Beam Spring 104+SSK Reproduction Project! First Batch In Stock, Shipping early next year after New Model F Project

[Ellipse]

Yep, you can update an existing order (it's $100 extra), as of now. Please email me for details.

[tiltowait]

Just to make sure: R2 is still on track to have USB-C, right? I don't see any advantage to USB-B.

[Ellipse]

Yes confirmed, all boards will be USB-C, but with the exciting possibility of a PS/2 add on board that forum member Rico is currently working on for the Leyden Jar controller!

The boards are still expected to go out later this year. The new beam spring mold has been approved and today the factory was approved to make a few thousand production parts before they make all several hundred thousand parts, just to make sure everything is good.

[BuGless]

I don't see any advantage to USB-B.

Robustness, perhaps?
The keyboards are meant to last decades. A USB-B PCB-socket is likely to survive that timeframe, a USB-C PCB-socket is likely to end up damaged in that same timeframe.

[genericusername57]

Yep, you can update an existing order (it's $100 extra), as of now. Please email me for details.

Cool, thanks for the answer. I'll give it a bit of a think and I'll write if I decide to go for it. Does that $100 include beamspring modules for the extra spaces you're creating?

[caesarcaliyugula]

Robustness, perhaps?
The keyboards are meant to last decades. A USB-B PCB-socket is likely to survive that timeframe, a USB-C PCB-socket is likely to end up damaged in that same timeframe.

That is true, but a similar durability can be achieved with aviator cables on USB-C, whereas the ecosystem compatability of USB-C cannot be achieved with B.

[tiltowait]

I don't see that a USB-B connector is necessarily more durable than a USB-C connector. It's also very atypical to use USB-B with a keyboard; I don't think I ever saw it outside of keyboards, scanners, and maybe a Zip drive.

Anyway, it seems the point is moot if it's still coming with USB-C. Looking forward to receiving mine.

[Muirium]

USB B is rated for less insertions than Micro, let alone C. It’s obsolescent at best. As well as bulky ugly.

I imagine the weak point with C is on the PCB hookup side. It’s small and fiddly.

[ritter]

I'm really looking forward to my beam spring... I'm not sure the USB question is a big deal for me personally, how often do you plug and unplug your keyboard? In any case, it's not a fast device, so the physical standard shouldn't matter too much; you can always put an adaptor on it if needed.

[BuGless]

That is true, but a similar durability can be achieved with aviator cables on USB-C,

The problem is not the cable, that is easily replaced. The problem is the demolition done to the PCB mounted USB socket on the PCB.

To be exact, violent janks at the cable perpendicular to the socket, that's what we're talking about here.
For a USB-B socket, what gets damaged the most is the cable, the socket is sturdily through-hole PCB soldered and has a lot of guiding plastics to absorb most of the shock.
For a USB-C socket, what is likely to happen is that the socket itself (metal only) gets deformed slightly (which is enough to wreak havoc on the alignment and small tolerances of the USB-C copper contacts), and the USB-C socket itself has a less sturdy PCB soldered connection, and there are no plastics involved to absorb the shock (so solder connections are more likely to crack).

[caesarcaliyugula]

That is true, but a similar durability can be achieved with aviator cables on USB-C,

The problem is not the cable, that is easily replaced. The problem is the demolition done to the PCB mounted USB socket on the PCB.

To be exact, violent janks at the cable perpendicular to the socket, that's what we're talking about here.
For a USB-B socket, what gets damaged the most is the cable, the socket is sturdily through-hole PCB soldered and has a lot of guiding plastics to absorb most of the shock.
For a USB-C socket, what is likely to happen is that the socket itself (metal only) gets deformed slightly (which is enough to wreak havoc on the alignment and small tolerances of the USB-C copper contacts), and the USB-C socket itself has a less sturdy PCB soldered connection, and there are no plastics involved to absorb the shock (so solder connections are more likely to crack).

And aviator cables would alleviate that by moving the violent yanks and the stresses they induce from the PCB to the middle of the cable.

