PCB Electromagnet vs Ferrofluid

[@CarlBugeja]

Thanks for all your comments and feedback 😊 I will talk in more detail about this 12-layer pcb coil in one of my upcoming videos & I will also put it on sale on my website flexar.io/store/p/pcb-actuator-12-layer-coming-soon

[@spearshaker7974]

U were almost able to through a drop straight up and down if u could lunch one up u could maybe float it or disrupt in someway I seen potential in there somewhere.

[@athreyanm5243]

Hey carl you should check out ECM PCB stator technology. You can design your own PCB motor!😎

[@valsodar6723]

What if you submerce your invention into en.wikipedia.org/wiki/Novec_649/1230 ? Can I help with overheating ? Maybe is necessary use more heat conductive base of PCB...

[@AppliedProcrastination]

What a pleasure it was to collaborate with you on this project, Carl! The 12-layer coil was super impressive, and I can't wait to see where you take this design next!

[@PCBWay]

Caught you here ❤, Simen

[@WalidIssa]

Use two coils with differential driving

[@PCBWay]

Caught you here ❤

Wouldn't that cover the display?

[@derarchitekt4925]

Is the glass-layer really needed between the fluid and the PCB? Choosing a Fluid/[Film]/PCB/Glass stackup could improve performance. It also would allow dissipating heat into the fluid, creating some sort of passive fluid-cooling. The circulation of the fluid through the change in the displayed image would further increase the distribution of heat inside the display (fluid). This in turn would allow for a higher voltage and hence result in increased performance. Further ideas for improvement: - Remove the need for a glass-back altogether and instead use the PCB directly as the backplate. - Thermally couple the PCB with heatsinks using thermal-paste for even greater driving voltage (see channel "tech ingredients" for example on cheap thermal-paste).

[@darylh8657]

Is the glass-layer really needed between the fluid and the PCB? Probably. You should protect the traces from the ferro-fluid, which could short them out, depending on how conductive the micro-particles are. Perhaps a thin layer of lacquer would work.

[@absalomdraconis]

@@darylh8657 : Either your lacquer suggestion or a lamination sheet would probably be the right choice.

[@derarchitekt4925]

A thin film that is applied in solid form could bring manufacturing benefits.

[@darylh8657]

@@derarchitekt4925 "thin film that is applied in solid form" The tolerances on a solid substrate would have to be high in order to achieve a tight fit. I envision some sort of schmoo that can be squeegeed on like the varnish we used to put on photographs back in the industrial age.

[@derarchitekt4925]

@@darylh8657 I agree - possibly a good balance between simplicity and effectiveness!

[@etienne3838]

You could add a magnet that produces a magnetic field equal to what your coils can produce. Powering up your coil will double the magnetic field. Inverting the polarity of the power supply, will create an opposite magnetic field that will cancel the one from the magnet. The main drawback is that your coil needs to be powered all the time (either in one polarity or another) and it could be a problem regarding their temperature...

[@siberx4]

I think this is basically what he did when we added the magnet to the back.

[@etienne3838]

@@siberx4 Yes you are right!

[@billfan7528]

The doesn't scale as well because the drive electronics for a huge grid will get a lot more expensive. Currently, the Applied Procrastination derive elestronics just drive the magnets with PWM at a single polarity, which can be done with just an LED driver and transistor arrray. The total cost is ~0.3$ per magnet. To switch polarity you would need to drive the magnets with H-bridges, which makes the cost ~1$ per magnet. For a big grid tripling the cost is pretty substantial

[@siberx4]

@@billfan7528 The drive electronics are not the dominant cost here; it's the labour of hand assembling a huge array of discrete drivers with wires/connectors/etc. If you could manage to make a viable PCB array of coils, it would come out way cheaper for the whole device even if you had a bunch of extra costs for specific magnets and more capable drive electronics.

[@max_kl]

I think it'd make more sense to talk about current instead of voltage, since it's the current that causes the magnetic field. That should make the different coils more comparable

[@3DprintedLife]

This is great! Can't wait to see how far you take the design

[@PCBWay]

Caught you here ❤ ! BTW, we will be super excited to see you bringing us more new Kool projects as well, Andrew!

