Worlds Largest Open-Source Ferrofluid Display - Fetch #8

[@crowdrctrl]

“Snake” would be a great game project, perhaps...

[@harviecz]

Also tetris or space invaders :-) Also this seems to be well suited for implementing game called "world of goo"

[@TroySchrapel]

Snake could work, except new "apples" would need to be raised into position from the bottom. Or you could build a border and have an apple pull into position from anywhere in the border?

[@ianmalcolm177]

Pulling the apples up could actually be a pretty cool effect

[@alvin41712]

Maybe Pong

[@UnderscoreEightHundred]

@@harviecz The world of goo was my childhood

[@TamDNB]

here from wintergatan, this project is so f'n cool

[@AppliedProcrastination]

Thank you!

[@szymek1567]

Thank you for open sourcing this :*

[@Metalguy40]

This is absolutely amazing! Came here after Martin’s MMX video.

[@verypotato6699]

Metalguy40 same

[@AnandKumar-vz9kk]

Me too

[@thekillingmachine4967]

Just saw this on reddit, what an incredible project, awesome work!

[@captainmaay]

This is too insanely cool, I desperately need to build one for myself ! It’s going to take me years... but so be it... it’s now on my bucket list

[@artinacart8035]

Incredible! This seems like something someone would wanna patent and keep under wraps, but making it open source is just awesome. This kinda stuff inspires me

[@gamingbud926]

NASA? Sweet! Congrats on the internship! I hope we'll be able to hear about some of the stuff you'll be doing there. :) So happy to see the version 1.0 final product!

[@AppliedProcrastination]

Thanks, Bud! I really hope I’ll be able to tell you about it as well! -Simen

[@kiranseusanker2426]

I think a clock is a great place to start with a ferrofluid display, you can improve small thing over time as tech improves. I would personally love to see snake on the display

[@peacockmoss1491]

This is amazing! Also one suggestion for a game is two player ping pong.

[@lukekamrath8795]

As soon as I saw it on Wintergatan I thought "I want one of those." and "How would I make it." Job done its open source SUPER AWESOME. A while back i watched a NileRed ferro-fluid video, he talked about how high quality ferro-fluid is not a mix of oil and iron, but nano-particles of iron with a special oil like coating that makes the particles not stick together. He makes some of his own and I think he just put it in a glass bottle with water and it never stuck to the sides. Long tall display to make tetris.

[@AppliedProcrastination]

Thanks, and welcome to our channel! Saying that ferrofluid basically is an oil with iron particles is a slight simplification, but not incorrect. We just want to get the point across that it reacts to a magnetic field, not explain the scale of those particles or that they’re coated in oleic acid to avoid clumping - those details are just not important for what we want to convey in this video :) we do love NileRed’s ferrofluid videos though, and hope we get the chance to make some from scratch ourselves one day!

[@GiGaSzS]

Just beautiful :O BTW Wintergatan brought me here ;)

[@johnolivera2369]

god, its like magic...WONDERFUL!

[@hashkafe]

this is the coolest project ive ever seen

[@AppliedProcrastination]

Thank you, glad you enjoyed it!

[@icedrago6500]

Whoa, so cool, subbed! You can also turn that into an abstract display when you play music

[@1337treats]

Your constraint of only being able to pull ferrofluid up from the bottom could be mitigated by hiding additional electromagnets along the frame which would allow you to drop ferrofluid down from the top or pull it in from the sides. Just a thought. I'd love to see you guys experimenting with more resolution. What would it look like to have a wave of ferrofluid fading across the display?

[@Chriss2370]

Yes, yes very well done indeed. Very emotional!

[@risfutile]

fetch-final-final... oh man, what a classic! 😂 Fantastic job guys!

[@thesuperginge1348]

"Stop trying to make Fetch happen!" Really cool guys, well done!

[@asheshshrestha]

super awesome.. congrats for hard and successful work

[@testingbeta7169]

this is seriously impressive

[@thoughtsofapeer]

I think it would be cool to have an info screen showing the weather, current song playing on spotify etc. EDIT: Oh, also, and I imagine this is really tough, but it would be so cool if you could have layers of glass with paint in different colours (CMYK), to display colourful things, even opensource artwork on rotation. But again, it might have to include the colours reacting to specific "wavelengths" such that it can pick up the right colour.

