The MOST DANGEROUS Fountain in the World

[@lajoswinkler]

Hats off. For all I know, this is the first time anyone did it on KZbin. It is a very demanding project requiring knowledge in several fields and not many people have it. Even less have both that and the time. Beautiful contraption. Regarding mineral oil, just abandon it. Mineral oil is used for shipping. It got popular by "eBay China sellers" because it's difficult to ship alkaline metals in a container that smells of hydrocarbons. Golden standard for keeping alkaline metals is purified kerosene. Reflux a small amount of NaK in kerosene under argon for an hour and you will get kerosene completely inert against NaK even under higher temperatures. That way, even if there are small amounts present in the fountain, it won't form dross.

[@hardwareful]

it's even more beautiful when you fill the fountain with mineral oil instead of Argon, then everything happens in slow motion edit: as for the electrodes, use an array of electrodes on each side.

[@cda32]

You need a current limiting supply! That one is just rated to 40A, but it is a constant voltage supply.

[@christiannorf1680]

That's probably the loveliest fire hazard I have ever seen

[@DanBowkley]

Make the terminal blocks much larger and with heatsink fins on them. They already have to be in really solid contact with the electrodes anyway, might as well use them to remove heat too.

[@drflash36]

And or, place a cooling fan or blower near the electrodes to cool them there as well.

[@TheFreak111]

@@drflash36 fan adds noise, I would say it's nice to hear only the liquid metal splash.

[@Enjoymentboy]

We've all seen the videos where a large chunk of sodium or potassium is tossed into a pond but what I've always wanted to see is a large piece, say 1kg or so, sealed in bottle with a sinker is dropped into a large body of water and the bottle burst open around 50-60ft underwater. We have seen so often what the reaction looked like when the metals are dropped into the water but I've never seen it where they are quite deep before they make contact with the water.

[@AdvancedTinkering]

I would love to try it. The problem is finding a body of water that allows for this experiment without violating any regulations. Maybe some time in the future there will be an opportunity.

[@littlegrabbiZZ9PZA]

​@@AdvancedTinkeringI got a boat, goes more than 12 miles from shore.

[@TheRealBanana]

After WWII the US had so much sodium metal laying around they decided to just toss a lot of it into a lake, in barrels weighing 3500lbs each. Theres some very old footage on youtube of it. Would love to see someone recreate that! 😆

[@craigpater6278]

​@@TheRealBananaI have seen that video on KZbin it didn't take me long to work out what would happen when 3500 pound containers of sodium metal crashed through the ice on the surface of the water in that video and the 3500 pound containers of sodium metal reacted with the water

[@craigpater6278]

​​@@AdvancedTinkeringI think that would be very interesting if it could be done since everyone who knows what alkali metals like sodium and potassium are knows how they react when they are on the water's surface but I don't think that that much is known about how those alkali metals will react with water if they react with water while they are deeper underwater when they react with it. I can sum up how alkali metals like sodium and potassium react with water simply for anyone who isn't familiar with what alkali metals are a large piece of sodium or potassium metal plus water equals explosion

[@marshallscholz8629]

A high current buck converter in constant current mode would work well for this. The different pumping speeds actually makes the fountain more interesting and you could connect a microcontroller to vary the current. If you want help choosing or designing one I can help.

[@ryanatkinson2978]

4:47 when the Terminator has a happy ending

[@berukadehikari2634]

This looks so cool! Now imagine a NaK lava lamp. Maybe with an oil with similar density as the NaK and a warm lamp at the bottom, it may be possible.

[@AdvancedTinkering]

Interesting idea! I will take a look into it!

[@jimsvideos7201]

That is gorgeous and terrifying.

[@阳明子]

The dark stained wood with the copper and liquid metal looks so good.

[@robertlapointe4093]

Assuming that your fountain is made of borosilicate glass (Pyrex, Schott, Kimax, etc.), there is an iron/nickel/cobalt alloy called Kovar that has the same coefficient of thermal expansion and is used routinely for glass/metal seals. This could be used to make larger electrodes to handle the high current of your fountain. For example, a piece of Kovar tubing (say 6 mm outside diameter and 25 mm long) could be sealed on one end by brazing to a piece of copper rod and the other end could be fused to the glass of the fountain. (You might want to check the compatibility of both copper and the brazing alloy with NaK.) As for the power supply, a simple mains frequency transformer with a full bridge rectifier would probably be best. For example: a Hammond185G10 (available from "Element 14" (aka Newark electronics here in the US) can provide 35 amps at 5 volts and can accept either 110 or 220 volts input for about $50. For that transformer, a Vishay PB4008-E3/45 (40 amps at 800 volts) bridge rectifier would probably work and is about $4. Maybe add a variable auto transformer (Variac) on the input side to control output, or a power resistor on the output to limit current. Your fountain might sing (at 100 Hz), if so, then an output smoothing capacitor could be added.