[@KX36]

As I have previously commented, you could try running FlexAR at 300kHz where it's mostly acting as an inductor, rather than DC where it's mostly acting as a resistor, to lower the heat losses. The caveat to my claim being that I only use coils as inductors and transformers, not as electromagnets. If you look at the phase angle when you show the LCR meter, the 2 layer board has a phase angle of 0 at 1kHz so it's pretty much entirely resistive. The 12 layer board has a phase angle around 10 degrees so at least that's starting to get a little more inductive at 1kHz and the best inductive frequency for that would be much lower than the FlexAR, although still over 1kHz. You could possibly try using the LCR meter to stimulate the coils at 1kHz, 10kHz, 100kHz and measuring the change in magnetic field strength between the frequencies on that other meter, although I'm not entirely sure how that particular LCR meter works, whether it is actually running at a simple single frequency as displayed or the waveform is more pulsatile/complex. Can't compare to putting a high DC current through them though as that's totally different than the low current stimulation from the LCR meter.

[@absalomdraconis]

You have to be careful using AC waveforms for this kind of thing, because at some frequency it'll cause repulsion or something similar.

[@andershestad284]

You could try to thermally couple the metal backing to the PCB with some thermal paste perhaps? If you're able to make the metal plate work as a large heat sink then you might be able to drive the magnet with a high voltage while keeping temps under control?

[@Reptex_cs]

Adding ferromagnetic metal to the center will improve the magnetic field quite a bit. Or possibly a thin ferromagnetic disc between the coil and petri dish? Might work too.

[@vitorhearteater]

I was thinking about that. The problem with ferromagnetic core he showed in the older video was how to place the core in the flex coil, but with the 12 layers been stiff would'nt be possible to place a core now?

[@AdityaPrakash-kt3rf]

Awesome collab Carl! And the new PCB coil has more green, which I always love!

[@veloxsouth]

When that intro started all I got was *word sounds* piss see bee coils *word sounds*. I understood the second time, but this guy's excitement for his project was really showing.

[@Culturedropout]

I always enjoy your videos because you're so methodical about your investigations, but are still excited and enthusiastic about your work. Watching you feels like watching science (or at least engineering) happening.

[@kevinbissinger]

can you put the coils inside the tank so the liquid keeps it cool?

[@dimitarnikolov3527]

I think you can design coils that go directly inside the ferrofluid mixture. That way the coils are both physically closer to the ferrofluid and water cooled. The solder mask would electrically isolate the windings from the wetter and the solder joints would be above the reservoir.

[@curtisbeef]

What about a hole in the center with a ~2mm thick piece of steel?

[@CarlBugeja]

I have tested this in this video kzbin.info/www/bejne/jpe5q2yFqd2tr9E - the material was not steel but back plating produced similar results

[@curtisbeef]

@@CarlBugeja ah shit. I must have forgot lol 👍great content as always👍

[@albygnigni]

@Carl you can join the 12 layers and the steel core concept

[@HeyBirt]

I wonder if you could deign your PCB electromagnet to incorporate the Halbach effect? (as a Halbach array) This would let you direct more of the flux in one direction...

[@johnnycernato4068]

Maybe he could just implement the horizontal components as neodymium magnets and switch the vertical one?

[@PCBWay]

Caught you here, Jeff ❤

[@Morgernstein]

Had the idea for trying to make a transformer from something similar to your pcb coils, just with a hole in the centre for an iron ring to be the core. Then to actually have a difference in turns you either use two different coil layouts or using two different coil pcb thicknesses

[@mth469]

what is the purpose for the little holes in the golden areas of the board? is it to release pent up heat ? i also see them around plated screw holes (vias) at the edges of boards.

[@justingilmore2176]

What software is that at 2:46? Does this allow you to simulate

[@immernochanders776]

you could make a pcb thatr ataches at the back to an heatsink that you can cool with a fan or just make it big enough that it can disipate the heat. with this method, you could potentialy run the pcbs at a higher voltage.

