Omnikey PCB and Fullsize Customs (Round 2 opening soon)

[Polecat]

I am also an engineer, but I tend to prefer mm, but for keyboards it is a bitch using 19.05mm for spacing, so I have to use inches...

You are correct about the spacing, but I was thinking because keycaps go in increments of 0.25U, the spacing would also follow that, which is true for most boards, this is the first exception I have seen. So it was unexpected, but it does make sense when you think about the spacing in inches instead of units. Center to center makes sense, but for layout it is easier to see a spacing than a center to center measurement when just laying out the blocks. I use both but each one has its' pros and cons.

The switches are laid out now, so I just need the locations of the screw holes in relation to the center of some switch...as the board edge is not going to be a reliable thing to measure from. Just use the same switch for each to make my life easier, or don't...if you hate me, lol.

I should have said I'm an *old* engineer, and we were doing layouts with press-on pads and tape and an Xacto knife. Millimeters were only used for nitpicky optical stuff (and spark plugs).

Looking good so far! You might want to check the spacing on the insert and delete for the Plus versions. Those two have the stems centered on the caps, so those switches should be centered at 3/4U and 2-1/4U, not 1U and 2U.

Here's a couple old photos of a bare Gen3 PC board that might be helpful:

keyboards-f2/just-picked-up-an-omnikey- ... ey#p342941

It's interesting to see the alternate switch locations marked "Europe". I've never seen a Northgate with a "European" layout, but it's possible that some were produced.

I'll get you some measurements for the screw holes ASAP, and I'll dig out a Plus and Ultra to make sure there's nothing worrisome on those.

[Polecat]

Ok, I've come up with some measurements for the PC board mounting hole locations. In the horizontal direction I used the centerline of the Enter key switch, since that's consistent between all of the different versions. In the vertical direction I used the lower of the two pins on the Enter key since that was easier to use than the centerline. On the Gen1, 2, and 3 boards the Enter key is mounted with the pins towards the bottom. On the Gen4 (Avant) it's mounted with the pins towards the top. These measurements are designated "left", "right", "up", or "down" from the above reference points, looking at the solder (bottom) side of the PC board. Holes are numbered left to right.

Gen1:

Hole 1 - 3-5/16" left, 3/16" down.
Hole 2 - 5/8" right, 1-11/16" down.
Hole 3 - 1-5/8" right, 1-7/8" up.
Hole 4 - 7-15/16" right, 1-7/8" up.
Hole 5 - 8-15/16" right, 1-11/16" down.

Ground screw (soldered) is 3/4" to the left of Hole 1. Holes are 5/16". Screws attach the plate to the upper case through these holes.

PC board measures 19-3/16" by 6-1/16". (2-1/4 down, 3-15/16" up) Any bigger and it would ground out against the bottom cover.

Gen2:

Hole 1 - 3-5/16" left, 3/16" down. (same as Gen1 Hole 1)
Hole 2 - 1-5/8" right, 1-7/8" up. (same as Gen1 Hole 3)

Gen3:

Hole 1 - 6-1/2" left, 2-1/16" up.
Hole 2 - 3-15/16" left, 1-9/16" down.
Hole 3 - 1-11/16" right, 2-1/16" up.
Hole 4 - 9-7/16" right, 2-1/16" up.
Hole 5 - 10-5/8" right, 15/16" down.

PC board measures 19-1/8" by 6-13/16". Screws attach the PC board to the plate through these holes, all screws soldered in place.

Gen4:

Same as Gen3, except Enter switch is inverted (pins up), so the vertical reference point isn't the same. Gen3 and Gen4 plates are different, because of the bottom row being different, but mounting hole locations are the same. Gen4 PC board is smaller in the vertical direction, but case and plate clearance is the same as Gen3.

I hope that all makes sense. Please double check against the photos in case I messed something up!

[BlindAssassin111]

Gen1:

Hole 1 - 3-5/16" left, 3/16" down.
Hole 2 - 5/8" right, 1-11/16" down.
Hole 3 - 1-5/8" right, 1-7/8" up.
Hole 4 - 7-15/16" right, 1-7/8" up.
Hole 5 - 8-15/16" right, 1-11/16" down.