[@AdvancedTinkering]

I looked at a few alloys suitable for that purpose but it's pretty hard or expensive to get them in the right dimensions. Thanks for the information about the power supply!

[@THYZOID]

Very cool!

[@AdvancedTinkering]

Thanks!

[@janewray-mccann2133]

Interesting alkalis, one positive one negative. The same alkaline potion (5 milligrams) makes up your sino atrial valve contained in our hearts. It maintains our static electrical link to our brains, as well as controls the constant pulse of our beating hearts, generating a voltage of 0.1 volts. It is our own battery on legs and last's a lifetime. These two toxic alkalis are the only reason we each and all exist. Interesting choice of lightweights you made, my favourites by far.

[@RaExpIn]

What an awesome project! So scary and beautiful at the same time... Looking forward to the next version! :)

[@AdvancedTinkering]

Thank you! I got the idea watching a video by "periodic videos" where they showed a short video of a fountain like that. But unfortunately it was destroyed.

[@RaExpIn]

I remember watching their video, back then, but as far as I remember they only had somewhat old footage of the fountain running. @@AdvancedTinkering

[@tracybowling1156]

That's really really cool!

[@Ziraya0]

Conductor heating is a function of resistance and current, resistance being a function of length; since you need to have such a high current, you should endeavor to minimize all distances you can to reduce heating, and the power loss that causes it. Tungsten looks to have a resistivity 3.3x higher than copper, so shortening the tungsten distances is worth 3.3x more. This is all to say, you should get those copper blocks as close to the glass as you can, to minimize the length of tungsten the power actually needs to travel through. It will also help to have the thermal mass of the copper block closer to the tips, so they can sink heat out of the tungsten rods more readily. A small fan would also be a good idea. I suspect the power supply is not very happy being in that box, you could put a small fan directly under the fountain jet blowing up along the bulb. Even if this is set to a low speed so it's not noisy, forcing the air to move at all should have a substantial impact on the duty cycle. A small universal motor fan would be simpler, since you can wire it into mains, but a lot of people aren't comfortable with touching mains even a little bit (which is fair).

[@benmcreynolds8581]

This is amazing! It's awesome to see the final project after seeing the beginning video awhile ago. With the custom glass blowing part of it. I love liquid metal. It's such a fascinating topic. I'm really curious to see what we can learn about how things behave under immense pressures

[@frankhaese_DrHaeseGroup]

It is a beautiful piece of chemical art. There is always something magic about NaK2.

[@FriedrichWinkler]

To cool the electrodes you could use bigger copper attachment chunks that are finned for surface cooling. As others have said ensure low resistance contact to reduce wasted heat. If you are willing to change the aesthtics you could add a cooling fan next to it or a ducted cooling fan in the base with the ducts routing towards the electrodes the tungsten doesn't care about the tempreture so your tempreture limit would propably be the liquid. Lastly instead of using a dead short power supply it could be worth investigating a high speed switching power supply that applies power with a PWM method that way there is less heating and the momentum of the liquid should make it look the same.

[@ViniciusVetor]

Use a higher voltage power supply (12V) and a cc-cv buck converter. You will be able to adjust the current having the voltage as consequence.

[@realedna]

FIrst he should measure the resistance of the load (fountain circuit). Then he can figure out how much power he needs for a certain current level to select a suitable power supply. Most buck converters would probably only go as low as 1.2V or even higher though, which might be already too much, as they measure against a voltage reference at that level. The CPU supplies found on PC motherboards would be the right kind as they supply high currents at very low voltage from intermediate DC rails. So maybe one has to have multiple phases as well and put multiple supplies in parallel to support higher currents.

[@zhou_sei]

very beautiful!

[@AdvancedTinkering]

Thank you!

[@dav1dsm1th]

How about using electromagnets instead of permanent ones - and route the circuit through their coils before going through the electrodes? This could increase the resistance - so the power supply doesn't trip - and maybe reduce the temperature of the electrodes too. Just and idea. Thanks for the videos.