[@AlexanderNielsen.]

Could this princip be used for a flipdot display? E.g. To lower the production price and make it more compact.

[@debashisdas7557]

If I am thinking r8 ten you are thinking about a panel of this PCB to control the final display, why not use PCB Thermal Vias to channel the heat to a separate heatsink.

[@luiswhatshisname7667]

Can you just stack the coil squares with lower currents to avoid heat and get stronger fields?

[@dianventer1791]

Why can't you use a 12V supply and just mount the PCB to a heatsink? This will make the final assembly neater as well since you'll need some way to mount the PCB anyways

[@CarlBugeja]

The back-core used in the video in a way acts as a heatsink - I have talked about this a bit in this video kzbin.info/www/bejne/jpe5q2yFqd2tr9E

[@dianventer1791]

In this application active cooling with a fan could also be an option since thickness is no longer limited. Thanks for the great videos!

[@yorgle]

What about putting the electromagnet coils right into the liquid/water... or perhaps in their own pocket inside the fluid container, but in like a plastic bag to isolate them?

[@Roobotics]

Maybe route out a hole in the center for an iron slug to seat into? Now that you are trying to make one magnet instead of a PCB motor for this project.

[@Jandodev]

This is amazing! I strive one day to make pcbs like you!

[@DanielSimu]

Do you think this pcb could drive a flip dot display?

[@Goodgu3963]

Have you done any experiements with how much force these can generate perpendicular to the PCB? I've been looking into a project of my own using flat linear actuators and some of your flexible actuators are similar to what I'm looking at. Awesome videos though, definitely earned my subscription. Can't wait to see more!

[@andymuzzo8568]

Another great video mate.

[@anaanshahid998]

Carl I have a question. can we make a really small PCB coil?

[@andycrask3531]

Just my two cents worth but would having round steel instead of rectangular help with the generated fields?

[@MegaBlackJoe]

Hi Carl, can you make a video on smd connectors for flexible PCBs?

[@timtrix1449]

Hey, nice video! Can you tell me which program are you using to calculate the coil turns?

[@youssefaly97]

Would you try high current coils? Thicker traces for more current and maybe continue the turns in series on other layers! I use DC-DC converters from Murata that output 12V at 50 Amperes and only use PCB trace coils for the transformers. They have 92.7% conversion efficiency so not so much heat is output. Granted the PCB is a lot thicker and has maybe 16 layers or more but still 50 Amperes! Keep on going, I love your work!

[@andrewkelley7062]

If something can extend magnetic fields then it can usually sense through them. Maybe you should look for a configuration where you can detect changes to the field. Probably through back emf.. The thing is if you buy last time I looked about $100 in magnets and had them mounted attached to those boards in the right way you could use the existing perminit magnet as a base line then the board as a secondary, then have the static be the physical object and have a rough cut mri sensor. No where near medical standards but maybe objects or given some work hidden metal frameworks like curit boards under plastic.

[@soupernerd8975]

You need to build the coils with a hole in the center so you can introduce iron in the middle.

[@musegeek]

What about series / parallel, like a 4x12 guitar amp cabinet?

[@michaelschecker2716]

Use more layer and may be use some layer (100% Cu) to spredd the heat to outside . Outside lots of via . Cause the coil is used with DC there is practically no magnetic short

[@IgnacioMoya]

Have you tried making your coils with a thicker pcb trace thus lowering its resistance? Maybe that would allow to push more current through your coils and increasing magnetic field. I will also suggest driving your coils with a current instead of a voltage source since current is what causes the magnetic field and you are not depending on resistance variance.

[@freedomcaller]

What if you put 2 series pairs in parallel?

[@youssefaly97]

Carl, how easy is it to route arc-like or circular tracks in Altium? Eagle in far from easy when it comes to circular or arc tracks. This is for a circular PCB

[@CarlBugeja]

It's pretty easy.. You can also design the track in cad and then import a dxf

[@youssefaly97]

@@CarlBugeja Thanks! That's more persuasion to move my 10+ years of Eagle background to Altium!! (which I tried by the way)

[@krit6847]

this is so exciting!!