Gen3:

Hole 1 - 6-1/2" left, 2-1/16" up.
Hole 2 - 3-15/16" left, 1-9/16" down.
Hole 3 - 1-11/16" right, 2-1/16" up.
Hole 4 - 9-7/16" right, 2-1/16" up.
Hole 5 - 10-5/8" right, 15/16" down.

The above ones in red are not possible assuming the enter has pins on the bottom, as these points are off the bottom of the PCB. And after checking my Gen 3 Ultra T, my enter has pins on the top which gives the measurements a 3/8" bump upwards, making them sit back on the board like they should be.

Can you check the heights of these holes again, to correct the positions? I really appreciate the extensive measurements and help, it makes the project possible.

[Polecat]

Gen1:

Hole 1 - 3-5/16" left, 3/16" down.
Hole 2 - 5/8" right, 1-11/16" down.
Hole 3 - 1-5/8" right, 1-7/8" up.
Hole 4 - 7-15/16" right, 1-7/8" up.
Hole 5 - 8-15/16" right, 1-11/16" down.

Gen3:

Hole 1 - 6-1/2" left, 2-1/16" up.
Hole 2 - 3-15/16" left, 1-9/16" down.
Hole 3 - 1-11/16" right, 2-1/16" up.
Hole 4 - 9-7/16" right, 2-1/16" up.
Hole 5 - 10-5/8" right, 15/16" down.

The above ones in red are not possible assuming the enter has pins on the bottom, as these points are off the bottom of the PCB. And after checking my Gen 3 Ultra T, my enter has pins on the top which gives the measurements a 3/8" bump upwards, making them sit back on the board like they should be.

Can you check the heights of these holes again, to correct the positions? I really appreciate the extensive measurements and help, it makes the project possible.

Measurements are correct, description of vertical reference point is wrong. Turns out the Enter switches on all four are mounted with the pins towards the top. Sorry for my mixup.

[BlindAssassin111]

Measurements are correct, description of vertical reference point is wrong. Turns out the Enter switches on all four are mounted with the pins towards the top. Sorry for my mixup.

Sweet at least all I need to do is move them up the 3/8" to make them sit in the correct position. Thank you for checking!!! And don't worry about mixing it up, I figured it was a simple mistake that made them all sit a tad low.

[BlindAssassin111]

Okay the PCB is laid out except for the LEDs as I still have no idea how QMK controls the lock lights...Anyone know? Besides that I have finished the entire PCB and will upload pictures once the LEDs are done.


Also, Polecat can you check the dimensions on the Gen 1, Hole 1 again? According to your Gen 1 pcb picture, hole 1 is above the up arrow and the soldered screw is above the right arrow key. But the dimensions put hole 1 just past the right arrow key, and then the soldered screw inside the numpad 4 key. Unless I am understanding you reference point wrong. How I understand it is you use the center as the point midway between the two enter pins horizontally and then used the lower pin vertically, am I correct?

[Polecat]

Okay the PCB is laid out except for the LEDs as I still have no idea how QMK controls the lock lights...Anyone know? Besides that I have finished the entire PCB and will upload pictures once the LEDs are done.


Also, Polecat can you check the dimensions on the Gen 1, Hole 1 again? According to your Gen 1 pcb picture, hole 1 is above the up arrow and the soldered screw is above the right arrow key. But the dimensions put hole 1 just past the right arrow key, and then the soldered screw inside the numpad 4 key. Unless I am understanding you reference point wrong. How I understand it is you use the center as the point midway between the two enter pins horizontally and then used the lower pin vertically, am I correct?

Correct on the reference points. Looks like I didn't subtract an inch from the Hole1 measurement (I started at 1 inch on my steel rule), so it should be 2-5/16 inch left and 3/16 inch down. Sorry again, and glad you're double checking things.

