> I have not yet printed and tested this exact model!
Sadly that sums up the 3D printable scene perfectly. So many times I’v seen someone creating X for Y but they don’t have Y to test it but “it should work”.
Fair enough, but this isn't that case. They have Y (it's in the first photo) and tested a previous version of the model. The changes are predictable (rotating letters and slight scaling) so I don't think it's unreasonable to be confident and not waste plastic.
Amazingly, I think you’re right, but not for (negative) the reasons you state.
I see the 3D printing scene a remix culture where you’re supposed to try new things. If you want something to work exactly for your situation, you might have to do a little work yourself.
In this case, here we have an idea of making new typeballs for IBM typewriters. Here is one post where someone is playing with the idea. At the top of the page, there is a link to a GitHub page for you to adapt it as you see fit.
But then there’s also a link to a new page: https://selectricrescue.org/ where someone makes many more varieties, also with code for you to try yourself.
I see 3d printing much like the world of early open source coding. Sometimes you get a completed project that works out of the box. But often, you’ll have to tinker and adapt it to your exact hardware. Once a new 3d model is shown to be viable, it can be mixed and matched to make something new!
I have, along with some other ones I created from different fonts. See https://www.youtube.com/watch?v=1kXnsvYfaF4&t=2160s (it was an aside to a longer Selectric project I did; this typeball project came out around the same time so I decided to make a few of them).
Watching this back, it partially works but some of the letters clearly need their positions adjusted, and some of them seem to have trouble imprinting fully on the page. That's besides the fact that the proportional fonts look pretty weird when printed monospaced.
I didn't really go further with this because the resin printed balls didn't hold up very well. A slot near the hole would always break open, and then the ball would spin freely instead of being indexed like it was supposed to. These days I've pretty much abandoned home resin printing entirely since it is extremely messy, and I've never been able to make parts with it that weren't ridiculously brittle, even with the supposedly "strong" or "tough" resins.
I've used commercial resin prints that Shapeways/Xometry/PCBWAY have and they are a lot tougher, so maybe they would work well.
He is not the author. The author’s post would still say I’ve never tested this and leave people uncertain about whether it’s actually a viable design.
I feel like as a matter of policy or at least convention, people really shouldn’t be uploading models that they haven’t at least printed. It’s disrespectful.
One of the Teletypes I used to work on had an eccentric-driven Wiffletree to position the typebox. This was necessary because of it's high speed[0] to decode the ASCII character set (it wasn't a Baudot machine). The selector cam at the end of the machine would do the serial-to-parallel conversion, then clutches would rotate the eccentrics powering the wiffletree to select the character.
If you look at the ASCII chart [1] you can select a character via binary tree - the Wiffletree being a mechanical equivalent.
The problem with this 3D-printed typeball is that without being injection-molded with an extremely durable plastic, it will not make good strike. I suspect it would also be torn apart (separating at the printed layers), beginning with the smaller common characters (e.g. the period/comma would quickly fail). The typeball hits the paper hard and it is under incredible rotational strains.
I've recently retired an IBM Selectric II, only because it is so damn finicky! When it is operating it is a fantastic machine, but it definitely needs a full-time technician to service (and they've almost entirely died off at this point, save for Berkeley). I got tired of the pulley creep [which eventually leads to gibberish output], only solved with continued maintenance (parts too fine for my electrician hands, and my nearsight is antiquating rapidly).
For my daily drivers, I still love my Smith Corona "Coronet Super 12", which has individual strikes and a powerful motor (for all your latenight raging / brainstorming). Can't [easily] change the font, though (which is why I have multiple S-C models).
> without being injection-molded with an extremely durable plastic, it will not make good strike.
I wonder how true that actually is. A properly 3d printed ABS part isn't really much if any less durable than an injection molded part. It's mostly just worse with respect to feature resolution.
>Mostly just worse with respect to feature resolution.
Inclusive to my injection-molded comment, but could be mitigated (e.g.) with an acetone bath (to "smooth" the layers/resolution).
The stress fractures will most-definitely grow parallel to the layers, though... watch a Selectric's mechanism (rotational; linear; impact), in slow motion:
Ah yeah, you would probably get stress risers along the layer lines. Hadn't considered that failure mode. I'd still expect wear and "mushrooming" of the type to be a problem before stress fractures are though. Plastic is pretty soft and gummy when you get down to it.
Interesting to see the timing on the mechanism too. It looks like it's already getting the rotation set up for the next character before the return stroke even happens. I suspect in normal use the ball would bounce off the page, avoiding smears.
EDIT: worst case you could always make a silicone mold and cast the type ball in an engineering resin of one kind or another. That probably fares better over the long term than 3d printed plastic.
>I suspect in normal use the ball would bounce off the page, avoiding smears.
>Plastic is pretty soft and gummy when you get down to it.
