Phantom Custom Keyboard Group Buy (CLOSED)

[bpiphany]

gimpster, does your problem occur on all columns? If it is just the J/K one you may have a short circuit between the columns.

[gimpster]

gimpster, does your problem occur on all columns? If it is just the J/K one you may have a short circuit between the columns.

As far as I have tested it appears to be all columns. I have not rigorously tested this, but I could make it happen on Z, X, M, N, ~, Q, R, etc. Up Arrow seemed okay from what I remember, but again, I didn't put it through a rigorous test matrix, I figured there was enough ways to test for a fix with what I had already discovered. Plus, until I actually get my switches then the only way I have to activate the keys is by shorting the pins with a scrap of wire.

[damorgue]

I was thinking of those small surface mounted once which you need a small sharp object to turn. I would need to open the case, screw them until I get the desired color, and close the case again, inc ase I want to switch color to go with different caps for instance.

Where are you going to solder the three LED legs? Are we talking about the stand-alone-LEDs now or the switch mounted? I made sure to add ground planes to make any modding harder

We'll see where i can place them once I get one. I want about three Phantoms but have only scraped together barely enough parts for one.

[litster]

Good news. BiNi has already shipped most of the parts 7bit needed to him. Some non-EU orders are going out today and will continue to go out as they become ready to ship. Please don't ask when yours will be shipped. There is only BiNi handling all the processing so please let him work on them as he can. Thanks.

[gimpster]

I was thinking of those small surface mounted once which you need a small sharp object to turn. I would need to open the case, screw them until I get the desired color, and close the case again, inc ase I want to switch color to go with different caps for instance.

Where are you going to solder the three LED legs? Are we talking about the stand-alone-LEDs now or the switch mounted? I made sure to add ground planes to make any modding harder

We'll see where i can place them once I get one. I want about three Phantoms but have only scraped together barely enough parts for one.

One thing that might be an issue is sizing. I haven't seen any multi-color LEDs in a 3mm size and I think early on they suggested that 5mm was too big to fit into a hole for the case...might be ways around that, though, depending on the design.

Good news. BiNi has already shipped most of the parts 7bit needed to him. Some non-EU orders are going out today and will continue to go out as they become ready to ship. Please don't ask when yours will be shipped. There is only BiNi handling all the processing so please let him work on them as he can. Thanks.

Yay!!

[gimpster]

Icarium has some parts up for sale for those looking:

http://deskthority.net/marketplace-f11/ ... t3352.html

[damorgue]

Icarium has some parts up for sale for those looking:

http://deskthority.net/marketplace-f11/ ... t3352.html

Not enough, I need more parts. If anyone reads this, feel free to PM me if you happen to have any to spare.

[MagicMeatball]

Still looking for a PCB. Hit me with a PM if you happen to have spares.

[damorgue]

One thing that might be an issue is sizing. I haven't seen any multi-color LEDs in a 3mm size and I think early on they suggested that 5mm was too big to fit into a hole for the case...might be ways around that, though, depending on the design.

You are quite correct. The only source that I hve been able to find is this:
http://www.alibaba.com/product-gs/46585 ... DIODE.html
and they will probably not sell them to me in these low numbers. Perhaps I can get a 5mm and enlarge the hole a bit, or get a very tiny smd one.

[7bit]

I'm looking at the source code for the Phantom and can't find out where the functions (usb_keyboard_press for example) are called from. Does anybody know where I've got to look?

Also, I would like to program some buttons as mouse buttons. I need to know which codes to send in that case.

Are the gaps in the USB keycode table (usb_keyboard.h) reserved in the USB keyboard standard, or are they just unused and can be used for special keys?

[bpiphany]

Are you looking at the code on github?

Gaps? I don't remember any gaps in that table of key codes. I added the ones PJRC didn't have from some USB HID specification.

I don't think usb_keyboard_press() is used at all. It is as simple as that =) usb_keyboard_send() is the function to use for sending a whole HID package. usb_keyboard_press() sends a package with one key code only. Releases it immediately, as well as releases any other keys that happened to be in the queue afterwards =P I say, just don't use it...