[BlindAssassin111]

Correct on the reference points. Looks like I didn't subtract an inch from the Hole1 measurement (I started at 1 inch on my steel rule), so it should be 2-5/16 inch left and 3/16 inch down. Sorry again, and glad you're double checking things.

So is that the only one that you accidentally did this on? I want to make sure in case I missed something being wrong as well. Your help is making this possible!!!

[Polecat]

Correct on the reference points. Looks like I didn't subtract an inch from the Hole1 measurement (I started at 1 inch on my steel rule), so it should be 2-5/16 inch left and 3/16 inch down. Sorry again, and glad you're double checking things.

So is that the only one that you accidentally did this on? I want to make sure in case I missed something being wrong as well. Your help is making this possible!!!

Well I worked twelve hours today and it's 10 pm, so I don't have time to take all four keyboards back apart again to remeasure every hole, probably won't until the weekend. That's why I suggested checking against the photos. My crappy camera doesn't take good enough photos to take measurements from, but I have a 3200 dpi flatbed scanner that I can lay the PC boards on and scan them if that would be better. Again I probably won't have much time until the weekend, bills to pay and all that.

[BlindAssassin111]

Well I worked twelve hours today and it's 10 pm, so I don't have time to take all four keyboards back apart again to remeasure every hole, probably won't until the weekend. That's why I suggested checking against the photos. My crappy camera doesn't take good enough photos to take measurements from, but I have a 3200 dpi flatbed scanner that I can lay the PC boards on and scan them if that would be better. Again I probably won't have much time until the weekend, bills to pay and all that.

You are good, was more just curious, I can check using the pictures to see if there are any issues. It is midnight here so...pretty late.

[BlindAssassin111]

Polecat, quick question. How would you recommend doing the break-away section? Currently I just have 0.039 in holes spaced .078 in center to center along the entire height of the PCB. I have never done a break-away section before, and especially not one that needs to be somewhat strong on the break-away.

Breakoffs.PNG (8.61 KiB) Viewed 13548 times

EDIT: Can you also measure the spacing between the numpad and nav cluster for the Gen1? I measured it on the gen3 and it is an odd 1.09375 inch spacing and wanted to see if by chance it is different on the older ones.

I at one point had long slots and the same holes separating the slots, but went to all holes to add more material between the PCB halves. Do you think this would be fine, too strong, or too weak? I found an article about sizing for break-away tabs, and the hole sizing is almost the same, just tighter spacing.

Finished routing everything and even found an error on the gen3 PCB where the LEDs are spaced incorrectly. The spacing on the opening on the top cover are equal but on the PCB they weren't and one of the LEDs is bent a tad to fit in the cutout. Kinda funny actually.

Also I was throwing the idea around of making a small break-away section above the left side function keys, and under the badge door for the boards that have it) to put the reset switch to program the teensy. Sound like a good idea or skip it?

[Hak Foo]

You might look at how the GH-122 did it-- it was designed with breakaway parts.

https://geekhack.org/index.php?topic=80 ... msg2309842

[BlindAssassin111]

You might look at how the GH-122 did it-- it was designed with breakaway parts.

https://geekhack.org/index.php?topic=80 ... msg2309842

Hmm, so they only did holes at the at the ends of the cut...So I assume you would need to score the pcb and then snap it. I could do something similar.

[Polecat]

Polecat, quick question. How would you recommend doing the break-away section? Currently I just have 0.039 in holes spaced .078 in center to center along the entire height of the PCB. I have never done a break-away section before, and especially not one that needs to be somewhat strong on the break-away.

EDIT: Can you also measure the spacing between the numpad and nav cluster for the Gen1? I measured it on the gen3 and it is an odd 1.09375 inch spacing and wanted to see if by chance it is different on the older ones.

I at one point had long slots and the same holes separating the slots, but went to all holes to add more material between the PCB halves. Do you think this would be fine, too strong, or too weak? I found an article about sizing for break-away tabs, and the hole sizing is almost the same, just tighter spacing.