[@AppliedProcrastination]

Weather station would be very cool. We don’t have high enough resolution to show song titles, though. The color idea probably won’t work, unfortunately.

[@qu1j0t3]

Love this project. Others have suggested "Snake" games but don't forget "Centipede"!

[@mskiptr]

Thank you for making it so awesome (and open)!

[@ericofire]

this is incredible

[@AppliedProcrastination]

Thank you!

[@Gwalchgwyn]

I am surprised that a sound-reactive graphic equaliser was not one of your first demos. Very cool idea and impressive result- wonderful work! :)

[@AppliedProcrastination]

We’ve had it in the plans for a while, but this felt like a good opportunity to actually do it :)

[@AppliedProcrastination]

I just noticed that this comment wasn’t on our latest video, so you might want to check that one out if you want to see music visualization!

[@namila007]

wow this is so cool. nice work guys

[@derpnerpwerp]

You could make the ferrofluid come out of different spots by adding an "emitter" at different locations, although it would be a lot more complex and would change the experience. The emitters could be fed from the reservoir. A simpler change could be to add side channels outside of the view which could be used to pull in fluid from all four sides of the display. This would allow you to do a game kind of like Galaga where you have things able to both descend and ascend across the screen

[@gerriepretorius4642]

Coming from Wintergatan. Very impressed by your machine

[@AppliedProcrastination]

Thank you!

[@Cairo_666]

Amazing! Great work, I hope to see more of this in the future. I think a good game to play on it would be the classic Snake game.

[@crazyegga]

Love It! Please implement "Pong". If you can draw from a bigger fluid tank, even Tetris or Snake might be possible!

[@playonwoods]

Saw your shoutout on Martin's channel. I can't wait to catch up on your content! I'm going to love it here.

[@BR.]

Amazing! Great job!

[@derekflynn6431]

Not many projects out there that are this unique, amazing job guys I’m seriously considering trying to build my own.

[@AppliedProcrastination]

Please let us know if you do! :)

[@humanperson8418]

Have you guys experimented with different density liquids, and using sound waves to make the ferrofluid 'dance', while also using the magnets to suspend it in place?

[@TheViralInfekT]

What a awesome idea. It looks great.

[@BlueMastic]

That is a really nice way to display things, not just as a clock. Also as a game suggestion: "Tetris upside-down" As the pixels need to start from the bottom but can go up this should be possible i think.

[@jasonl3445]

So awesome!

[@cobralyoner]

this is SO cool!!!

[@JRwhite1991]

Here from Wintergatan, awesome project and congratulations on the internship! PS Flappy Bird xD

[@AppliedProcrastination]

Thank you!

[@Skovjuul]

Fantastic project! Can't wait to try and work on it in 2020.

[@louisrnd4065]

Just wow from France ! 😮

[@foxpup]

Black and White Digital Lava Lamp - That's what this feels like. :-) Gr00vy Man!

[@ozzlemcschnozzle7250]

How does tank width affect everything? and have to considered keeping the top row of magnets allways on for quicker fluid availability

[@logannelsen4932]

I just came across this awesome project, and I have an idea! Use the top (and maybe sides too) as a reservoir. The fluid could be brought to the top via magnets in the bezels.

[@logannelsen4932]

Second idea: create a full-panel reservoir behind the magnets that can “spill over” in desired locations.

[@migfer7]

amazing project.

[@ChrisBigBad]

Is there a video where Fetch just does a bunch of animations? I'd like to marvel at it building something from start to finish.

[@AppliedProcrastination]

Hmmm. We don’t really have a video like that, but you can check out the videos we’ve posted on Twitter while developing it. Those are often longer and contain the entire animation (sometimes still with a bit of cross-camera editing though).

[@mtrdesigns7844]

Nice work guys! I work with ferrofluid as well so I understand how much work you put into this. Looks great 👍

[@bernard2735]

You had me at ‘Applied Procrastination’. Like and subscribed.

