Get the top HN stories in your inbox every day.
gibolt
smnscu
I really enjoy Alec's videos and quirky humour. Beware though, this kind of channel is a massive, addicting time sink. Other similar channels that HN might appreciate (the first two are quite well-known, the last one I found recently and is definitely bingeworthy):
- Smarter Every Day https://www.youtube.com/channel/UC6107grRI4m0o2-emgoDnAA
- Tom Scott https://www.youtube.com/user/enyay
- Practical Engineering https://www.youtube.com/user/gradyhillhouse
iCarrot
Also check Ben Eater's for computer stuff https://www.youtube.com/c/BenEater/videos
mastax
- Marco Reps - Electronics, Metrology, Machining, etc. With a dry German sense of humor. Like electric This Old Tony https://youtube.com/c/MarcoReps
TonyTrapp
> Beware though, this kind of channel is a massive, addicting time sink
Luckily he only posts new videos every few weeks. ;) And almost every one of them is time well spent.
jamestenglish
Subscribed to all 3 of those, one of my new fav's
Stuff Made Here: https://www.youtube.com/channel/UCj1VqrHhDte54oLgPG4xpuQ
ziml77
Tech Ingredients is another good one with nice, long videos: https://www.youtube.com/user/TechIngredients
jedimastert
Stuff like this is why I have difficulty giving up YouTube. I've given up basically every other algorithmic feed (except for HN, I suppose) but the amount of passionate and engaging education and curiosity content on YT will have me hooked for a while I suppose...
smnscu
YES! I'm also really struggling with giving up YouTube due to all the wonderful content available there. I wish there was a widely-used alternative like Floatplane to double-post your videos for your supporters.
Benlights
I've found that watching Alec's videos at 1.25X speed made them more enjoyable and listenable.
s0rce
After watching this video I feel like every website is showing me flip flap displays.
teh_klev
I watched that last night. Alec's a tremendous educator and come a long way since his early videos (which are also still good). I really enjoyed his deep dive into how dishwashers work :)
sva_
Frankfurt airport has a huge split-flap display to show departures[0]. I love looking at it changing the data, and hope they'll never replace it with a digital one.
rjsw
The sound it makes lets you know when something has changed, you don't need to keep looking up from whatever else you are doing.
axolttl88
They were installed as useful & beautiful features in the Qantas First Class lounge at Sydney International Airport, and are much loved:
https://theluxurytraveller.com/2019/03/review-qantas-first-c...
jcadam
Still? Good to know :) Last time I went through Frankfurt was back in '06.
argvargc
Genuine question: what are some practical advantages of a split-flap display over, well... any other kind of dynamic character display?
The only two I could think of are:
- they don't emit light, so maybe there could be some kind of light-sensitive situation where this would be required (though in that case e-ink would surely surpass it).
- the sound passively-alerting those present to updates. It's not obtrusive or attention-getting like a bell or spoken announcement, so only those wishing to notice the updates can listen for the sound, while others can go about their business unperturbed.
Anyone know of, or could imagine any others?
goodpoint
- Visibility angle is much better than any other alternative. This is crucial when one big display is used in a very large room instead of many small displays.
- Image sharpness and contrast is excellent. You can read text on an old flip display far away that you would struggle to read on a modern LCD if you have mild astigmatism.
- Perfectly readable on a sunny day.
- They can show colored logos and symbols (e.g. train type, airline, highlight warnings)
- Better failure mode than LCD: single digits fail independently from each other and can be replaced. But it's less likely that the entire display fails.
- They are either working or stuck - no fading out like CRTs.
protomyth
- oddly pleasant sound indicates change in data
danohuiginn
Nostalgia, really -- I have fond associations of these with childhood train journeys.
I could just about imagine a genuine use case in some kind of hostile environment. You could probably make one of these very resistant to water/temperature/radiation, with the controller kept safely out of the way.
mjhagen
I used to have one of these, but left it above the radiator. All flaps warped beyond repair :( So if you're going for resistance to harsh environments you'd have to source a better material for the flaps, at least.
grishka
Hm, after watching and listening to videos of these displays at airports and train stations (don't think I've ever seen one in person), I thought the flaps in the commercial ones were made of thin sheet metal because the flipping sound was somewhat sharper than the OP version.
argvargc
That sparked an interesting idea: would these work in a zero-gravity situation (mechanically I mean), and if not, how to adapt them so they would?
(And agree with the nostalgia - I'm sure for many the sound is associated with a quiet anticipation or excitement of embarking on a journey.)
ridaj
I think they'll work... Somewhat? As the card is pushed through the face, there's a small amount of spring energy that it stores by bending, which gets released when it makes it through the opening and propels it forward + downwards. Normal gravity just adds to that motion and keeps the flap moving down. Zero g doesn't help as much but the card should already have momentum out of the gate. In zero g the problem may be more dealing with the rebound as the flap hits the bottom of the frame - it may bounce back up and float. May need some dampening or magnetic trick to keep it down...
In any case they probably won't work upside down.
grishka
> how to adapt them so they would?
Add rotary springs (is this a valid term?) to the flap hinges so they're forced down once released from the top edge of the frame.
martin_a
They probably won't.