Finished routing everything and even found an error on the gen3 PCB where the LEDs are spaced incorrectly. The spacing on the opening on the top cover are equal but on the PCB they weren't and one of the LEDs is bent a tad to fit in the cutout. Kinda funny actually.

Also I was throwing the idea around of making a small break-away section above the left side function keys, and under the badge door for the boards that have it) to put the reset switch to program the teensy. Sound like a good idea or skip it?

I measure 1-1/8 inch center to center between the right nav column and the left numpad column on Gen1, 2, and 3. All the same in any case.

My opinion is that the "breakaway" section should be cut with a fine toothed saw or shear. The original PC board material from the '80s was a phenolic material, but most modern PC boards are made from fiberglass and resin, which is much stronger and would probably not break cleanly, even with a row of holes or slots. Even after scoring the fiberglass material might not break where it's supposed to, and I would hate to see someone ruin a brand new board trying. I would suggest a printed line with no traces too close, and let folks cut the board if used for a 101 version. That would also make it stronger on non-101 models.

I've never played with a Teensy, so I'm not the one to ask on that.

[BlindAssassin111]

I measure 1-1/8 inch center to center between the right nav column and the left numpad column on Gen1, 2, and 3. All the same in any case.

Hmm, I measured with calipers and got the 1.09375 measurement, I was using the pads even. But if they are the same and you used a different tool to measure, maybe they are all the 1.09375in?

[BlindAssassin111]

Update:
I have finished laying out the entire pcb and have gone through and checked all switch and LED positions. All I need is to have someone besides me and polecat measure the spacing between the clusters on their omnikey in order to get a larger pool of data for the dimensions. Once this is done, I can get a final quote, and start the GB.

What I need measured: (all are center to center or measured from the same point on two switches)

• - num lock to page up (mine is 1.09375 in)
  - insert to backspace (mine is 1.46875 in)
  - F2 to tilde (mine is 1.23475 in)

I will start designing the sandwich case as well for specific layouts, in order to allow people to make a custom omnikey board. But I can't do that until I get the aforementioned dimensions.

Thank you all for input and the massive amounts of information in order to make sure I get this done right.

[Polecat]

I measure 1-1/8 inch center to center between the right nav column and the left numpad column on Gen1, 2, and 3. All the same in any case.

Hmm, I measured with calipers and got the 1.09375 measurement, I was using the pads even. But if they are the same and you used a different tool to measure, maybe they are all the 1.09375in?

Just for curiosity, what did you use for a measurement between the nav cluster and the alpha section? Not center to center, because the keys in the alpha area are bigger than 1U, but the space between? That should be the same as the space between the nav cluster and num pad. And the same as the spaces between the F-key blocks (which are 1U) on Gen3 and 4 boards.

I'll try to remeasure everything later today or tomorrow.

[BlindAssassin111]

Just for curiosity, what did you use for a measurement between the nav cluster and the alpha section? Not center to center, because the keys in the alpha area are bigger than 1U, but the space between? That should be the same as the space between the nav cluster and num pad. And the same as the spaces between the F-key blocks (which are 1U) on Gen3 and 4 boards.

I'll try to remeasure everything later today or tomorrow.

I used SF12 and print screen, as well as backspace and insert. Since I looking for spacing of a whole cluster, it doesn't matter if one thing is 2U and the other 1U, as the pin to pin measurement still tells me the spacing I need. Only reason I brought up backspace is for those that have no function row to measure off of, so it is the only way they can help in that regard, so I am going to use that for all of them to get a consensus.

[Myoth]

What layouts are going to be supported ?

ISO would be very nice, please add it !!

[Polecat]

I used SF12 and print screen, as well as backspace and insert. Since I looking for spacing of a whole cluster, it doesn't matter if one thing is 2U and the other 1U, as the pin to pin measurement still tells me the spacing I need. Only reason I brought up backspace is for those that have no function row to measure off of, so it is the only way they can help in that regard, so I am going to use that for all of them to get a consensus.