[@danielbrazell]

I wanna play Doom on this.

[@dandejy6520]

Actually more suitable for space invaders !

[@tomikun8057]

bad apple

[@NicksStuff]

This is magnificent

[@irpyc]

Really beautiful! With some lighting integrated it could be even more amazing. Maybe some background lighting? EL (electroluminescent) panel could work great, since they are really flat and would have about enough light for use in not so bright room. From the side, a nice continues light stripe could provide neat reflection. Like some "COB LED Stripes" (the continus one where you don't see each light dot) Preferably Warm light. And good CRI (90) would worth the expense for such great object you just made.

[@irpyc]

For background lighting you could alternatively try to use the rear glass as light panel guide. For this you could just try to put the same cob led strip directly on to the edge. Then it would need some structure on/in the glass to catch the light. But it can be tricky to get homogenous light, especially near the edges/LED where it would be brighter. Gaffer/duck tape can do the trick for trying it out (with the fabric structure you can press it down more or less on the glass to achieve different brightness), a more proper solution could be to sand blast the glass.

[@tiagotiagot]

Would it be possible to make the eletromagnets double as "metal" detectors, to fine-tune to magnetic field to get exact amounts and more gradual changes?

[@Chryoss]

Eyegasmic !

[@reedpropes2037]

Pac-Man!!! Or galaga, but you’d have to have a reservoir at top with a larger tank by flipping and leaving on a few rows of magnets then dragging a few at a a time down, but I’d LOVE to see galaga. I’d also love to see live/ wireless phone programming on this or just simple drawing on the phone that the ferro display replicates

[@YeshuaGod22]

would a small container of ferrofluid mounted between the pickups on a guitar react to the field distortions caused by the strings vibrating?

[@davidprock904]

Instead is setting up the electro magnetics in a square grid, put them in a honeycomb/ hexagon pattern. ALSO! will the ferro fluid flow up a capillary tube? If so you could always have some available to appear from the top!

[@AppliedProcrastination]

We tried a hexagon pattern, but it made the movement algorithm a lot more complex, so we decided against it for the full size display.

[@davidprock904]

@@AppliedProcrastination what about the other half of my idea, using a capillary tube for each column of electro magnet, have an overhead magnet to pull it out of the tube, and then unpower it to let what you pulled out the tube to fall

[@AppliedProcrastination]

It’s a bit too complicated for us to realize it in practice. Not sure if it would work, but I’m fairly sure that WE wouldn’t be able to make something like that with our resources and knowledge.

[@davidprock904]

@@AppliedProcrastination so take a different approach, have a small pipe like fish tank tubing, just something shaped similar to a candy cane, for each column of magnets, keep it in the back, and have it curve over the top and slightly downwards. Have permanent magnets always keep Ferrofluid in the vertical and slightly covering the top of the candy cane, and an electro magnet on the downwards slope/end. When you unpower the electro magnet, sure some of it will pull back to the natural magnets, but gravity will also pull some out the tip of the curved end! Let me know if you like this idea!

[@monberg2000]

Came from Wintergatan - will stick around :)

[@AppliedProcrastination]

Glad to have you here!

[@ВладимирЗенин-ч4к]

Hint: keep the ferrofluid storage on top and release a small amount just above the pixel you want to turn black

[@CJT3X]

How do you get it back to the top when power goes off?

[@danielawesome36]

In an alternate timeline, this was how televisions worked. Until the discovery of LED's that is.

[@g.o.j._productions]

dude iits been 3 years.

[@drcaiius]

fascinating! re game design, it looks like you should be able to use the outer ring of magnets as a border to redistribute the ferrofluid from the bottom, thus opening up a lot more design options.

[@YukioShinoda]

Really interesting project. I feel something life in the movement.

[@guitarBoyJunior]

Amazing job, guys! This looks truly amazing :) My suggestion for a game is "Blobby Volley". It's a relatively simple but hugely fun game. Although diagonal ball trajectories could be a problem... Anyways, good luck at NASA!