[thefarside]

For the new model F and B keyboards is the USB C connector still on the controller like the previous versions? If so my recommendation would be to wrap a zip tie tightly around the cable right before it goes through the hole in the case. This will protect it from any yanks by having the zip tie hit the case acting as a bump stop.

[BuGless]

And aviator cables would alleviate that by moving the violent yanks and the stresses they induce from the PCB to the middle of the cable.

Yes, they would absorb some of the short-term shocks, but the average perpendicular force exerted on the socket will not change, which still is likely to cause more damage on the shorter all-metal USB-C sockets.

[Ellipse]

Blog update posted:

Please see today's blog update post for a summary of the various goings-on with the project in recent months:
https://modelfkeyboards.com/blog

Interesting point Muirium - I did not know that about the rated insertions. However I do find that replacing USB-B connectors is something that I can do at my lower level of soldering skill, so if it fails it is more fixable in my experience. I had to replace the USB-B connector of the Fairbanks scale I use for the keyboard projects as the one that came with the scale was broken (I bought it used). I don't think I would be able to successfully do any complicated surface mount soldering as would be needed for a USB-C connector.

[enthdegree]

Is a beam spring F50 macro pad likely?

[Ellipse]

The beam spring project is open to additional designs with a minimum order quantity. For a completely different design I'd guess there would have to be at least 10 or more orders which would be a tall order for a $400-$500 B50 macro pad. Maybe there would be greater interest in an ortholinear / split board?

[caesarcaliyugula]

The beam spring project is open to additional designs with a minimum order quantity. For a completely different design I'd guess there would have to be at least 10 or more orders which would be a tall order for a $400-$500 B50 macro pad. Maybe there would be greater interest in an ortholinear / split board?

I'd personally be interested in a board that's the length of the B104 but still has extra macro keys for reprogramming. Perhaps just with the extra function row?

[Ellipse]

caesarcaliyugula I would recommend the B122 as you can set the 10 keys on the left or the extra keys on top to be macro keys. Better to have a few extra keys I would say! The B122's extra width is about 1.5 inches or so and the board I believe has the same length dimensions as the B104.

[caesarcaliyugula]

caesarcaliyugula I would recommend the B122 as you can set the 10 keys on the left or the extra keys on top to be macro keys. Better to have a few extra keys I would say! The B122's extra width is about 1.5 inches or so and the board I believe has the same length dimensions as the B104.

I do have the B122 on order at the moment, as it happens. It fits most of my needs, but I'll be placing it on my desk's keyboard tray, and the extra 1.7 inches of length means it will have to sit over my mousepad. Not a deal-breaker of course, but as long as we're making wish lists for the perfect custom order...

[dodddummy]

Now that we are now on the back side of the year, I'm curious if the thinking is B122s will ship in 3rd quarter or 4th.

[Ellipse]

We are still looking at later this year for all of the boards including the B122 (they should arrive in one container shipment, all the various new F and Beam keyboards). I would say closer to the end of the year.

[Ellipse]

Project update:

After the sample order was assembled and tested, the updated beam barrels (part A and B) for Round 2 were approved for full production and I placed the order this week. These took a while to get right but I am glad everything is good now.

Hopefully the S45500 material sheets arrive in the coming weeks, when the A and B parts are completed, so that the tens of thousands of beam spring modules can start to be assembled.

I have also ordered the capacitive PCBs for all 8 beam spring variants and the F122, as well as the Leyden Jar version 3 controller for testing.

The 8 beam spring variants: B62 Split Shift and Regular Backspace, B62 Split Shift and Split Backspace, BSSK ANSI, BSSK ISO, B104 ANSI, B104 ISO, B122 ANSI, B122 ISO

The factory is also preparing the tooling to bend the aluminum cases. Then they will produce a full working sample keyboard of each design for verification, and then the case and inner assembly plates production can start.

[ritter]

Thanks for the update Ellipse; another step towards completion. Very exciting.