I'm not making myself clear, as usual. What I meant was if your measurement, say between F10 and PrintScreen, was 1.09375 inch, or 1-1/8 inch? It should match the measurement between the right column of the nav cluster and the left column of the num pad. (you can't use pin to pin on the backspace...unless you correct for the 2U keycap of course...)

[BlindAssassin111]

What layouts are going to be supported ?
ISO would be very nice, please add it !!

I was never planning on doing ISO, as I have never seen an ISO omnikey. For ISO I would need two switches on backspace and the weird enter correct? I will give it a thought but idk...

[BlindAssassin111]

I'm not making myself clear, as usual. What I meant was if your measurement, say between F10 and PrintScreen, was 1.09375 inch, or 1-1/8 inch? It should match the measurement between the right column of the nav cluster and the left column of the num pad. (you can't use pin to pin on the backspace...unless you correct for the 2U keycap of course...)

I think I am still misunderstanding you...But will see if this helps at all.

I was getting a smaller measurement than what you had said you got, mine was 1.09375" and yours 1.125", I was getting the same value between nav and alphas, as I was between nav and num. The backspace measurement I am correcting to if it were two 1U keys next to each other by subtracting 0.375". And if you add 0.375" to my 1.09375", you get the same as the backspace I measured, 1.46875". So I just wanted someone else to give input to what it actually could be.

[Myoth]

What layouts are going to be supported ?
ISO would be very nice, please add it !!

I was never planning on doing ISO, as I have never seen an ISO omnikey. For ISO I would need two switches on backspace and the weird enter correct? I will give it a thought but idk...

https://prnt.sc/imhvys

this wouldn't be hard to implement and I think you would actually bring a lot of people into this project there IS no ISO on omnikeys, I'm searching for one of those sweet keyboard myself and I would actually love to find one, and get an additional pcb to make mine in ISO.

[BlindAssassin111]

https://prnt.sc/imhvys

this wouldn't be hard to implement and I think you would actually bring a lot of people into this project there IS no ISO on omnikeys, I'm searching for one of those sweet keyboard myself and I would actually love to find one, and get an additional pcb to make mine in ISO.

Hmm, the only issue is I would have to re route an entire row to accomadate the "\|" key, I don't have room to add one without doing that.

EDIT: Nevermind found a spot in the matrix that would be possible.

[Polecat]

I think I am still misunderstanding you...But will see if this helps at all.

I was getting a smaller measurement than what you had said you got, mine was 1.09375" and yours 1.125", I was getting the same value between nav and alphas, as I was between nav and num. The backspace measurement I am correcting to if it were two 1U keys next to each other by subtracting 0.375". And if you add 0.375" to my 1.09375", you get the same as the backspace I measured, 1.46875". So I just wanted someone else to give input to what it actually could be.

Ok, that's 1/32 inch smaller on both measurements, now I know what to look for.

[Polecat]

I was never planning on doing ISO, as I have never seen an ISO omnikey. For ISO I would need two switches on backspace and the weird enter correct? I will give it a thought but idk...

Typically ISO has an extra 1U key next to the Enter, and a shorter left shift with an extra 1U key to the right of it. Sometimes called a "102" layout. Exact key designations depend on the country and specific layout. The plate would need to match, of course. I have an ISO Focus board I can check to see if the switch for the Enter would need to be relocated to use a Focus key cap for the Enter. It may vary depending on the Enter keycap being used. For the PC board the (smaller) left shift would need to move to the left and the two 1U switches added. Backspace is usually unchanged at 2U, but there may be exceptions. In summary, the PC board, plate, and keycaps would all need to match to make this work.

[BlindAssassin111]

Typically ISO has an extra 1U key next to the Enter, and a shorter left shift with an extra 1U key to the right of it. Sometimes called a "102" layout. Exact key designations depend on the country and specific layout. The plate would need to match, of course. I have an ISO Focus board I can check to see if the switch for the Enter would need to be relocated to use a Focus key cap for the Enter. It may vary depending on the Enter keycap being used. For the PC board the (smaller) left shift would need to move to the left and the two 1U switches added. Backspace is usually unchanged at 2U, but there may be exceptions. In summary, the PC board, plate, and keycaps would all need to match to make this work.