[@deltab9768]

I was thinking you could use two MOT’s modified into electromagnets (with the secondaries removed and the bottom cut off the cores and used to complete the magnetic yoke of course.) And then have them in series with the primary of a third MOT (with its secondary removed and replaced with a single turn, two parallel loops of 4/0 107mm^2 that feed the actual NaK current.) I’m thinking that would be a forceful potassium jet indeed.

[@peter360adventures9]

Beautiful but dangerous. Awesome and nice to see your 3rd version.

[@brfisher1123]

The alkali metals have to be one of the most gorgeous looking metals (along with gold) on the periodic table especially cesium when unoxidized of course lol 😂

[@Grateful.For.Everything]

Dude, this hits deep, how cool is it that we can take two solids and combined they just transition right on over into liquid!!! That’s wild, that’s incredible! The shit we’ve figured out Man, Thank You for sharing, the fountain was super cool!

[@robot797]

I so wish I could have one of those!!!

[@DuckPerc]

This guy watches the same youtube woodworkers I do and I'm here for it.

[@AdvancedTinkering]

Haha, probably.

[@CatboyChemicalSociety]

railpumps are amazing

[@Fluorineisgreat]

Very nice!

[@AdvancedTinkering]

Thank you very much! However, I'm a bit embarrassed that I didn't put more effort into transferring the metal under Schlenk conditions and without oil into the fountain. I definitely could have done that better.

[@NonEuclideanTacoCannon]

I remember I used to see ads for a product for de-soldering heat sensitive components. It would form an alloy with the solder that had a drastically lower melting point. I was going to point that out as another example of a eutectic mixture, but on second thought I think most solder is already a eutectic mixture.

[@flomojo2u]

Moving to thicker tungsten TIG electrodes seems like the best option unless you wanted to mount a fan directly below the assembly (In the base), and cut slits or just a large round hole to blow upwards. I'm afraid you're probably going to need a larger PSU, though going to a linear/transformer supply would likely be the easiest. When trying to create very small voltage drops for high power laser diodes, I've used lengths of fairly thick solid wire to provide a small resistance without building up too much heat. You really need to measure the resistance of the fountain with a special 4-wire low-resistance meter (Search for YR2050, it's the cheapest option, though definitely not super cheap), which can measure fractions of a milliohm. Any multimeter is useless for such low resistance. Once you have the resistance, you can see what the peak current is going to be and how much additional resistance you can add without dropping too much power for the fountain to function.

[@martys9972]

As an engineer, I have never heard of a "left-hand rule". The normally taught right-hand rule is based on current being defined as the flow of positive charges. The triad that you show for the relationship between current, magnetic field, and force is only correct if current is redefined as the flow of electrons. In that case, the triad that you show would be correct.

[@AdvancedTinkering]

Yes, there are two definitions. The "conventional current" assumes, that the current flows from the positive terminal to the negative terminal. As you said, the electron flow is actually in the opposite direction. Depending on which definition is used, the right-hand rule or left-hand rule mist be used.

[@elitearbor]

I enjoy watching videos like these, of experiments I'd never in a million years try. From a distance. On the other side of the planet.

[@unusualfabrication9937]

"it's not mercury - it's worse!"

[@lambda7652]

To address the mineral oil. Would it be possible to flush it a view time with a solvent and and then evaporate the last bit s of solvent under vakuum? Have you Considered using no electrodes and power this only via induction. like in a Asynchronous induction motor? As feasibility test could you place a vile full of NaKa in Stator from a 3 phases Induktion motor (with the rotor removed) and see if its starts spinning?

[@T3sl4]

Would be interesting to have current flow between the electrodes and simultaneously magnetize an electromagnet wrapped around in place of the permanent magnets; the MHD version of a series-wound or universal type motor. This would take some machining to build such a magnet though (effort which could be used to improve the permanent magnets, just as well!).

[@slyfoxchemistry]

It is super amazing I love wat ching it I need to make one

[@nettobin]

A better way of transferring liquids without exposure to air is to use a double ended cannula. Place one end of the cannula into the air space above the NaK in a suba sealed jar, which will have a positive pressure of argon. As argon is flowing out the other end of the cannula place the end into the suba sealed fountain with the needle air relief (as you have). When argon flushing is complete simply lower the end of the cannula into the NaK and the pressure of argon will pump over the NaK into the fountain. No glove box required! Hopefully I have explained this well enough, as it is a very useful technique to have when handling air sensitive materials. Edit: I just googled it, its called a cannula tranfer if you want to see pictures.