[@AppliedProcrastination]

Yep, we should’ve probably mentioned that diagonal motion doesn’t work... Only up/down and right/left. Thanks!

[@theoriginalmichaelvincent]

@@AppliedProcrastination Yep, we should’ve probably mentioned that diagonal motion doesn’t work... YET! (There I corrected this for you. ^_^)

[@AppliedProcrastination]

Lol, Thanks 😂

[@johnacsyen]

Awesome, subscribed

[@voxeledphoton]

Man I'd love to make an upside down tetris with this :V

[@FiZiX610]

Could it do a version of Pong? Snake would be cool too. Or maybe a very simple 1-button platformer.

[@jeffersonsilveira1452]

Have you tried multiplexing the magnets? I thought if you could do that, it would reduce the pcb size a lot. The magnets act like low pass filters so I think it is possible

[@AppliedProcrastination]

We loose too much duty cycle by multiplexing, unfortunately. But it’s a valid suggestion

[@Sideshowspike]

Such a cool and amazing thing you've made. New subscriber here, brought here by MMX and Martin at Wintergatan

[@AppliedProcrastination]

Thank you! We’re stoked to have you here :)

[@Sideshowspike]

@@AppliedProcrastination I"m stoked to be here, just my type of stuff!

[@isaac879]

I love the display! I was wondering if you have tested using PCB electromagnets (where the magnets coil is formed from the PCB tracks)?

[@AppliedProcrastination]

We haven’t tested it, but we have had some dialog with Carl Bugeja to see if we can test his PCB magnets

[@isaac879]

@@AppliedProcrastination I've subscribed and would love to see if you get any useful results!

[@ChrisM243]

Wintergatan brought me here as well. I immediately thought of "World of Goo" seeing this Ferrofluid.

[@HyperIonMake]

Dude holy crap. It looks like its a bad cgi animation, but it's real.

[@AppliedProcrastination]

Haha! Good luck to the CGI animator who wants to make ferrofluid, I guess? 😂

[@greggv8]

How about a game of "Angry Ferrofluid"? Have a tower on the right and a blob on the left. Pull back with a joystick then let go. The blob bloops across the magnets in a very crude approximation of a parabola until it either hits the tower or impacts the 'ground'. If it hits the tower, the magnets above the impact level get turned off.

[@AppliedProcrastination]

Hmm. That could possibly work, but it would be hard to detect if ferrofluid had, unintentionally, fallen off a magnet - making the tower invisible. Would probably need a visual feedback system for that (which would be nice for other purposes too)

[@dunkdon158]

It would be very hard, but in the long distance future, if you guys could make even smaller magnets, there would be more “pixels” and the ferrofluid would not have to jump from point to point I could crawl. Also, to prevent dropping, what if magnets at the top of a blob held slightly stronger than the one below it and so on?

[@dominicmillerca]

This is totally crazy, last week, I saw a bottle of ferrofluid on Amazon, the one with a little bit of ferrofluid in a clear liquid to play around with magnets. I immediately thought of making a magnetic display with it using electromagnets and maybe Arduino or Raspberry Pi. So I order a lot of electromagnets, I'm still waiting for them. And this morning, KZbin just suggested your video (maybe because I searched for a way to make a clear liquid to host the ferrofluid a couple of days ago... the most mysterious part of the project for me). Anyway, I didn't know someone already thought about doing this, I really thought I had an original idea... again. (facepalm) So, I've decided to build it from scratch without looking at your PCP design or your code. I want to see what I can do by myself and see how my version will compare to yours. I want to congratulate you guys for the idea and the project, it looks so good! The end result is very impressive and fun to watch. When I'm done with mine, I'll come back to share the result. ❤

[@AppliedProcrastination]

That sounds awesome. Best of luck with your project :)

[@Lumin3k]

Well This is Awesome !

[@daveayerstdavies]

Is it possible to match the density of the clear fluid more closely to the ferrofluid? What if the ferrofluid didn't immediately fall to the bottom? Could you clear the display by actively sending all the fluid to one end without help from gravity?

