laser mounted on a X-Y table
would work.

[Casainho]

> When I tried etching PCBs, I got some reasonable results using photo
> resist. I printed the design on transparent plastic using a postscript
> laser printer. I found that the accuracy of ink jet printers
> inadequate then, they may be better now. An error of only a few
> percent was enough to mean a 40 way connector no longer fitted.
> Getting the exposure and chemical treatment requires a bit of
> practice, I never got good enough to achieve consistent results.
>
> Then of course there is the problem of drilling, aligning two sided
> boards, through hole plating, solder mask, silk screen, chemical
> stains everywhere... I concluded that the cost of the equipment and
> time was not worth the results. I guess it is different if you do a
> lot of boards and can fine tune the process.

I wish I had already invested on the technology for PCBs. Well, I
tried with the CNC I bought 1 year ago, but I ended up doing the 3D
printer with that. I even milled and drilled one board for Bob, for
the Mini2440 Lyre prototype.

Right now I have a laser/toner printer with me. I tried a few paper I
have here, from magazines, and I got a few problems. The printer
prints a lot of toner, with high quality! Today I will buy from
Staples "photo paper" which seems to be the best/cheaper for this.
I already bought online the "press-n-peel" paper, but it can take 2
weeks or more to arrive, that's why I prefer to do with local
resources and if possible at home.

If I can do the imx233 expansion boards with toner, then I may try to
drill them with CNC (or maybe for now I will do it by hand since I
don't have the CNC ready to work now).

[Opendous Inc.]

I have experimented with many papers and have created many PCBs
using the toner transfer method and Pulsar's Toner Transfer Paper
works best:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1003-ND

Considering the time you will spend, chemicals you will use, and
cost of PCBs the paper is not that expensive relative to your overall
costs.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1017-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1018-ND
http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_70201_-1

If you must try other papers the most important thing to remember is
heat+pressure. You can use a cheap clothes iron and any paper to
transfer toner but you have just about sit on the clothes iron for
several minutes to make sure you produce enough heat and pressure. A
laminator is very useful to save your strength:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1031-ND

Or, if you want some exercise then properly transferring toner is
great Isometric exercise. If your arms don't burn and your PCB is not
too hot to touch then your transfer will fail. If you do it properly
you will be able to remove the paper using a toothbrush.
http://en.wikipedia.org/wiki/Isometric_exercise

It helps to heat your Ferric Chloride when etching:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=GC392-ND

Make sure you use appropriate materials (polyethylene tank, vinyl
gloves) so that you do not dissolve the wrong things.
http://www.mgchemicals.com/products/416es.html

[Opendous Inc.]

http://www.pulsarprofx.com/PCB/a_Pages/1_Menu/overview.html
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1003-ND
http://www.coolcomponents.co.uk/catalog/product_info.php?cPath=48_51&products_id=320

Farnell sells Ferric Chloride in Europe:
http://uk.farnell.com/mega/141311/ferric-chloride-xtals-0-5kg/dp/141311
http://uk.farnell.com/cif/ar412/ferric-chloride-36c-2-5l-jerrycan/dp/4148009


On Jul 14, 3:24 pm, "Opendous Inc." <opendous.supp...@gmail.com>
wrote:

>   I have experimented with many papers and have created many PCBs
> using the toner transfer method and Pulsar's Toner Transfer Paper

> works best:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...

>
>   Considering the time you will spend, chemicals you will use, and
> cost of PCBs the paper is not that expensive relative to your overall

> costs.http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_7...

>
>   If you must try other papers the most important thing to remember is
> heat+pressure.  You can use a cheap clothes iron and any paper to
> transfer toner but you have just about sit on the clothes iron for
> several minutes to make sure you produce enough heat and pressure.  A

> laminator is very useful to save your strength:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...

>
>   Or, if you want some exercise then properly transferring toner is
> great Isometric exercise.  If your arms don't burn and your PCB is not
> too hot to touch then your transfer will fail.  If you do it properly

> you will be able to remove the paper using a toothbrush.http://en.wikipedia.org/wiki/Isometric_exercise
>
>   It helps to heat your Ferric Chloride when etching:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=GC39...

>
>   Make sure you use appropriate materials (polyethylene tank, vinyl

> gloves) so that you do not dissolve the wrong things.http://www.mgchemicals.com/products/416es.html

[Opendous Inc.]

Very Important! Do not pour used Ferric Chloride down a drain as it
may etch your pipes, not to mention poison your water.

http://www.mgchemicals.com/techsupport/ferric_faq.html


On Jul 14, 3:33 pm, "Opendous Inc." <opendous.supp...@gmail.com>
wrote:
> http://www.pulsarprofx.com/PCB/a_Pages/1_Menu/overview.htmlhttp://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...http://www.coolcomponents.co.uk/catalog/product_info.php?cPath=48_51&...
>
> Farnell sells Ferric Chloride in Europe:http://uk.farnell.com/mega/141311/ferric-chloride-xtals-0-5kg/dp/141311http://uk.farnell.com/cif/ar412/ferric-chloride-36c-2-5l-jerrycan/dp/...