I don't think most people could break the typewriter ball with their bare hands, either by pulling or compressing ("squishing"). Just tried this with my least-favorite font — 275lb blue-collar electrician can't break it (barely any deflection).
Not sure what the plastic's composition is, but it is absolutely RIGID (as it must be). The thing can fly letters onto the page (I can do sustained 80wpm, with bursts into 100+ — thing could go twice my output rate on a clear-headed day, mechanically).
Selectrics are just absolutely modern marvels (still)!
>engineering resin
The ball's plastic is harder than cured JB-Weld™ (standard 2-part mixture).
Oh yeah on a human scale there's plenty of VERY hard plastics.
It's still pretty soft and gummy when compared to a LOT of materials. It's all relative and whatnot. You'll often find machinists using descriptors like "chewing gum consistency" for stuff like copper. On that scale plastic is pretty soft and squishy.
> I have not yet printed and tested this exact model!
Sadly that sums up the 3D printable scene perfectly. So many times I’v seen someone creating X for Y but they don’t have Y to test it but “it should work”.
Fair enough, but this isn't that case. They have Y (it's in the first photo) and tested a previous version of the model. The changes are predictable (rotating letters and slight scaling) so I don't think it's unreasonable to be confident and not waste plastic.
Amazingly, I think you’re right, but not for (negative) the reasons you state.
I see the 3D printing scene a remix culture where you’re supposed to try new things. If you want something to work exactly for your situation, you might have to do a little work yourself.
In this case, here we have an idea of making new typeballs for IBM typewriters. Here is one post where someone is playing with the idea. At the top of the page, there is a link to a GitHub page for you to adapt it as you see fit.
But then there’s also a link to a new page: https://selectricrescue.org/ where someone makes many more varieties, also with code for you to try yourself.
I see 3d printing much like the world of early open source coding. Sometimes you get a completed project that works out of the box. But often, you’ll have to tinker and adapt it to your exact hardware. Once a new 3d model is shown to be viable, it can be mixed and matched to make something new!
I have, along with some other ones I created from different fonts. See https://www.youtube.com/watch?v=1kXnsvYfaF4&t=2160s (it was an aside to a longer Selectric project I did; this typeball project came out around the same time so I decided to make a few of them).
Watching this back, it partially works but some of the letters clearly need their positions adjusted, and some of them seem to have trouble imprinting fully on the page. That's besides the fact that the proportional fonts look pretty weird when printed monospaced.
I didn't really go further with this because the resin printed balls didn't hold up very well. A slot near the hole would always break open, and then the ball would spin freely instead of being indexed like it was supposed to. These days I've pretty much abandoned home resin printing entirely since it is extremely messy, and I've never been able to make parts with it that weren't ridiculously brittle, even with the supposedly "strong" or "tough" resins.
I've used commercial resin prints that Shapeways/Xometry/PCBWAY have and they are a lot tougher, so maybe they would work well.
Yeah, feels a bit like "I'm sharing this code I've written here. I haven't tried to compile it yet, but it should work."
That approach does work for LLMs, though.
Looks like the same person made a typeball for Cherokee, and that comes with a picture of the printed output:
https://www.printables.com/model/441262-cherokee-typeball-fo...
So maybe Comic Sans (and Wingdings) will work as well.
They also link a version that works they say. https://selectricrescue.org/
Also includes Papyrus!
It’s not far fetched though. With careful design, prints often do work on the first try.
Sure. And it's given away for free, so complaining feels wrong.
But is it _too_ much to ask that they print their tweaked version 2 and test it before publishing it?
:sigh:
Why don’t you print and test it and contribute to the documentation? Too much to ask as well? ;)
He is not the author. The author’s post would still say I’ve never tested this and leave people uncertain about whether it’s actually a viable design.
I feel like as a matter of policy or at least convention, people really shouldn’t be uploading models that they haven’t at least printed. It’s disrespectful.
Published is better than screenshot
The article says "I tested a previous revision", so this isn't the first go around.
> they don’t have Y to test it but “it should work”.
SW development process 101. What are users for ? /s
Works for shipping commercial software!
(Slightly /s)
Bill Hammack (engineerguy) has an excellent video on the IBM Selectric titled IBM Selectric Typewriter & its digital to analogue converter.[1]
[1] https://www.youtube.com/watch?v=bRCNenhcvpw
One of the Teletypes I used to work on had an eccentric-driven Wiffletree to position the typebox. This was necessary because of it's high speed[0] to decode the ASCII character set (it wasn't a Baudot machine). The selector cam at the end of the machine would do the serial-to-parallel conversion, then clutches would rotate the eccentrics powering the wiffletree to select the character.
If you look at the ASCII chart [1] you can select a character via binary tree - the Wiffletree being a mechanical equivalent.
[0] For a Teletype
[1] https://en.wikipedia.org/wiki/ASCII#/media/File:USASCII_code...