[@trench01]

you can use 2 magnets if you use the tractor beam magnet.

[@MsUnderattack]

Your argument on coil stacking is somewhat off the point. The current over the coil decides the magnetic field intensity. This means that trying to drive coils with different internal resistances at the same voltage gives a misleading result. This is seen with coils connected in series, it doubles the resistance but you should keep the current the same so that increase in resistance shouldn't matter. Since temperature depends on I^2R and the cooling power of the coil, and you doubled the coil (cooling surface) and R, the temperature should stay constant. For example consider a coil with very low resistance. In your explanation, it would have lets say 5V over the coil, and a large current through it. That large current would be squared so it would cause high heat. But this kind of coil should be more ideal not "bad because it heats at 5V". You can pass more current through it because it will pass the same current compared to higher resistance coil with less voltage drop so it will heat less.

[@MsUnderattack]

Some intersting trivia: You can even write the specific heat formula and find a differential equation for the heating action under constant current. You will see that there exists a current value that makes the system thermally unstable. (Formula is in the form of R=...e^(-t/tau)+... and you see that over some current tau

[@Pixelcrafter_exe]

Maybe it's somehow possible to submerge the pcbs into the fluid to cool the magnet and to get rid of the layer of glass

[@cosmefulanito5933]

It has very important misconceptions. It is true that if you connect two coils in parallel, the current doubles. Simply download it from the circuit that excites them and voila. The same goes for series coils. Turn up the current of the circuit that excites them and solved them.

[@nadiaplaysgames2550]

what if you stack 4 coils two series set in parrle

[@mikolavision]

engr to engr: excellent work!

[@bloodlinegaming9962]

cant you send some voltage through the fluid? you should be able to make a modular circuit like that with liquid switches lol would be so cool

[@motherjoon]

Is it possible to stack the 12 layer in series and power with 12 volts?

[@motherjoon]

It may be possible to have soldering gold fingers purely for structural support to stack multiple of these coils together. A huge stack of them (50+) may even make a great tesla coil secondary.

[@CarlBugeja]

They can be wired in series but like I explained the series resistance would created a drop in the field

[@livdbest2942]

Great video 👍👏👏

[@jimjjewett]

I wish there were a way to offer corrections to subtitles. it took me several times through to figure out why it mattered that this was his first "red layer board", but the drawing indeed had red tracks. Eventually, I realized he was saying "twelve-layer board", and the automatic caption generator was even sloppier than my ears.

[@CarlBugeja]

It should be update now - sorry for this

[@jimjjewett]

@@CarlBugeja Alas, this is a problem on many channels. I just mentioned it here because "red channel" with red tracks had me fooled for more re-listens than usual. But thank you for the extra care.

[@ConsultingjoeOnline]

Awesome 👌 👏 👍

[@philsponsel3824]

Hi Carl :-) i wonder what a Peltier cooler + your 12 layer Flexar on 12V can archive? Or better saying general cooling your flexar and pumping up the volts. Another idea about cooling (looking at the ferrofluid display) is that you just put the flexar in the liquid so that the liquid is cooling the flexar and you can pump up the volts. idk what liquid is used and if it would work but it may be worth a try? I'm sure that one day you'll find the perfect use for your flexar and you'll revolutionize the (magnet)world :-)

[@Beregorn88]

I don't understand why you use a constant voltage source, instead of a constant current one...

[@samuelstanfield3029]

could you try to make a ferrofluid speaker

[@jasontemlett]

The moment when you pop onto KZbin and somebody's busy working on the exact same project as you are

[@csabaczcsomps7655]

Put one ligth source back to display, use smal leyer of ferofluid, ligth can pas , the buble is the pixel, You got 2 color display

[@trench01]

the spiral is not efficient since you counter the rotating magnetic from one coil to another to cause friction which results in heat and thus inefficient. do that on gears and it wont work as efficient as well. You need another coil in the spiral to take the emf directly and you can just let it go nowhere or redirect it maybe from behind to give an additional boost. OR maybe a tesla coil might help. Sure it will be less output but might be able to handle more input.

