HyperMicro madness

User avatar
matt3o
-[°_°]-

20 Dec 2014, 11:37

can't you simply use D6 on a column instead of a row? I believe Hasu said it could work on col (give me 5 mins I'll try to find the discussion back)

User avatar
7bit

20 Dec 2014, 11:41

D6 is pin 12!

Look how many tracks go off at the top. One pad less and the whole thing becomes quite a mess.
:cry:
Attachments
HyperMicro_800.png
HyperMicro_800.png (39.35 KiB) Viewed 5546 times

User avatar
matt3o
-[°_°]-

20 Dec 2014, 11:52

blobbing the resistor is fine by me :P

User avatar
suka
frobiac

20 Dec 2014, 12:03

D6 should be just fine as it is if you put the diodes correctly (cathode on square pad) and strobe the columns in your firmware.
Regarding the mouse buttons I am not sure the Omron switches supplied with the PCB are really easy to interface due to lack of space in front of the lowest row. Front-facing ones could be easier to access, but I have no idea if there are any lighter ones than the typical square pushbuttons.

User avatar
7bit

20 Dec 2014, 12:16

Not sure what you mean.
:?

There are 2 rows of solder-pads for the mouse switches (Cherry DG and compatible, like the Omrons you got from me).

The front row is to upright mounting and the back row is for 90 degrees mounting which requires the variant with 90 deg bent pins.
:-)

User avatar
matt3o
-[°_°]-

20 Dec 2014, 12:19

I would definitely go with the 90deg!

User avatar
suka
frobiac

20 Dec 2014, 12:24

I also prefer 90° mounted switches, but the ones I know require quite some force and I don't know whether there are lighter ones available.

Could work though without moving the board around on the table, if there is enough friction from good rubber feet I assume

User avatar
7bit

20 Dec 2014, 12:30

These switches exist in both variants, I just have to find them ...

User avatar
ماء

20 Dec 2014, 13:17

looks dual sgg 8-)

User avatar
suka
frobiac

20 Dec 2014, 16:20

Here are three alternatives for the mouse buttons I found at a local shop:
Left one mounts well, but cap will not clear the housing when pressed completely. Plus it feels cheap...
Middle one is the standard type I thought of earlier, needs to be bent a little to fit the PCB and also have some of the back mount stabilizers trimmed off to fit in front of the switches.
Right is a microswitch by Cherry that sits firm and nicely but needs to have its left leg cut off in the v0.0 pcb version. Feels best and should be easy to interface with the lever, but it also protudes over the edge.

So for now I think I'll go with the middle one...
(Conrad PartIds: 707759 1.29€ / 704849 0.45€ / 704083 2.09€ if you're interested)
HyperMicro_MouseSwitches.jpg
HyperMicro_MouseSwitches.jpg (170.64 KiB) Viewed 5486 times

User avatar
7bit

20 Dec 2014, 16:50

How much further back, could that 3rd switch be mounted?
:?

Also:
Align the holes in the PCB with those in the switch?
:o

User avatar
suka
frobiac

20 Dec 2014, 17:08

Too late to check since it's already mounted now :-) But there is not really any more space between its pins and the first MX casing to utilize.
The left and right ones show above fit perfectly, the middle one I now used everywhere needs to have the contact pins bent to the front but that works easy enough. And only the middle location needs some metal removed on the switch back, the others can be mounted right in the gaps between any MX switch cases.

User avatar
7bit

20 Dec 2014, 17:49

OK, no problem.

All those measures are from the documenttaions and nothing is measured from real switches.

After all, great to see that things fit, right from the beginning!
:ugeek:

BTW:
Does anybody know a plate manufacturer here in Europe?

neverused

20 Dec 2014, 18:35

matt3o wrote:blobbing the resistor is fine by me :P
Since the resistor is there to limit the voltage supplied to the LED, wouldn't you run the risk of blowing it if the resistor is bypassed only?

Perhaps removing and blobbing both would be best?

User avatar
chzel

20 Dec 2014, 18:42

neverused wrote:
matt3o wrote:blobbing the resistor is fine by me :P
Since the resistor is there to limit the voltage supplied to the LED, wouldn't you run the risk of blowing it if the resistor is bypassed only?

Perhaps removing and blobbing both would be best?
If both of you are referring to my comment about a blob of solder, I mean that because the resistor is so small, you can make a blob of solder to heat up both pads simultaneously and easily remove the resistor, not short it.
Since the resistor is in series with the led, removing one of them is enough!

neverused

20 Dec 2014, 18:45

Ok yes removal would be the way to go. I interpreted what I read as implying a short, but I agree that part of the circuit should be left open to use D6.