[@bite28]

Ttheir band is pure rebellion! Next reaction - October Ends Dark 😍

[@andrewevenson2657]

Any specific reason you wanna use sodium instead like mercury or something? Is mercury so heavy it’s hard to pump, or just like the idea of the metal being flammable lol.

[@AdvancedTinkering]

You are right that mercury would be too hard to pump due to its density. The other reasons are: 1. I like the idea of a flammable metal being used ;) 2. I would rather work with a non toxic flammable metal than mercury. I know that mercury metal isn't that toxic, but it gets everywhere if you spill something and it's a pain to clean it up.

[@MysticalDork]

The oil contamination problem could be reduced by using a lighter hydrocarbon (such as hexane) and vacuum-distilling it off after adding the nak to the pump. Even mineral oil could probably be removed this way, although it would likely take a long time even if you were to heat the assembly to 150c again. The nak itself wouldn't care, due to its extremely high boiling point and low vapor pressure. You may be able to use multiple smaller tungsten electrodes for better current-handling, either grouped together or maybe as several smaller "pump" sections and run them electrically in parallel but fluidically in series. That combined with some passive heat-sinks on the outside should reduce the heating issue. I would be happy to help make a yoke for V3, I have access to a milling machine and other metalworking tools.

[@me0101001000]

This is the scariest fountain I've seen in a long time

[@kemster9495]

This is very interesting have you tried any Alkaline metal isolation like magnesium or calcium.

[@AdvancedTinkering]

Thanks! Not sure if I understand you correctly but I isolated cesium and rubidium by reducing their chloride salt. I also made a few videos about that.

[@fredrikkleven9913]

I think you could solve the electrode problem by using multiple small tungsten rods instead of just one. If you need a focused point you could weld them to a single large tip and just have multiple small rods where you seal.

[@DisorderedArray]

You could just make a big batch of nak in a schlenk flask under nitrogen and without oil. It would react with moisture and any stray o2, but the oxide normally forms a film on the top and on the glass surface, so you could then syringe pure nak from within the bulk liquid, via a cannula through a rubber stopper. Then use schlenk technique and draw in some shield gas, and do the transfer. Argon would also be much better than N2. Also, I would say a heat gun is much quicker and more effective than the heating wire for pre drying the glassware.

[@AdvancedTinkering]

Yes, that's my plan eventually. Right now I sadly don't have the space to set up my schlenk line. And working only with seperate valves will be very inconvenient.

[@jonathanwienke512]

Instead of having two thick electrodes, consider using an array of thinner electrodes, perhaps three or four in a row on each side, so that each electrode is only carrying a portion of the total current. A current-limited power supply that allows you to set the maximum current is also a better option than a voltage controlled supply for this application, as the pumping force is directly proportional to the current.

[@jonathanedelson6733]

Really beautiful. In addition to DC MHD systems, you can also drive liquid metal with an induction motor approach and completely eliminate the electrodes. But this would require winding coils and an AC control system. DC is far simpler.

[@pypes84]

If you'd stood the syringes on end for a few hours could you have gotten the mineral oil to separate out and then discard it before injecting the rest of the syringe into the fountain?

[@cheeserdane]

Making me think about Thunderfoots liquid metal flame thrower experiment.

[@Chor_Ultra_Pro_Max]

You can use a big 110/220V to 1.5V Transformer capable of producing at least 40A or you can rewind a random transformer (of at least 500W) with thicker wires to get the desired output and then use a rectifier with capacitors to get the required DC voltage and current. This way you can run your fountain continuously and avoid intermittent working because of the SMPS.

[@AdvancedTinkering]

I allready thought about rewinding a microwave oven transformer. But I need something with a small enough form factor to fit into the base. Do you have any ideas or even a model number of a transformer that could work?

[@UltrAgejmr]

I suppose you could use a ferrite core transformer rather than iron core, they are way smaller and lighter but also harder to drive.@@AdvancedTinkering

[@DrakkarCalethiel]

@@AdvancedTinkering A torroidal tranny might do the job nicely. Easy to put a couple turns with heavy wire onto it without the need of removing any existing windings.

[@T3sl4]

@@UltrAgejmr Which is what the MeanWell uses. The operating frequency is quite far from mains frequency, however, and thus the complexity. Hence it's worth buying a ready-made one.

[@UltrAgejmr]

​ @T3sl4 I mean yeah, but you can still make your own that fits the best for your application like 12 to 2 volt and even make it work on batteries.

