43s status

[walter b]

(04-05-2014 04:58 PM)Richard Ottosen Wrote:  Duly noted by the hardware department. Dealt with in the Laser-cut prototypes.

I will forward the reminder to injection molded case department. :-)

Thank you for forwarding. Isn't that a job for your intern?

d:-)

[anetzer]

(04-05-2014 05:04 PM)walter b Wrote:  Thank you for forwarding. Isn't that a job for your intern?

As long as the product presentation event hinges on that feature...

"One more thing..."

[orcinus]

So, here's an odd idea... (regarding hardware)

The biggest hurdle in manufacturing the hardware is the keyboard (at least the way i see it). So why not gut the 30B for its keyboard PCB and keycaps only, transplant it into a custom designed and 3D printed (or later injection molded) case and redesign the 43S layout to fit (with overlays on the keys and the case, obviously, as with 34s)?

Yes, it's a big compromise, but it might provide a nice stepping-stone towards a "true" 43S.

[brouhaha]

(04-10-2014 09:45 PM)orcinus Wrote:  So, here's an odd idea... (regarding hardware)

The biggest hurdle in manufacturing the hardware is the keyboard (at least the way i see it). So why not gut the 30B for its keyboard PCB and keycaps only, transplant it into a custom designed and 3D printed (or later injection molded) case and redesign the 43S layout to fit (with overlays on the keys and the case, obviously, as with 34s)?

Yes, it's a big compromise, but it might provide a nice stepping-stone towards a "true" 43S.

The PCB and switches aren't a problem. If you're willing to settle for keys that have the wrong legends, as would be the case if you used 30b keys, then manufacturing new keys isn't a problem either. I've 3d-printed keys with no legends without any trouble, and injection-molding them without legends isn't that big a deal. It's getting quality legends that's difficult. We're not going to get double-shot molding. The next best choices for long life are laser marking or dye sublimation. The only processes with low NRE are pad printing or adhesive overlays.

[Manolo Sobrino]

(04-11-2014 03:30 AM)brouhaha Wrote:  The PCB and switches aren't a problem. If you're willing to settle for keys that have the wrong legends, as would be the case if you used 30b keys, then manufacturing new keys isn't a problem either. I've 3d-printed keys with no legends without any trouble, and injection-molding them without legends isn't that big a deal. It's getting quality legends that's difficult. We're not going to get double-shot molding. The next best choices for long life are laser marking or dye sublimation. The only processes with low NRE are pad printing or adhesive overlays.

Or you could wait until this technology gets cheaper:

Objet500 Connex3

And just print them in colour .

[walter b]

(04-11-2014 10:52 AM)Manolo Sobrino Wrote:  Or you could wait until this technology gets cheaper:

Objet500 Connex3

And just print them in colour :).

Sure - when did you want the 43S? 2020?

d:-/

[anetzer]

(04-11-2014 03:30 AM)brouhaha Wrote:  It's getting quality legends that's difficult. We're not going to get double-shot molding. The next best choices for long life are laser marking or dye sublimation.

Two things I found. Perhaps some ideas might spawn from this:

Geeckhack

SBS Keycaps (Youtube)

If I understand right that gamers have gone into modding their keycaps just for the heck of it, I would assume that a sizable bunch of calculator keys should also be feasible...

"Just to shift a few tons..."

a.n.

[Manolo Sobrino]

(04-11-2014 11:46 AM)walter b Wrote:  Sure - when did you want the 43S? 2020?

d: -/

OMG, a release date!!

Seriously, Shapeways can print for you now in something called "full color sandstone". It's a question of time until they use ordinary resins with printers like that. I'd just keep an eye on it, the field is growing at full speed.

[orcinus]

(04-11-2014 03:30 AM)brouhaha Wrote:  The PCB and switches aren't a problem. If you're willing to settle for keys that have the wrong legends, as would be the case if you used 30b keys, then manufacturing new keys isn't a problem either. I've 3d-printed keys with no legends without any trouble

I never said the PCB is the problem, i've said the keys are

I've been 3D printing various stuff and, trust me, while printing button- and key-caps is a neat thing to show off your project at a hackerspace, it is nowhere near something you'd want on an actual, non-prototype, usable machine.

Finding decent switches is not impossible, but is relatively hard. If you're scavenging the keys, you might as well scavenge the switches as well.

The PCB is then a no-brainer.

Now overlays... remind me -how are overlays suddenly a problem, after years of WP-34S' existance? And, like i've said, it's just a first step. A calculator with the keys and overlays in the same vein as 34S make for a much better first step (and clearer message) than a single heat-shrink wrapped membrane keyboard, does it not?