>
> On Jul 14, 3:24 pm, "Opendous Inc." <opendous.supp...@gmail.com>
> wrote:
>
> >   I have experimented with many papers and have created many PCBs
> > using the toner transfer method and Pulsar's Toner Transfer Paper
> > works best:http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-...
>
> >   Considering the time you will spend, chemicals you will use, and
> > cost of PCBs the paper is not that expensive relative to your overall

> > costs.http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-......

[Casainho]

1st: I were trying to capture PWM signals on LPC2103, and I got a high latency on interrupt code, more than the 2us impulse signal I were trying to capture, so I decided to quit for now.

2nd: I received a specific paper I bought on Ebay for toner PCB printing, the paper looks very good for this application but I ended up to undestand that my printer prints to much toner. I am sending a picture os the tests I did, in one board we can see the phrase "PCB PRINT..." that I printed with Epson DuraBrite ink/printer. The problem is that I just could print one time, I couldn't replicate again.

3nd: There are people printing/drawing with a Etch Resist PCB Marker Pen, using RepRap Mendel: http://ultimachine.com/content/printing-circuit-boards-mendel (I am sending a picture).

Now I will go to market and try to find a 0.3 or 0.4mm marker pen and try print with my Mendel or even with my CNC... maybe I will get better results than with toner transfer...

[Opendous Support]

Some tips to make things better:

1) Print at the highest resolution and turn off any toner saving
features - you want MORE toner

2) Sand your PCB using 1000 grit sandpaper, then clean it with soap
and water, then clean it with either 99% Rubbing Alcohol (Isopropyl
Alcohol) or Acetone

3) Use Green TRF film - after transferring the toner repeat the heat
and pressure step with green TRF film which will fill in all the voids
in your transferred toner. Green TRF film will only stick to the
toner and not the copper.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=182-1021-ND

4) Most importantly, press harder with your clothes iron or send the
PCB through your laminator several more times

> --
> You received this message because you are subscribed to the Google Groups
> "rockboxplayer" group.
> To post to this group, send email to rockbo...@googlegroups.com.
> To unsubscribe from this group, send email to
> rockboxplaye...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/rockboxplayer?hl=en.
>

[Casainho]

I am still working (but slowly) on getting technology for PCB printing. Right now I found a very thin pen (0.1mm) of pigment ink (the same as Epson printer, I guess). The same pen is also available from STAEDTLER in 0.5mm. I bought the pen on staples: http://www.staedtler.ca/pigment_liner_ca.Staedtler

I already tested that pen/ink on PCB, and later dried on oven, it seems almost strong as the one Epson DuraBrite pigment ink.

Now I am printing a pen holder with a spring, for place it on my RepRap Mendel. I already have a program that takes a PNG image from PCB layers and generates the GCode for RepRap Mendel.

[Casainho]

I tried to use "cad.py" script to generate the path for plotting the PCB on copper board. I tested first on a paper and I found that "cad.py" generates bad optimization paths :-(

The best paths I know are generated by Skeinforge, for 3D printing... and so I decided to see if I could import in some way the PCB gerber to Skeinforge... and I ended up by using Gerberview + Inkscape + Blender until I got a 3D mesh on Blender :-) -- unfortunately, the final 3D mesh in manifold although Skeinforge looks to deal with it perfectly.

I am now testing Skeinforge path generation... it have a few profiles like for milling and extrusion (3D printing)...

Workflow for Gerber to 3D mesh:

- open the gerber file with Gerbview and export it to SVG file;
- open SVG file with Inkscape and export to PNG file with 600DPI;
- open the PNG file on Inkscape and apply the "Trace bitmap (SHIFT + ALT + B)", remove the bitmap image, select all nodes and do "Union CTRL + +)", save the file as SVG;
- import the SVG file on Blender. Hit Alt + C so Blender transforms curves to mesh. Enter edit mode, select all vertices, remove duplicated, exit edit mode and extrude in Z direction 0.5 units. Scale the final mesh on X and Y to correct values (measure on KiCad).

[Casainho]

I were able to mill one face of the PCB. I hope soon to mill the 2
faces of the PCB and mill it. I will post pictures in the end.

[Casainho]

I am having problems when trying to mill the PCBs. The bits I have
seems really bad and not cutting anymore... I bought them from Ebay,
cheap, from China.

Bob, are you interested in buying one PCB for you? -- maybe Matt can
buy it while buy another PCBs. I would like to buy at least one
boards, so I can finally have the hardware working, or I will never
finish this project.

I am not seeing any commercial interest/viability on this project. I
wish I could earning money by selling online this hardware.

If any off you guys need an enclosure for this board, just ask. My
printer is ready :-)

--
Cumprimentos,
Jorge Pinto

[Opendous Support]

it is summer and people have more important things in their lives.

Give the project a chance to mature before discounting it.