[@T3sl4]

Would a wash with petroleum ether be feasible? Even just to thin out the mineral oil, would be a start. I suppose with a few cycles of adding it, shaking around, draining off the excess, then a final evaporation step under vacuum, it could get pretty clean? More opportunity for oxygen to get in, of course.

[@blg53]

You have a forth problem. The copper electrodes that are fused into glass will expand from heat potentially causing the glass to crack. And with the currents involved you can't avoid heating of the electrodes. So you can't run this device for long anyways. Also, related to the large currents, you can't ignore inductivity of the circuit, a similar problem to what spot welding devices have. So there would be potentially high voltage spikes at moment of disconnection of the current. So to protect your power supply from high reverse voltages I suggest you include a relevant TVS diode reverse-parallel to your fountain electrodes.

[@Amateur.Chemistry]

NaK is really just a hardcore version of mercury

[@mduckernz]

Sort of, yeah…. but sadly it lacks the rather enjoyable property of large quantities feeling uncannily heavy that mercury has!

[@geoffreyentwistle8176]

Given the terrifying toxicity of mercury, I'm inclined to think of NaK as the less dangerous metal... At least you can see the fire and run.

[@leonardotheuseless4188]

@@geoffreyentwistle8176 mercury in its metallic form isnt that toxic really. Its the organic mercury thats really scary

[@TotallyFred]

For the pin size, you could have multiple thin pins in parallel on each side.

[@y33t23]

3:09 Where can one buy a heating band like this one? I've been looking all over the internet but I couldn't find one

[@deltab9768]

I’m wondering, what is the reason for tungsten electrodes? Would it not be possible to use a metal of lower specific resistance such as copper?

[@FUZxxl]

I wonder if this would work better with a constant current supply instead of a constant voltage supply.

[@bdunderscore]

Yeah, using a lab power supply with CC mode would resolve the shutdown issue.

[@Alexander470815]

You should be able to use a Meanwell HRP-200-3.3 it will not hiccup go into hiccup mode but just push a constant current through.

[@realedna]

It's hard to predict the right output voltage without measuring the resistance first.

[@Alexander470815]

@@realedna it's not needed. The power supply will just push a constant current, be it at 0V, 1V or 2V etc

[@AdvancedTinkering]

Thanks for the advice! I will take a look into that power supply.

[@daytimetundra7757]

You should check out the potassium bullets someone shot. They are so awesome.

[@MrMartinSchou]

NaK makes for the funniest water gun ammunition.

[@jercos]

You might get the supply you have to work better unmodified by lowering its output voltage... many supplies have a trimpot next to the terminal block, and the resistance of the liquid metal alone might be enough if the voltage can be lowered enough.

[@jaxjackson4100]

Have you thought about gold plating the part of the tungsten electrodes in the glass? Gold should seal the two materials better.

[@AdvancedTinkering]

Interesting idea. I will try that once I have my PVD setup. But it's actually not the seal that's problematic but the different coefficients of thermal expansion. With a thicker tungsten electrode the difference could cause the glass to crack when cooling down or when the electrodes heat up.

[@jaxjackson4100]

@AdvancedTinkering I see. I didn't think about thermals. What if you tried a hollow or tube electrode, say one that could be cooled with a heat pipe going to a larger radiator. Would keep with the no moving parts theme. Though I am not sure of the sizes for either of parts. If it works. You could have a nice-looking ring radiator around the lower section of the NaK fountain. For esthetics. Beyond that, you could try or look into using a high temperature tempered glass tube as a sleeve or spacer. Sorry, I'm not a glass worker. I'm not even sure if something like that would be possible, let alone work. Either way, love the NaK fountain. Hope you are able to solve the minor issues you are having. Cheers, mate.

[@bottlekruiser]

instead of using thicker electrodes, could you maybe use more in parallel? vacuum tube glassblowers do that all the time

[@Convolutedtubules]

Some capacitors in parallel, might stop the short circuit protection from kicking in.

[@AdvancedTinkering]

Interesting, I will give it a try.

[@user255]

Wont work. The output current is exactly the same, if not more. Without capacitor there is only resistance from the "pump", add the capacitor and now you have "temporarily additional resistor" parallel. IE more current will flow. Once the capacitor reaches equilibrium state, it does not affect the current flow at all.

[@Convolutedtubules]

​@@user255It will be interesting to see nevertheless.

[@user255]

@@Convolutedtubules Well, I think the metal is not pumped as high, but the pumping pulse lasts bit longer. Depends on the capacitance and power supply cut off speed. It might just not work at all.