[brouhaha]

(04-11-2014 11:10 PM)orcinus Wrote:  I've been 3D printing various stuff and, trust me, while printing button- and key-caps is a neat thing to show off your project at a hackerspace, it is nowhere near something you'd want on an actual, non-prototype, usable machine.

I've been 3D printing calculator parts for more than six years, using printers from Stratasys, Makerbot, and Lulzbot. I've shown a calculator with 3D printed keys at multiple calculator conferences, and no one complained about the quality of the keys. They were designed to be similar to the 41C and 12C keys. The surface texture isn't as smooth as injection molded, but that can easily be taken care of with acetone. It really doesn't matter since the 3D printed keys are too expensive for anything but prototypes; production will have to be injection molded.

(04-11-2014 11:10 PM)orcinus Wrote:  Finding decent switches is not impossible, but is relatively hard.

Actually that's not even slightly hard. Search for "tactile switches" on Digikey or Mouser. There are hundreds of different ones that are suitable, and some are even low enough cost for a premium calculator, though not for a cheap blister-pack calculator for Walmart.

Richard Ottosen and I have been building calculator prototypes for more than ten years, working toward something that could actually be sold. While we're always happy to hear new ideas, it seems a bit strange to be told that things that we've already accomplished successfully are not usable, or that things we've done easily are "relatively hard".

[brouhaha]

(04-11-2014 10:52 AM)Manolo Sobrino Wrote:  And just print them in colour .

I've done two-color 3D printing. The resolution is nowhere near good enough for calculator key legends, and that's unlikely to change any time soon.

Almost all 3D printers have vary good positioning resolution, but that doesn't result in the ability to print parts with extremely fine detail, due to the nature of the printing processes, especially plastic extrusion.

[orcinus]

(04-11-2014 11:23 PM)brouhaha Wrote:  It really doesn't matter since the 3D printed keys are too expensive for anything but prototypes; production will have to be injection molded.

1. Who's talking about production?
2. Were the keys you printed flat or profiled?
3. Did you show them outside calculator conferences and were they actually used?

(04-11-2014 11:23 PM)brouhaha Wrote:  Actually that's not even slightly hard. Search for "tactile switches" on Digikey or Mouser. There are hundreds of different ones that are suitable...

I assume you've tried them all and found the ones that are suitable?

(04-11-2014 11:23 PM)brouhaha Wrote:  Richard Ottosen and I have been building calculator prototypes for more than ten years, working toward something that could actually be sold. While we're always happy to hear new ideas, it seems a bit strange to be told that things that we've already accomplished successfully are not usable, or that things we've done easily are "relatively hard".

This will sound terribly disrespectful, but... After ten years, have you finally gotten to the point where something can actually be sold?

And by that i honestly am not trying to be disrespectful - *i'd* rather have a solid prototype than a hacked product and i appreciate the effort involved. But the reality of it is - no one will buy a prototype with 3D printed keys. But people will buy a factory made product hacked into something else, despite all its shortcomings.

And that's my original point (which, granted, might be completely off base). A tiny step sideways is sometimes better than standing still and talking about that big step forward.

[orcinus]

(04-11-2014 11:32 PM)brouhaha Wrote:  I've done two-color 3D printing. The resolution is nowhere near good enough for calculator key legends, and that's unlikely to change any time soon.

Not only that, but the repeatability of multi-color prints is terrible with FDM printing. SLA printing keys in resin, with "troughs", then filling those with paint might work (and be fairly long-lasting), but would end up very pricey and time consuming. And powder+binder process resolution is just too low for text, as you've said.

[brouhaha]

(04-11-2014 11:42 PM)orcinus Wrote:  

(04-11-2014 11:23 PM)brouhaha Wrote:  It really doesn't matter since the 3D printed keys are too expensive for anything but prototypes; production will have to be injection molded.

1. Who's talking about production?

I am.

Quote:2. Were the keys you printed flat or profiled?

As I said, they were basically the same shape as HP-41C and HP-12C keys.

Quote:3. Did you show them outside calculator conferences and were they actually used?

I've shown them outside of calculator conferences. Aside from people that have tried it, I'm probably the only person who's used them extensively, because only one person was willing to pay the exorbitant price of a 3D printed calculator.

Quote:

(04-11-2014 11:23 PM)brouhaha Wrote:  Actually that's not even slightly hard. Search for "tactile switches" on Digikey or Mouser. There are hundreds of different ones that are suitable...

I assume you've tried them all and found the ones that are suitable?

No. We've tried two or three of them. Many are suitable; the only reason we've tried several is because we have tried several different case designs that had different requirements for the size and shape of the switches, and especially the actuators thereof.