[genericusername57]

Project update:

After the sample order was assembled and tested, the updated beam barrels (part A and B) for Round 2 were approved for full production and I placed the order this week. These took a while to get right but I am glad everything is good now.

Hopefully the S45500 material sheets arrive in the coming weeks, when the A and B parts are completed, so that the tens of thousands of beam spring modules can start to be assembled.

I have also ordered the capacitive PCBs for all 8 beam spring variants and the F122, as well as the Leyden Jar version 3 controller for testing.

The 8 beam spring variants: B62 Split Shift and Regular Backspace, B62 Split Shift and Split Backspace, BSSK ANSI, BSSK ISO, B104 ANSI, B104 ISO, B122 ANSI, B122 ISO

The factory is also preparing the tooling to bend the aluminum cases. Then they will produce a full working sample keyboard of each design for verification, and then the case and inner assembly plates production can start.

Thanks for the update! Very much looking forward to my R2 board

Just a question regarding the possibility of a "F77" style cutout between the nav cluster and the delete-end-etc-cluster above, would paying for that include a full set of beamspring modules to cover the board or would you need to order extras to fill in the blanks you create by going for such a configuration?

[Ellipse]

genericusername57 it would be included in the $100 additional cost.

[genericusername57]

genericusername57 it would be included in the $100 additional cost.

Thanks a lot for the answer, I probably will go for that configuration. And thanks again for this project, I adore your commitment to vintage keyboards and bringing them to public again

[Ellipse]

The S45500 thin material for the beam spring modules has shipped to the stamping factory. This "small" amount of material is enough to supply several hundred thousand modules. Finding the material was not too difficult but finding a supplier willing and able to produce such a "small" quantity of the material (i.e. less than one metric ton!) was challenging.

The material is a super-thin sheet that is supplied on a roll. Each sheet is a fraction of a mm in thickness.

2023 08 02 S45500 material (1).jpg (361.07 KiB) Viewed 183259 times

2023 08 02 S45500 material (2).jpg (513.97 KiB) Viewed 183259 times

2023 08 02 S45500 material (3).jpg (394.87 KiB) Viewed 183259 times

[jsheradin]

S45500

How was this material chosen? I assume real IBM modules would have been made from spring steel of some sort. I'm not a metallurgist but the composition of S45500 seems a bit different from any of the commonly available spring steel varieties. The datasheet doesn't list springs as one of its typical applications.

Was there any consideration or testing in regards to work hardening of the thin area around the center melt rivet? Of the few failed beamspring modules I've come across they always broke right in the same spot. If IBM was using actual spring steel and still seeing failures, I'd be very curious how these modules hold up after a year or two of use.

I think the final physical properties are also greatly influenced by the processing the raw material undergoes. Cold rolling it this thin might make it brittle but tempering it too much might make it too soft for this application. I think stamping typically hardens the material near the shear line making crack inception easier. There might need to be a tempering process after stamping to remedy this. I've run into issues on work projects where laser cut, wire EDM, and stamped components all behave wildly differently under certain circumstances.

[Ellipse]

jsheradin the team over at the Silo project posted an elemental material analysis of the IBM beam spring metal, so we know what it was. I reposted this analysis along with the S45500 composition earlier in this thread to do an element-for-element comparison, and I also posted a brief history of this relatively-new material (at the time IBM was developing the beam spring keyboard).

I have found that the plastic rivet around the metal should be ideally the same dimensions as what IBM did (not too big or too small; to minimize issues), and that the bigger issue of failure I have noticed is due to over-bending the material during assembly at the factory, as happened with a number of the Round 1 modules installed in the keyboards (one reason for including replacement modules with each board). This was fixed with the updated Round 1 modules included in the current first aid kits and for the Round 2 modules, though I will still include replacement modules with each Round 2 board to account for damage sustained in shipping.

Yes there were some processes done after cold rolling to reduce potential brittleness in the material.

The springs themselves are made from more standard spring steel, not from this S45500 material.

[enthdegree]

Is the F50 numpad dutch auction still live or did I just lose?