[@fano72]

Can you build a plasma tube with NaK? It would be interesting how its light looks like!

[@lajoswinkler]

Overwhelmed by sodium lines. You couldn't see anything but that unless you had a cobalt glass or didymium glass in front.

[@ikocheratcr]

If you want to try the "add resistor" option, use a piece of nichrome wire. You can get it from any heater that has it. you cannot solder it, so a connection like the one you used here will be needed. Then put it under water, just use a bucket with water. A resistance of 60mOhm will be too much, so it is not much what you need. If you just want to give it a try without the nichrome, use a piece of copper wire, and again, put it in a bucket of water to keep temperature under control.

[@T3sl4]

A purpose-made resistor will do fine; ribbon-wound ceramic or aluminum-case (heatsink required!) types would be typical examples for this value and rating (~50mΩ 100W).

[@realedna]

It will just waste the energy from the overvoltage and require a bigger supply than actually needed. The key is to provide the current at the right/low voltage level as the resistance through pure metal is surely very low. Also it would be an advantage to be able to control the current (and therefore the effect) directly, rather than the voltage, which doesn't matter. You only need the voltage to overcome resistance here. Adding more resistance is counter-productive. He doesn't want to build a heater...

[@MsRagidandy]

Perhaps you can use an isolating oil that evaporates when heated under vacuum. Then you can remove the oil after adding the metal.

[@adrianhenle]

A cannula transfer would be the best way to deliver the eutectic without picking up mineral oil, IMO

[@Intelligenthumour]

To deal with the heat of the electrodes why not use multiple and either run air over it or have heatsinks attached to the electrodes to dissipate heat?

[@senorjp21]

I think you could make an unregulated DC power supply with just a transformer and full bridge rectifier. You want high current and low voltage - the "circuit" has low resistance so you don't need much voltage. The power supply is switching off because the regulation of the output voltage is out of spec, but you don't actually care about voltage regulation. A transformer that is 100:1 will take your mains down to a few volts and the FBR will give you 100hz pulsed DC which I think will work for your application.

[@realedna]

He would need to measure the resistance R of the circuit at least to plan for a fitting supply, that can deliver the power of P=I²*R. So not any transformer will just work, it could overheat as well depending on the resistance at the output voltage.

[@TorPaulin]

Have you tried Galinstan alloy? It's a mixture of gallium, tin and indium metal.

[@alexpotts6520]

Mercury fountains are probably more dangerous because they require much higher voltage/current because of how dense mercury is, it requires a lot of energy to get it to move. You don't have the fire hazard but you do need to keep the fountain airtight anyway due to the toxic vapours, and because mercury is so dense it's much more liable to break the glass. Really the only advantage of mercury I can think of is that it looks prettier because it doesn't stick to the glass like NaK does.

[@fdgaming]

could you not suspend the juicy metal in acetone or 100% alcohol to wash off the oil before adding it into the glass ? or include it into the fountain then evap it of with the same heater and vacuum you used for water

[@Ishimanian]

Should add some flux for the connection between the copper ribbon and the copper caps so it’s actually using all of that carrying capacity instead of just whatever touches when crimped.

[@makegrowlabrepeat]

Could you introduce some sort of trap for the mineral oil?

[@xenomancer1]

How to fill: Add an argon purge to the NaK container to create an inert atmosphere above the mineral oil. Purge and fill the canula in the fountain septum or in the argon above the mineral oil several times to eliminate oxygen from the syringe barrel and leave it filled with argon. Use this argon in the syringe to displace any mineral oil that enters the canula after dipping the end into the NaK. Withdraw NaK into the syringe and avoid withdrawing any mineral oil. Remove the end of the canula to the argon above the mineral oil. Allow mineral oil clinging to the outside of the canula to drip to the end and tap the droplet to the side of the NaK container to remove it. Eject a small amount of NaK fluid to clear the canula of any internal mineral oil and then withdraw argon, preferably next to the end of the purge line, until a bubble enters the syringe barrel. Move the canula end to the septum, wiping the outside with a towel or cloth to clean it of mineral oil as you move it. Pass the end of the canula into the septum, using the towel or cloth to clean the canula from the sides just before the septum as it passes in. There should be minimal mineral oil in the system if this method is used.

[@sherannaidoo2712]

Try using a DC welder as your power supply mate.

[@TheFreak111]

What about using a shorter chain alkane like pentane instead of mineral oil and evaporate it later? Pour it in cold, heat it a bit and argon flush + vacuum to remove it?