After a flap has passed the upper lip, which is kind of holding it back to have a precise "flipping point", gravity will take care of moving it down.
No gravity, no flap. ;-)
michaelt
The traditional uses of split-flap displays were sports arena scoreboards, airports and train stations.
i.e. places where you want a very large display that's readable in direct sunlight.
If you were equipping a train station or airport in the 1980s, your options were split-flap displays, CRTs with sunshades over them (maximum size ~20 inches), a multi-CRT video wall ($$$$$$$$), a vane display (numbers only) or a flip-dot display. And a lot of places that were equipped with them in the 1980s still had them in the late 1990s.
These days you'd use an LED display, or a sunlight-readable TFT.
mastax
I had never realized that the original Sony Jumbotrons were CRT (one CRT per pixel) https://youtu.be/AclwH64eAkU
tyingq
Worth noting not all LED displays are sunlight readable. Typically, outdoor requires old fashioned through-hole, separate LEDS, and at least 1/4 duty pwm refresh.
dspillett
> And a lot of places that were equipped with them in the 1980s still had them in the late 1990s
Many UK rail stations had them into the early to mid 00s. If I'm not misremembering, the boards at London Kings Cross were not moved to different tech until at least 2005. Manchester Piccadilly had them replaced around 2001 (at the time I was passing through there a fair bit, travelling between York & North Wales).
vidarh
When I was a kid our hifi-system had a split-flap clock built in (around 1980 or so)
wayoutthere
As someone who used to travel a lot by rail in the not so distant past, every station that replaced one of these boards was a downgrade. Most stations had maybe 1 or 2 of these signs with the rest of the smaller ones already LCDs, so it’s not like there weren’t already places to get the info you couldn’t show on a sign like this.
Part of the problem with the LCD signs is they show too much information. It’s not really an advantage to show the next 30 trains incoming on the big board when only the closest 2 or 3 have gates assigned. They also tend to be themed in colors that are harder to read (Amtrak does blue on white, which is harder to read than white on black). Some stations tried replacing the split flaps with single-character LCD displays to mimic, but they always ended up breaking faster (and never getting fixed) for some reason.
From what I recall in an article about these signs, industrial versions were ungodly expensive to begin with, but the real reason they’re being replaced is that the companies that originally made them are all out of business and you simply can’t get parts for them. It’s cheaper to replace an entire LCD panel than it is to fix one of the mechanical signs once and you don’t have to wait 6 weeks with a broken sign for parts to be manufactured.
gattilorenz
similarly to an e-ink display, it doesn't take energy to show the current/last status, it's required only for transitions. If there's a blackout, the display is still visible.
goodpoint
> If there's a blackout, the display is still visible
And dangerously out of date in airports and train stations.
When they break down they had to be covered to prevent misleading passengers.
undefined
mschuster91
- they don't require energy to maintain their state - a property which they share with eInk displays, but a split-flap display is unrestricted in size.
- assuming a fixed output selection, a split-flap display can also achieve far greater visual quality than other competitors like flip-dot displays
- defective units can be automatically detected (by comparing the position of the indicator magnet), whereas stuck pixels in flip-dot, matrix or LC displays require a human to spot them
theodric
Zero power needed to retain image, not encumbered by e-ink patent and manufacturing monopoly, simple and repairable, and you can build it yourself at home
agys
Readability and ultra low energy consumption (plus: they are awesome and produce a soft sound like a distant waterfall).
Also, contrary on what I have seen on LED matrices or other RGB/pixel based displays: they won’t display advertisement which otherwise manages to sneak in – too close to information.
0xffff2
I wonder how their energy consumption actually compares. Electric motors moving relatively heavy physical cards is far from free. I wouldn't be surprised at all to find that an eink display beats the pants of of a split flap display in terms of energy consumption, but I don't have the necessary equipment to test it.
chrisdalke
This is great! Love seeing hardware hacks like this on HN :)
Beyond just being a cool project, the detail in the README on Github is really amazing and I've bookmarked this as a great reference for releasing open-source hardware as a git repository. CAD files, a bill of materials, detailed usage instructions, etc.
gonehome
I've always loved these - this company makes one that looks great, but isn't open source (there is an API that has yet to be released): https://www.vestaboard.com
It's also quite expensive, but has a clever design where the individual split-flap modules can be replaced.
There are also custom ones you can have made (probably even more expensive), but they're mostly targeting enterprise: https://www.oatfoundry.com/split-flap/
It'd be neat to have one you could hook up to some market API to make a little stock ticker.
Very cool to see work an open source design.
gorgoiler
Fantastic.
Glyphs can be loaded in an arbitrary order. For a given corpus (e.g. city names, American English digraphs) how does glyph order affect the time taken to update the display? Is there an optimal ordering?
If your display could accommodate 5% more glyphs, which would you add and why?