[@matjazwalland903]

Hello I looked at some of your videos but I miss what your goal is. Do you want to increase the power of magnetic field on a large surface or in small? Regarding the magnetic field increase, it would suggest that you change the clean copper for a mixture of copper and iron. Such Materal reckon with acid should therefore be necessary to handle the circuit. Another idea would be to do the holes in the circuit. Through which a metal tile would have to come out of the bumps. The layout of the holes in the circuit I leave your imagination

[@shallowthunder2972]

are u from malta

[@___aZa___]

Yay another Video ^^

[@szymongrzelak1658]

Well, you could try going full Intel/Nvidia/AMD, add cooling and 24V.

[@espero_dev]

Hey just put more voltage through it that’s how I make super magnets it might work for your project

[@KillerKaninchen]

Can you pulse Electromagnets like you can do with LED'S

[@hitendrasaini3465]

You are always a genius, I think you have to work on nano technology, The time need this technology, Make all your ideas which can only be seen under microscope, For e.g make a spy quadcopter of smd size, You can power it my a smd peltier module or smd solar panel which work on sun heat and light. Make a smd coil and smd propeller, Think if your nano project work on sunlight than how much time you can backup your nano projects.

[@Derlaft]

Can you put PCBs inside the fluid? That will cool them down naturally :)

[@henrychan720]

Just strap a heatsink to the back?

[@alluriman]

why not use them as actuators and just push small magnets onto the glass

[@Sir-Dexter]

How about a PCB Electric Generator

[@zerstaerker]

dammit how cool is that? how about replacing coils in general that way?

[@DissanayakeDinal]

Cant you use those mini magnetic balls

[@thestariusproject]

OHH I tested it this weekend with your PCB and I have no good result ...

[@OualidMars]

NICE BRO ❤️♥️

[@Jkauppa]

multi tube frequency analyzer audio microphone/speakers

[@k.kannankrishnaswamy95]

I want five coil

[@DarkDragonEgg]

Layer two in serial and two of those in parallel (:

[@denizzzka]

Ferrofluid can be "tuned" to any buoyancy. This guy made a ferrofluid with almost zero buoyancy: kzbin.info/www/bejne/fYGqeqiJeN14fNU

[@MECKENICALROBOT]

ʕ ͠°ʖ̫°͠ ʔ⸝^nice! I love your thinking…. One of these days I’ll jump back on my college ferrofluid project, I’ll reaching out to see what hardware you’ve made by then.

[@Niohimself]

Epilepsy warning at 3:00

[@kelghou2437]

cool video

[@Abell_lledA]

One is caught up playing the impassioned protagonist in one’s Subjective Narrative of Self🎈

[@raspberrypi4970]

Distortion from a Ferro BlackHole

[@vaposhack]

Sorry man, unsubscribing due to KZbin's decision to hide dislike counts. Hope to see you on another platform.

[@LambGoatSoup]

Is that it? Do way higher

[@AJB2K3]

Star Trek Fans Are going to Hate you, Marvel fans are going to love you.

[@ATLTraveler]

Wtf are you saying I can't understand anything... also what is a pissibe

[@ferrofluiddisplay8107]

🖤

[@Electrons1]

4.

[@kubeek]

Uhmm, what is a pissiby?

[@antumurikks4861]

if i like we can make world. just say yes

[@VoltsandVodka]

I always hear furry-fluid

[@victorschlim4644]

man, could you please sponsor me an altium license? lol Cannot afford even with 30% off....

[@rafaeltavares6928]

Dude... Adding an actual magnet defeats the purpose doesn't it? 😑

[@WaschyNumber1]

🖖 👍

[@an_what]

Ah hey guy why not epoxy the magnets into place so they don't attract one another I just solved your problem without watching the video

[@mohammaddarbandi1303]

#altium

[@aggabus]

Man fix your accent