Yeah, the first thing that came to mind when I read ISO was the weird boards with the split backspace. I thought that was the standard, but apparently not.

Now something I need to figure out is what orientation for the ISO enter? rotated or non-rotated or both? Hard to know as SP alps caps are non-rotated, where as older sets generally are rotated, and supporting both would be hard for trace layout reasons.

EDIT: Managed to get both rotated and non-rotated ISO enter on the board, traces look kinda funky, but was much easier than I expected actually.

ISO_Enter.PNG (37.11 KiB) Viewed 13444 times

Updated the ISO enter again, was off a tad on placement, treated it as 1U instead of 1.25U like I should have.

[Polecat]

Yeah, the first thing that came to mind when I read ISO was the weird boards with the split backspace. I thought that was the standard, but apparently not.

Now something I need to figure out is what orientation for the ISO enter? rotated or non-rotated or both? Hard to know as SP alps caps are non-rotated, where as older sets generally are rotated, and supporting both would be hard for trace layout reasons.

EDIT: Managed to get both rotated and non-rotated ISO enter on the board, traces look kinda funky, but was much easier than I expected actually.

Quick peek shows my Focus FK-2002 has the rotated switch. That's a likely key cap donor since (we think) Focus and Northgate were made by the same manufacturer. The PC board has pads for rotated and non-rotated, presumably for the BAE on the FK-2001 version. Note the extra pads for the small right shift key with the 1U switch next to it. The left shift is done the same way.

Edit - just saw your reply with the ISO Enter added, you managed that more quickly than I could take two photos. Must be getting old.

[BlindAssassin111]

Just want to throw this out there...The number of combinations for the setup of this board is ridiculous.

There are:

• - With/Without Left function keys
  - BAE,ANSI, and ISO enters
  - 3 Different right side bottom rows (4 x 1U, 2 x 1.5U with 1U, and 2 x 1.5U with a gap)
  - Split and long right shift
  - Split numpad plus
  - T-nav, Ultra nav with bottom 2 x 1.5U, Ultra nav with bottom 3 x 1U

Totaling: 216 different layout combinations (288 if you replace the left, right and down arrow with 2 x 1.5U keys, if you really wanted to)

And because alps don't support multi-layout plates as nicely as cherry switches, there are a boat load of plate layouts...144 to be exact. So I am not sure what to do about designing the plates as there is no logical way to do this, and swillkb can't build to the weird cluster spacing these keyboards have.

EDIT: I even started to use keyboard-layout-editor to get the layouts started, then realized there are more than I originally thought, only managed to do 12 layouts before doing the above math, needless to say I gave up.

[//gainsborough]

Just want to throw this out there...The number of combinations for the setup of this board is ridiculous.

There are:

• - With/Without Left function keys
  - BAE,ANSI, and ISO enters
  - 3 Different right side bottom rows (4 x 1U, 2 x 1.5U with 1U, and 2 x 1.5U with a gap)
  - Split and long right shift
  - Split numpad plus
  - T-nav, Ultra nav with bottom 2 x 1.5U, Ultra nav with bottom 3 x 1U

Totaling: 216 different layout combinations (288 if you replace the left, right and down arrow with 2 x 1.5U keys, if you really wanted to)

And because alps don't support multi-layout plates as nicely as cherry switches, there are a boat load of plate layouts...144 to be exact. So I am not sure what to do about designing the plates as there is no logical way to do this, and swillkb can't build to the weird cluster spacing these keyboards have.

EDIT: I even started to use keyboard-layout-editor to get the layouts started, then realized there are more than I originally thought, only managed to do 12 layouts before doing the above math, needless to say I gave up.

I'll have to check my plate on the gold label 102 or the PLUS, but I think that plate could support an ANSI enter over BAE, you'd just have to remove the stabilizer-insert. It has the same dimensions as an alps switch!