User avatar
beltet

22 Dec 2014, 00:00

OK, I haven't read up on the hypermicro because I haven't been interested in 40'ish KB until recently. Can you sum this up? What is the main features? Track point? Mouse buttons?

User avatar
Muirium
µ

22 Dec 2014, 00:17

Yes: trackpoint and mouse buttons! It's got an optional pointing device, if you have the parts.

Other than that, it's a Teensy 2 powered highly configurable 12x4 unit canvas with a wide variety of compatible layouts built in. (Plenty of switch footprints.) PCB mount MX switches are advised, and there's footprints for optional Cherry PCB mount stabs. No Alps / Matias support in this version.

Image
http://deskthority.net/group-buys-f50/h ... t4185.html

And because 7bit sent one to Suka, there may be a nice case in the works for it!

User avatar
suka
frobiac

23 Dec 2014, 12:47

The latest case iteration is complete: it now features additional supports, reset button access, mouse button cutouts, a trackpoint area, rubber feet cavities as well a customizable cable/usb connector outlet and is still fully modular, so different case heights, switch layouts or rotations of the PCB are simple to incorporate.

Still missing:
- optional covers at the sides for layout options without the caps filling the whole rectangular area
- modular mouse button assembly
- modular usb/cable assembly
- separate led mounts
- and of course all the cosmetic stuff like bevels, logos and whatnot everybody comes up when prototyping
hm_case_1.png
hm_case_1.png (17.21 KiB) Viewed 5325 times
I hope to get an actual sample early next year to confirm my measurements and provide a base for more decorative case options. Pricing should be in the 50-60€ range if 3D-printed.

User avatar
matt3o
-[°_°]-

23 Dec 2014, 13:11

3d printed?

you sure just €50?

User avatar
suka
frobiac

23 Dec 2014, 13:12

That's what the shapeways quote is! Was amazed myself, to be honest...

User avatar
Muirium
µ

23 Dec 2014, 13:27

That's for the usual white material? How good is that stuff for painting, etc. for a finish?

Just out of curiosity, I wonder how much a metal one would cost. Or if that's as bad as I think, then CNCed wood.

Oh yeah: still nothing in the mail. Tomorrow's the last chance this year, most likely!

User avatar
matt3o
-[°_°]-

23 Dec 2014, 13:46

the "white material" can be sanded and painted. Just give a good base of primer first. It can be both an enamel or acrylic based primer (I believe Americans call it "gesso", to be added to the Italian words weirdly used in English).

If the 3d print is too rough you can also use some spray filler first.

User avatar
suka
frobiac

23 Dec 2014, 13:52

Muirium wrote:That's for the usual white material? How good is that stuff for painting, etc. for a finish?

Just out of curiosity, I wonder how much a metal one would cost.
Yes, standard white or black material. Normally a little rough, but as Matt3o pointed out it can be post-processed quite nicely.

Steel varieties were around 280€ <cough>

User avatar
Muirium
µ

23 Dec 2014, 14:02

Bugger!

3D print material is okay but not great. Definitely lighter than I like in a case. But better than no case at all, of course. I'll keep a close eye on your progress.

User avatar
7bit

23 Dec 2014, 14:04

Maybe aluminium is cheaper?
:?

Or just use a metal cookie box!
:evilgeek:

User avatar
Muirium
µ

23 Dec 2014, 14:07

Or perhaps better one of these:

http://deskthority.net/keyboards-f2/epiphany-t9366.html

Tins are too flimsy for a keyboard. They flex far too much, and are way too light besides.

User avatar
Khers

23 Dec 2014, 15:08

Yet another way of improving the finish of 3D-printed stuff is to have it exposed to acetone vapour which melts the surface of the object and smoothens it a bit. I've done this on a couple of things I've used for work and they have turned out pretty nice, mind you that was on stuff written in ABS on our MakerBot. I have no experience using Shapeways, but they are probably at least an order of magnitude better.

I would also like to add a disclaimer: don't do this if you don't know what you're doing. Acetone is not particularly nice to breathe, I always do this work in a fume hood. Also, given a flash point of -20 degC, heating acetone can easily go wrong.

User avatar
suka
frobiac

23 Dec 2014, 15:35

Acetone has no effect on PA, nor does any other readily available chemical I am aware of.

User avatar
Khers

23 Dec 2014, 16:04

Ok, didn't know what type of plastics Shapeways use.

Post Reply

Return to “Workshop”