[@makegrowlabrepeat]

Can you use tungsten welding electrodes? They can get extremely hot without melting

[@n-steam]

extreme temperatures pose a challenge to the glass they're encased in

[@urdnal]

Hey you're the only one I think I can get to look into this: beta phase titanium 3 gold intermetallic. 75% titanium, 25% gold. Beta phase is supposed to be 4 times harder than steel? I can't find much info on it, except the paper describing its discovery, let alone a photo of what it looks like.

[@number_three]

I'm sure many people have probably asked this before, but in case nobody did: did you see der8auer's recent video on the NaK CPU cooler? It was basically going to be a commercial product using this type of setup, which would have been concerning safety-wise to say the least.

[@skittlesryan7862]

Those electrodes heating up makes me think this is more akin to a timed incendiary device.

[@alext6933]

I want to drink it.

[@jamest.5001]

A microwave transformer rewound using a diode full bridge rectifier, the pulsed DC will allow the electrodes to cool and wrapping the transformer with 4 gauge welder cable, 2, 3-6 turns should supply the voltage required , 1.5-3v maybe a scr dimmer to control the input power to the transformer. To limit the output . Or possibly just use the transformer core, use a zvs driver, to pulse the windings on the core. And use enough 4gauge secondary winding to get the voltage. A 18 gauge winding connected to the zvs driver should output a fair amount, not sure on that one, but definitely a few mosfets driven by a 555timer circuit with a 2.5v push pull into the transformer, using a 12 gauge center tap run it at 30 to 80hz to limit power , possibly get different effects . The mosfets push/ pulling the primary center tapped winding, with a 4-6 gauge secondary, (with 2.5v in on the primary it will probably require 8 gauge windings with a center tap. The 4gauge secondary is it limit the voltage drop from resistance, being it is a low resistance load.) But the simple power supply would be to rewind the secondary of a microwave transformer. Keep adding a half wrap on the core until the voltage after the full bridge rectifier is reached, using a dimmer with a scr with zvs, to adjust the power , also the pulsed DC will allow electrodes to run cooler. To get larger electrodes in the glass. Use a fork shaped here it passes through the glass. Or possibly use a bolt with a hole diamond drilled through the glass, if it must be tungsten drill a hole 90% through it, and press a copper plug inside to lower the resistance so the electrical connects to copper, the next step would be silver or gold. Press a silver pin into the tungsten, or melt it and pour it in. Difficult in a tiny piece of hard metal. Heating the tungsten, and chill the silver is about the only way, getting a 2mm silver pin, into a 3mm tungsten pin, with the end being 0.8-1.0 mm thick, over 25-35 mm the resistance should be a fair amount, removing 80-90% of it should reduce heat, but making a stainless screw, to fit a hole in the glass, , make it large enough. That a hard steel nut can be lowered inside, and a long with diamond paste lap the glass flat and smooth, so a stainless bolt with a silver plated copper/brass bolt can be threaded inside, a 3-4mm bolt with a 1.5mm bolt inside it's alot of work, what if the glass is made conductive ? Would the resistance be low enough to be useful? But a diamond cut hole in glass, could a small bolt not be sealed with epoxy, and can't stainless be used over tungsten? Lots of questions, maybe this can be made with PYREX type glass, with the electrodes placed in the glass mid shaping. Then shrink the area. And form the top and allow it to cool extra slowly, just some crazy ideas! ✌️

[@charlesurrea1451]

I think if you're electrodes had an element, much like a wire winding turret, it would be easy enough to seal in the glass.

[@OmniversalInsect]

I always found the idea of a liquid metal fountain interesting, there is a mercury fountain in Spain but this honestly looks much better and maybe even more dangerous than mercury.

[@EthanReesor]

A cesium fountain would be amazing

[@ThylineTheGay]

seems you have a NaK for it

[@MrKingdavis13]

i would replace your electrodes with modified spark plugs from a high energy ignition system or a modern car that has resistors built into the plugs already. If you make the threaded inserts sealed in the glass you can experiment with different spark plug designs as well so it would be wise to pick a common thread and depth for that reason. It would also be interesting to see the pump work normally shooting to metal to a height that keeps it in an argon zone but

[@MrKingdavis13]

occasionally have the metal reach beyond the height of the safe zone where it can react with the air and ignite. sorry i accidentally hit enter rathe that the shift key. Lol

[@wakebreaker]

NaK!!!