[@AppliedProcrastination]

We got some tips from a viewer yesterday on how to further increase the density of our suspension liquid to attempt getting neutral buoyancy, but we probably won't be able to try that in a while.

[@saumoon]

Venom's clock, this is.

[@YeshuaGod22]

maybe you could stick an array of nozzles along the base to quickly squirt ferrofluid upwards

[@fernandomartinez125]

OMG. This is soooo cool. I want one.

[@wenzhengqu349]

It's a great product, and I was wondering if it could be rendered in a three-dimensional space by different magnetic fields? Like a cuboid?

[@AppliedProcrastination]

I don’t think that would be possible. It could be possible to get a blob to hover in 3D space, but I can’t see a way to make 3D drawings possible, since it wouldn’t be possible to isolate a magnetic field to only affect a single blob.

[@umrkhayam]

This is so cool!

[@gazzacroy]

wow.. now that's really cool guys.

[@odinborson308]

Excellent

[@SuperMolo666]

Pong would be amazing on this display

[@wassup5196]

It's so satisfying and cool

[@JeremyChung]

this is so cool!! the back end electronics look so complicated, is there an easier way to do this?

[@xSaDii]

I'm sorry, newbie question. wouldn't it be easier to make 2d rails (Like the ones on leds displays) to control all those electromagnets instead of hundreds of wires?

[@unagisama5476]

Imagine playing World of Goo on there!!!!

[@davidgoff5883]

Does the Ferro fluid have to be black? Is there a way to colour it? Or add a metal fleck?

[@AppliedProcrastination]

I’ve seen commercial products with colored ferrofluid, but we haven’t experimented with how to actually do that. Could be a future project!

[@OliverSchmelzenburg]

super cool!

[@peterheynmoller2581]

really great project.. your wiring looks a bit complicated though.. wouldn't it have been a lot simpler to multiplex all of this? perhaps with diodes and capacitors over every electromagnet to keep the current a bit more stable?

[@AppliedProcrastination]

The wiring would still be quite complex in a multiplexed setup (all the electromagnets still need to be connected) and we found the duty cycle to be too low for multiplexing anyway.

[@peterheynmoller2581]

@@AppliedProcrastination yeah, that's true.. it just is a hughe mess to connect all of that! correct me if I'm wrong but wouldn't it be okay to compensate for the lowered duty cycle by simply increasing the pulse power proportionally? My thought would be that the limitation of the electromagnet is a thermal one (if you are not too close to saturation I guess?) So at a reasonably high frequency it wouldn't matter how much current you run through it as long as the average power doesn't exceed the maximum? I don't know that much about electromagnets so sorry if my thoughts are a bit dumb xD

[@xanderjames8682]

Theoretically could you make it so tht it climbs through a system and make it like a refilling reservoir for turbine generator?

[@davidgoff5883]

Led lighting to the screen in between the magnets?

[@AppliedProcrastination]

That’s definitely something we’ve wanted to do, but first things first at the moment. Got to get the display itself to the standard we want it first

[@DrOne-sw6oz]

GUYS RESPECT!!!!!! Amazing piece of ....Art?? :P fantastic. simple but really really coo land unique.

[@HmmmmmLemmeThinkNo]

I hope with all this stuff going on in the world that your internship has been going okay. Regardless, this has been a wonderul journey so far, and I look forward to whatever comes next.

[@nityaghosh5074]

I am loving it

[@AJMansfield1]

This could probably be done more compactly and at a higher resolution if you used an array of wire-wound inductors as your electromagnets, mounted directly to a PCB with the drive electronics behind the tank. Something like an RLB1014-104KL looks like a good candidate, 10mm diameter at only 20 cents each qty 5000. Standard ferrite inductors might not be arranged to project their magnetic fields as efficiently as purpose-designed electromagnets, but they'll project enough, and you could also significantly lighten the load by adding more salts to make the ferrofluid near-neutrally-buoyant.