I couldn't resist.
https://github.com/rbanffy/3270_type_element
Now I need to get myself an IBM Selectric, or make a daisy wheel printer element for my typewriter (which will eventually become a terminal).
The problem with this 3D-printed typeball is that without being injection-molded with an extremely durable plastic, it will not make good strike. I suspect it would also be torn apart (separating at the printed layers), beginning with the smaller common characters (e.g. the period/comma would quickly fail). The typeball hits the paper hard and it is under incredible rotational strains.
I've recently retired an IBM Selectric II, only because it is so damn finicky! When it is operating it is a fantastic machine, but it definitely needs a full-time technician to service (and they've almost entirely died off at this point, save for Berkeley). I got tired of the pulley creep [which eventually leads to gibberish output], only solved with continued maintenance (parts too fine for my electrician hands, and my nearsight is antiquating rapidly).
For my daily drivers, I still love my Smith Corona "Coronet Super 12", which has individual strikes and a powerful motor (for all your latenight raging / brainstorming). Can't [easily] change the font, though (which is why I have multiple S-C models).
¢¢
> without being injection-molded with an extremely durable plastic, it will not make good strike.
I wonder how true that actually is. A properly 3d printed ABS part isn't really much if any less durable than an injection molded part. It's mostly just worse with respect to feature resolution.
>Mostly just worse with respect to feature resolution.
Inclusive to my injection-molded comment, but could be mitigated (e.g.) with an acetone bath (to "smooth" the layers/resolution).
The stress fractures will most-definitely grow parallel to the layers, though... watch a Selectric's mechanism (rotational; linear; impact), in slow motion:
https://www.youtube-nocookie.com/embed/RTtKaqIpOJc
https://www.youtube-nocookie.com/embed/izZ02t2UEGc
https://www.youtube-nocookie.com/embed/1ctlRduNCn4
Ah yeah, you would probably get stress risers along the layer lines. Hadn't considered that failure mode. I'd still expect wear and "mushrooming" of the type to be a problem before stress fractures are though. Plastic is pretty soft and gummy when you get down to it.
Interesting to see the timing on the mechanism too. It looks like it's already getting the rotation set up for the next character before the return stroke even happens. I suspect in normal use the ball would bounce off the page, avoiding smears.
EDIT: worst case you could always make a silicone mold and cast the type ball in an engineering resin of one kind or another. That probably fares better over the long term than 3d printed plastic.
>I suspect in normal use the ball would bounce off the page, avoiding smears.
>Plastic is pretty soft and gummy when you get down to it.
I don't think most people could break the typewriter ball with their bare hands, either by pulling or compressing ("squishing"). Just tried this with my least-favorite font — 275lb blue-collar electrician can't break it (barely any deflection).
Not sure what the plastic's composition is, but it is absolutely RIGID (as it must be). The thing can fly letters onto the page (I can do sustained 80wpm, with bursts into 100+ — thing could go twice my output rate on a clear-headed day, mechanically).
Selectrics are just absolutely modern marvels (still)!
>engineering resin
The ball's plastic is harder than cured JB-Weld™ (standard 2-part mixture).
Oh yeah on a human scale there's plenty of VERY hard plastics.
It's still pretty soft and gummy when compared to a LOT of materials. It's all relative and whatnot. You'll often find machinists using descriptors like "chewing gum consistency" for stuff like copper. On that scale plastic is pretty soft and squishy.
Doing the devil's work.
Comic sans is quirky and wholesome. I'm tired of pretending I don't like it.
This is very cool
Had just recently looked up IBM Selectric typeballs and the possibility of 3D print custom ones but did not expect so many active projects around it.
Pretty nice time for nostalgic tinkerers to be alive ...
Dave Hayden of Austin Selectric Rescue is working on several type elements for the Selectric II: https://selectricrescue.org/
As I understand, the recent IBM Plex Mono face has been requested too.
It’s even the first link on this page. In addition to Comic Sans, he’s also made a font ball for Papyrus!
I can’t imagine using any of these (even when typewriters were king), but I love that they exist.
What do you think you’re running from? The disease is inside you!
https://achewood.com/2007/07/05/title.html
Don’t approve of the font but do approve of hacking old tech
Is this still post-modern or is it post-irony already?
It's vintage now and gives me warm feelings.
Is this intended to be offensive?
“You will live to see man-made horrors beyond your comprehension.”
[dead]
[dead]
[flagged]
I too irrationally hate incredibly legible typefaces.
https://dyslexichelp.org/why-is-comic-sans-good-for-dyslexia...
Cyanide and Happiness reference, for graet justis.[0]
(maybe not NSFW, but... eh. edgy. it's up to you.)
[0] https://explosm.net/comics/dave-comicsans
Definitely public typography enemy number one, followed a bit behind by Papyrus
[flagged]
Every key punch will translate into a bunny punch
https://topher1kenobe.com/phlog/graphics/bunnypunch.png