(With apologies to anyone revising for an exam this summer.)
agys
We made a small project with all the 6 letter words from the Italian dictionary (as an homage to the inventor and producer of the split flap display: Solari from Udine) ordered in a way to produce “the least effort”.
grishka
You can only flip in one direction, so the time taken to update is entirely dependent on the previous state. Worst case you'd need to display the previous character (B -> A for alphabetical order) and would have to flip through the entire alphabet to get to it.
nine_k
> You can only flip in one direction, so the time taken to update is entirely dependent on the previous state. Worst case you'd need to display the previous character (B -> A for alphabetical order) and would have to flip through the entire alphabet to get to it.
This only applies if you have only one copy of each letter.
You can consider having more copies of frequent letters mixed into the stack, if the stack's thickness allows.
eurleif
Wouldn't that only help with the average case, not the worst case? Either you have multiple copies of every character (in which case you're still flipping through the same number of characters as the original stack), or some characters don't have duplicates.
abetusk
Scott Bezek has a great video on "How a Split Flap Display Works" [0].
He also sells blank "flaps" on Tindie [1].
The documentation on how to build the Split Flap is also really awesome [2], including how to cut your own flaps [3].
I've been following this project for a while. It's pretty awesome. I estimate about $10 per character (as opposed to the $180/character (!!) another poster mentioned through Alibaba), with a large portion of the cost being the custom cut "flaps" (in small quantities). It seems like the cost could get in the $2-$4 range per character but I couldn't figure out how to do it without a significant amount of labor involved in cutting the flaps.
FYI, there's also a company called Oat Foundry that sells large split-flap displays [4]. I seem to remember the cost being in the $10k+ region but I don't see a price and I could be way off base.
Oh, interesting, I see Vestaboard also has a (smaller) one for around $3k [5].
[0] https://youtu.be/UAQJJAQSg_g
[1] https://www.tindie.com/products/scottbez1/blank-split-flap-d...
[2] https://github.com/scottbez1/splitflap/wiki
[3] https://github.com/scottbez1/splitflap/wiki/Cut-flaps
murderfs
> I've been following this project for a while. It's pretty awesome. I estimate about $10 per character (as opposed to the $180/character (!!) another poster mentioned through Alibaba), with a large portion of the cost being the custom cut "flaps" (in small quantities). It seems like the cost could get in the $2-$4 range per character but I couldn't figure out how to do it without a significant amount of labor involved in cutting the flaps.
Laser cutting a bunch of sheets at once could be one option, but material selection might be a problem: PVC releases chlorine when laser cut, which will do all sorts of horrible things to optics/the meatbag operating the machine. Maybe a cutting plotter will work, since you're using really thin sheets?
mschuster91
> Laser cutting a bunch of sheets at once could be one option
Use a waterjet cutter instead. You'll have some issues with microplastics in the water, but these can be filtered.
abetusk
Show me a water jet cutter that's priced at the hobbyist level (and isn't a full time job to upkeep) and I'm there.
abetusk
Laser cutting is straight out for the hobbyist, because of the chlorine, as you correctly point out. I think the way to do this is either with a CNC or with some die punch solution.
Doing a single cutout of PVC with a CNC would be fine, but doing 40x23x7 is probably not.
Doing a die punch is maybe possible but I don't really know how to create a custom punch and what tooling to use.
atrus
It's too bad there isn't a bigger market for this, steel rule dies aren't too expensive to get made. And used clickers are pretty cheap too.
But only if you were doing a bunch of signs!
jiveturkey
> Laser cutting is straight out for the hobbyist, because of the chlorine
Just don't use PVC blanks.
thunderbong
This HN post has been submitted by Scott Bezek linking his github source code
anonu
I've looked into buying ready-made solutions. Vintage split-flap mechanisms go for close to $1000 per character on eBay. There is a company I found in China that manufactures these, but its still around $300 or so per character. Displaying a tweet would cost close to $100k after shipping, etc...
tartoran
The randomness factor is quite mesmerizing, each flap search ending at a different timing along with the mechanical aspect make this a very fun thing to watch. Old tech was so cool..
oxplot
These screens are awesome. I did a design a few years ago, aiming for super low cost and easy build. Made a few units but it needs more refinement: https://splapr.com/
Animats
Moving parts made of acrylic, though. Will not wear well.
Delrin, maybe?
It's now possible to buy commercial split-flap displays from China. $180/char.[1]
michaelt
For some applications you wouldn't worry too much.
A score display for a sports team might only do one or two rotations per week.
Of course, for a clock or airport flight display, it'd be a different matter!
jiveturkey
eh? It's DIY. Make them out of whatever you like.
Animats
They're intended to be laser-cut, which limits the materials.
Those flaps and wheels really want to be injection-moulded from ABS, with round pins at the flap ends. That would be cheap if you were making enough of them.
(I miss TechShop. They even had an injection molding machine good enough to do this.)
jiveturkey
yes i very much miss TechShop. i was a lifetime member. too bad they tried to grow too big, too fast. i think they misunderstood their business, at a fundamental level.
the techshop 2.0 guy was a travesty.
techdragon
This is just awesome on so many fronts. I love the automatic update of the BOM using GitHub actions, it’s such a nice touch.
Get the top HN stories in your inbox every day.
Related video on Flip Clocks from Technology Connections: https://youtu.be/ZArBfxaPzD8