[@AppliedProcrastination]

I love the idea, but I highly doubt that an inductor would provide enough lifting force for this. We have already over saturated the saltwater-brine to an extent where the salt creeps out through about a cm of silicon that’s coated with epoxy resin. Maybe it could work if we experiment with the chemistry until we achieve neutral buoyancy (which would be hard as we aren’t chemists), but I still think the effect wouldn’t be as cool with lower pulling force from the magnets.

[@AJMansfield1]

@@AppliedProcrastination For neutral buoyancy, the way to get a denser brine is to dissolve several different salts rather than just sodium chloride. Cody's lab has a video about how these sorts of multi-solute systems work ( kzbin.info/www/bejne/o6PcgpV5Z5p6b80 ), but long story short it's possible to create incredibly dense aqueous solutions if you use the right combination of salts. Saturated NaCl brine has a density of around 1.18 g/mL, but it's possible to dissolve nearly another 0.23 g/mL of ammonium sulfate without affecting the NaCl, for 1.41 g/mL. Swap sodium chloride for sodium bromide, and that 0.18 g/mL NaCl becomes 0.41 g/mL NaBr, to get a final density in the neighborhood of 1.64 g/mL with both that and ammonium sulfate. Just NaBr plus (NH₄)₂SO₄ might not be quite enough to get true neutral buoyancy with heavier ferrofluids (Ferrotec EMG 900 is 1.74 g/mL), but you only need to reduce the load not eliminate it, and you could absolutely float more typical 1.2 - 1.4 g/mL ferrofluid if you wanted.

[@AppliedProcrastination]

Interesting. I’ll keep this in the back of my mind until we decide to make a new tank. Thanks!

[@AJMansfield1]

@@AppliedProcrastination For the coils, running some back-of-the-envelope calculations the inductor I listed _should_ have a field strength of around 5000 gauss running at its rated 40mA, which isn't that far off from the field strength from the sort of magnets you used, and is well above the saturation point of even high-end ferrofluids like Ferrotec EMG 900. The key is to drive adjacent coils in opposite polarities, so their magnetic fields combine rather than cancelling out; the magnets you used are arranged to guide both ends of their field out the front, but here each coil is only has one of its poles pointed up. You could also add a metal midplate between the PCB and the inductors, both for cooling and to guide the field around between the bottoms of the coils (improving the efficiency). At 40mA and 300 ohms, they can each put out 480 mW of heat, so cooling is probably required in a dense grid; the way to do this is to drill clearance holes for all of the inductor leads to match your PCB pads, and then just adhere the plate to the PCB and thread the inductor leads through to solder. You'll probably want to add plastic sleeves to the leads to prevent shorts as well, if the plate you use is more than a few mm thick. Mu-metal would be the material of choice for magnetic reasons, although you could probably use some kind of flexible ferrite sheet to make it easier to handle. A slightly more intensive enhancement beyond that, is to grind down the tops of the inductors to remove the upper bobbin cap, once the board is fully assembled, to help the field project further. Once the board is assembled and the inductors soldered, you can pot the top side in some kind of rigid epoxy resin, and carefully sand the whole assembly down using a sheet of sandpaper and a flat surface.

[@AppliedProcrastination]

We're still skeptical that regular coils will provide enough pulling force, but we placed an order for 50 x the RLB1014-104KL you suggested - just to do some testing for fun. They seem like the only coils on DigiKey (in a small form factor) that can handle 12V without crashing and burning. But where did you get the Gauss-numbers from? We're already using the "checkered pattern" polarity-trick on our current magnets. That's a complete necessity in order to move ferrofluid between magnets.

[@ferrofluidsolutionprovider1865]

This product is great, and surprisingly, it may be better to use plastic containers because it is safer and more possibilities.

[@AppliedProcrastination]

What kind of plastic do you use, and what type of suspension liquid do you use? Have you written a writeup?

[@ferrofluidsolutionprovider1865]

The type of plastic can be AS ABS PP and so on. I have not written an article about these, I have a video to show. We provide commercial products

[@TouchWorld3G]

You are right, i really just want this to be a clock :D hopefully a lot cheaper. Do you know what the material cost of the first version are?

[@AppliedProcrastination]

Roughly $1500

[@honsonabader]

how this amazing video has so little likes and views????