A mouse is probably some other HID device, and needs more code to be added. You'll need to ask someone of the guys having implemented this stuff before. The Teensy should be able to report as both a mouse and a keyboard at the same time I think.

[7bit]

I've downloaded your latest code.

So if it is not in use I will throw it out. The idea I've got in mind is to add further modifier keys and implement them at keyboard level, so I need as many keycodes as possible.

Here are the gaps:

Code: Select all

#define KEY_RESERVED1    0-3

#define KEY_A            4
#define KEY_B            5
...
#define KEY_CLEAR_AGAIN         162
#define KEY_CRSEL               163
#define KEY_EXSEL               164

#define KEY_RESERVED2           165-175

#define KEYPAD_00               176
#define KEYPAD_000              177
...
#define KEYPAD_OCTAL            219
#define KEYPAD_DECIMAL          220
#define KEYPAD_HEXADECIMAL      221

#define KEY_RESERVED3           222-223

#define KEY_LEFTCONTROL         224
#define KEY_LEFTSHIFT           225
...
#define KEY_RIGHTALT            230
#define KEY_RIGHTGUI            231

#define KEY_RESERVED4           232-65535

[bpiphany]

The firmware keeps track of which switch IDs that are pressed, not the key codes. They are translated on sending. You probably don't want to start sending internal codes across to the OS =)

I don't think there is really any reason to throw out unused code. It should be trimmed off by the compiler anyway, no? I've just kept the usb_keyboard.c untouched. I added the extra key codes to usb_keyboard.h, but that is all.

Edit: There is mouse code on the PJRC pages. I haven't looked at it at all though.

[7bit]

The firmware keeps track of which switch IDs that are pressed, not the key codes. They are translated on sending. You probably don't want to start sending internal codes across to the OS =)

Call them USB scan codes!


As far as I understand these are sent to the computer, so the quastion is are these reserved for special purposes or are the just unused and can be used for additional stuff.

I don't think there is really any reason to throw out unused code. It should be trimmed off by the compiler anyway, no? I've just kept the usb_keyboard.c untouched. I added the extra key codes to usb_keyboard.h, but that is all.

Unused code irritates me. What is it good for? Keep it in your version, but I throw it out, knowing now it does not do anything.

Edit: There is mouse code on the PJRC pages. I haven't looked at it at all though.

Yes, I've found that but hoped someone else (with access to a Teensy ) can tell me what to do ...

[maoiste]

You should take a look into the source of the doxkb he has done mousebuttons and other mouse stuff and has multiple layers:
https://github.com/doxkb/tmk_keyboard

[bpiphany]

I think it was from this document I took the "scan codes" page 53

But again. You only need to keep track of 'pressed[mu_mod_key]' is set or not. No need for a scan code, since you are not sending them to the OS.

[litster]

This is the original source.

ANSI150-Source.zip

(67.31 KiB) Downloaded 123 times

[bpiphany]

I really think the code on github is better in every regard. I don't know what gimpster's problems arises from though. Lit, you should try it out. And tomorrow I will test it on a Windows machine, and perhaps a Mac if I can find one... But that will be another, slightly different, Teensy keyboard hardware.

[litster]

Just finished building my second Phantom. I downloaded the new firmware source code from github, built it, and downloaded to the Phantom. Everything seems to be working correctly. I am running Windows 7.

[gimpster]

Sorry for the late follow-up on the firmware issue. I did some more testing with my VMs. I found that Windows 7 works just fine so far as I am able to easily test right now. However, my Ubuntu 12.04 VM went nuts. Random characters that were nowhere even near the key being activated, even managed to start another app. So, I'm still sticking with my original theory so far that the Mac just needs to be handled specially to play nice with the special things it wants to do with keyboard input. I'll work on it and keep the repository updated as I make changes.

I did finish prepping the mounting plates, though. I had to take two tries to get the metallic paint to come out looking decent. One of them is going for a purple theme (with purple LEDs), for the wife, and the other is going for a blue theme. The first attempt did not create a very metallic finish at all, it was crap. But I did get to test the fitment of the Costar stabilizers on it and they still were able to fit into the cut-outs without any trouble. The second attempt at painting came out much better, the blue especially. Not sure why the purple doesn't sparkle as much but it'll still look good.