Quote:

(04-11-2014 11:23 PM)brouhaha Wrote:  Richard Ottosen and I have been building calculator prototypes for more than ten years, working toward something that could actually be sold. While we're always happy to hear new ideas, it seems a bit strange to be told that things that we've already accomplished successfully are not usable, or that things we've done easily are "relatively hard".

This will sound terribly disrespectful, but... After ten years, have you finally gotten to the point where something can actually be sold?

No, it's not yet to the point of being sold. It's almost to the point of providing loaner units to the firmware development team.

This has been a spare time project, and the objectives have changed over the years. When we started, we never dreamed that there would be any third-party firmware on part with WP-34s. We were only building a platform to run HP calculator microcode.

Richard and I are professionals. Had this been a commercial endeavor, and had we worked full time on it, and had a suitable budget, we would have been able to produce a professional grade product in less than 18 months.

Quote:And by that i honestly am not trying to be disrespectful - *i'd* rather have a solid prototype than a hacked product and i appreciate the effort involved. But the reality of it is - no one will buy a prototype with 3D printed keys. But people will buy a factory made product hacked into something else, despite all its shortcomings.

That depends on exactly what is meant by "hacked into something else". There's an existence proof that people will buy an HP 20b or 30b with replacement firmware and vinyl overlays. There's not any evidence that people will buy something cobbled together from pieces of other calculators.

Quote:And that's my original point (which, granted, might be completely off base). A tiny step sideways is sometimes better than standing still and talking about that big step forward.

Gutting an actual HP calculator for keys would be a diversion that wouldn't get us any closer to a product.

[walter b]

(04-10-2014 09:45 PM)orcinus Wrote:  So why not gut the 30B for its keyboard PCB and keycaps only, transplant it into a custom designed and 3D printed (or later injection molded) case and redesign the 43S layout to fit (with overlays on the keys and the case, obviously, as with 34s)?

Simply since the 30b features one row of six keys less than the 43S will do.

d;-)

[walter b]

(04-12-2014 03:01 AM)brouhaha Wrote:  It's almost to the point of providing loaner units to the firmware development team.

... who are eagerly waiting for them (maybe wrong grammar-wise since 'are waiting' is said to be for a moment in time but that moment is ... hmmmh ... a bit strained and we've read such annunciations more than once. Anyway the team is looking forward to those units )

d:-)

[sa-penguin]

(04-11-2014 11:46 PM)orcinus Wrote:  SLA printing keys in resin, with "troughs", then filling those with paint might work (and be fairly long-lasting), but would end up very pricey and time consuming. And powder+binder process resolution is just too low for text, as you've said.

What about - milling?
Take a 3D printer bed that moves XY, replace the plastic injector with a drill holding a very fine milling ball, and you have a turret mill.
Result: characters are engraved into the plastic.
Optionally followed by a syringe-based paint. Then a clear lacquer layer.

http://www.architectscornerstore.com/ima...Categories

[colinh]

(04-12-2014 09:30 AM)sa-penguin Wrote:  

(04-11-2014 11:46 PM)orcinus Wrote:  SLA printing keys in resin, with "troughs", then filling those with paint might work (and be fairly long-lasting), but would end up very pricey and time consuming. And powder+binder process resolution is just too low for text, as you've said.

What about - milling?
Take a 3D printer bed that moves XY, replace the plastic injector with a drill holding a very fine milling ball, and you have a turret mill.
Result: characters are engraved into the plastic.
Optionally followed by a syringe-based paint. Then a clear lacquer layer.

http://www.architectscornerstore.com/ima...Categories

Or how about printing, then milling, and then printing with a different colour to fill the milled glyphs.

If you were to attach the mill, rather than replace the printer head (extruders), you could print-mill-fill all the keys without intervention. Then a quick acetone vapour bath and you're done.

I've only just (like - yesterday) started out on FDM printing, and only have a single extruder so it'll be a while before I can actually do this myself...


BTW, how many calculators does the marketing department envisage making/selling?

[colinh]

Oh, and here's how Leica fill the engravings on their cameras and lenses... :-)

http://us.leica-camera.com/assets/media/img2771.jpg

[sa-penguin]

(04-13-2014 10:11 PM)colinh Wrote:  Oh, and here's how Leica fill the engravings on their cameras and lenses... :-)
http://us.leica-camera.com/assets/media/img2771.jpg

Bingo!
I was also thinking of a technical drawing pen, loaded with white ink [for black keys] but then there's the secondary function labels.
Flourescent yellow would be easier to source as a Revell paint 32312, for example.
See: http://www.wonderlandmodels.com/products...us-yellow/