IMG_0978.jpg (360.59 KiB) Viewed 4661 times

IMG_0979.jpg (443.95 KiB) Viewed 4661 times

IMG_0980.jpg (442.25 KiB) Viewed 4661 times

IMG_0982.jpg (330.22 KiB) Viewed 4661 times

IMG_0983.jpg (364.95 KiB) Viewed 4661 times

IMG_0984.jpg (440.22 KiB) Viewed 4661 times

[dirge]

That looks great! I may do pink

[gimpster]

Latest update on this. I still don't have my switches from Bini, and I've decided that I want to use Cherry stabilizers on these boards, so yesterday I desoldered 3 Wyse boards to salvage parts. Today, I sat down to start working on the firmware to make it play nice with the Mac and I decided to put in a couple of the Cherry black switches that I salvaged, to make testing easier. After putting those in, suddenly the keyboard works fine. I've tested it in OS X, Windows 7, and Ubuntu 12.04. All of the keys that have an actual switch connected (and not activated by just shorting a piece of metal between the contacts) works beautifully. So, on the plus side, there's nothing for me to "fix" right now. On the down side, I'm tantalizingly close to having a fully-functional Phantom (albeit, without a case to mount it in) but I can't complete it because I don't have my switches yet. Cherry blacks aren't bad, but they're not my favorite, and they won't stay in the keyboard permanently. I'm tempted to desolder one of my Dolch boards and put the blues in...

[7bit]

I don't understand this.

How can you have the plate and PCB and I have nothing?

[gimpster]

I don't understand this.

How can you have the plate and PCB and I have nothing?

Because apparently Bini's house is too far from the post office. I only have these because I drove 4 hours to his house and picked them up.

[7bit]

You should have picked them all!

The switches must be at Bini's site, I'm quite sure. I will check tomorrow.

7bit.

[gimpster]

You should have picked them all!

If only I had known...

[gimpster]

Well, my wife is impatient, and since BiNiaRiS is hand-delivering the key caps... She gave the Cherry black switches a try and it turns out she rather likes them, so my work spent desoldering those Wyse boards yesterday was more useful than originally thought. Plus, I got myself a new sculpture.



I still don't have a case, but her keyboard is now completely assembled and fully functional. I need a 7x spacebar for her and a proper set of keycaps for the 1.5x mods, but at least it's usable. She wanted both styles of LEDs, so we put blue ones under the keys, Leopold-style, and violet ones in the frame, Filco-style. I couldn't get a good picture of the violet LEDs, but I used these.

[bpiphany]

Looking good =) If this doesn't spring to life soon you will have the most expensive keyboard in history, adding up what everyone has paid for it =D

It still would be interesting to know what you could possibly have done to get that erratic behaviour you experienced before. I didn't get it with my breadboard set-up, which should be pretty equally prototypic as your PCB without switches =P

[7bit]

Looks nice! You should have ordered the 7BIT layout.

[gimpster]

Looking good =) If this doesn't spring to life soon you will have the most expensive keyboard in history, adding up what everyone has paid for it =D

It still would be interesting to know what you could possibly have done to get that erratic behaviour you experienced before. I didn't get it with my breadboard set-up, which should be pretty equally prototypic as your PCB without switches =P

Yah, it is pretty baffling. As far as I know, the controller doesn't care about line voltage fluctuations so any way you short the circuit should trigger the same behavior...I'm just glad it's working now.

The last update for US folks, at least, is that BiNiaRiS is in a holding pattern awaiting key switches to either clear customs or complete re-transport. Everything else sounds like it's sorted and ready to go. So, hopefully they'll start shipping out Real Soon Now™

Looks nice! You should have ordered the 7BIT layout.

Thanks! I really did give the 7bit layout a hard ponder, but I couldn't think of anything "cool" to do with the extra switches, and at the end of the day, my biggest goal is just to be able to put a complete set of vintage Cherry keys on the thing. It's amazingly expensive to be able to do that.