[@elektronikvideos-bremen2873]

👍Ich würde den Magneten durch einen Elektromagneten ersetzen, der vom Hauptstrom erregt wird wie bei der Reihenschlussmaschine.

[@marioskot4883]

What if you mix NaK with Caesium?

[@buckstarchaser2376]

I don't understand why you didn't set the power supply voltage low enough that the current draw was within its limits. It appears to have a readout for amperage so that you can find what current you need to get a good flow going without overheating. Then, you could simply use that current as your "setting", as a magnetic field depends only on amperage. Voltage does not factor in, besides being able to push the amperage through the circuit. -I think that if you do another version of this fountain, I would suggest to put levels in it. Have electrodes in the line of the jet that perform different effects, and are connectable from the outside. -You could have a level (or 3) of accelerators, and then possibly a set of beam-steering electromagnets. You could make spirals by changing the deflection while the stream is in flight. Put a flat-bottomed Erlenmeyer flask upside-down at the top, and use the bottom (top, in this case) as a target for the metal jet. Then, make transparency overlays to put on the flask, and hook the whole thing up to a microcontroller. Program some games and practical applications into the microcontroller and ???, then victory. -If you pump RF energy into the metal while you're at it, you could make a pretty weird antenna with it, and light up florescent bulbs outside the setup. Or, I'm sure there would be a way to get an antenna length sweep along with a frequency sweep, for the most impractical radio scanner possible. -Put two complete systems in the same container, but without the fluids (normally) touching. Pump them both into another upside-down Erlenmeyer flask, but at angles away from eachother. Then, offset the separate (isolated!) power supplies by a high voltage to create a unique Jacob's Ladder that is so awesome, Styro Pyro would probably burn something in your honor. -Instead of the legs that are on one side of your current setup, make a second return tube, but this one goes to the top (of a shorter display tube). The goal is to get the metal to fly up and into the top port, and it loops around back to the reservoir to make a continuous ring flow of the metal. Perhaps it would pick up momentum. If it does, then the head-height would increase rapidly. If so, you could change the layout so that there are multiple nozzles below your main jet (in some kind of pattern), and put your electrodes above these nozzles. The side nozzles are normally below the surface of the metal, so when they are revealed by the fluid being looped into the top return tube, the momentum would make a multi-jet fountain with an infinite loop in the middle to give it momentum. -Make a horizontal version with the reservoir on the bottom. The jet and return are horizontal, facing eachother. From mid-way along the bottom tube, put a somewhat wide up-tube. Put both the oil and the metal in this one. From one end, you flow the metal up, where it jets horizontally across the spot with the "oil tube", and into a receiver cone. I'm trying to describe a glass aspirator pump (and you could just buy one and melt it into your system) but this one would necessarily be a closed system. The jet of metal makes a low pressure zone when it crosses the gap, and it pulls the oil into the flow. The oil then separates from the flow in the bottom tank, and rises up to the up-tube I previously described. I'm not sure, but that may be the most clear representation of how an aspirator pump works that has ever existed, so that's cool. I guess.

[@buckstarchaser2376]

-Put a magnetic "pinch" coil low in the stream. Now you can vary the column width in real-time. Now hook the coil to an amplifier output and put patterns, speech or music into the amplifier to watch a visual display of the sound. Arrange an array of different-colored, columnated LED bulbs so that they strike the jet at an angle that depends on how wide the jet is at that moment/location. Now, you have a multi-colored light display of the waves in the flying metal mirror. I think that the mirror would change behaviors in relation to the sound frequency as well as the linear rate of the jet. Some tinkering may be required.

[@RikestRik42]

Seems like you have a NAK for this kind of thing.

[@MaxWithTheSax]

Maybe using multiple thin electrodes would be a solution to the overheating problem

[@AdvancedTinkering]

Interesting idea! But I think fusing multiple electrodes close to each other to the glass would also be pretty hard. But it's worth investigating.

[@GermanMythbuster]

❤❤

[@snowdaysrule]

Can someone explan to me how two solids mix together? Like from an energy perspective, what drives this process forward? Like you take two solid hunks of metal and touch them together and then you get a liquid? Where is the solvent?

[@1Oldroot]

You should add water to the fountain too lol.

[@Skwisgar2322]

Instead of thicker electrodes, you could maybe do multiple smaller electrodes?

[@AdvancedTinkering]

I think it will be hard to fuse multiple electrodes close to each other into the glass.

[@jackspinetta4580]

You have tò use a current generatore not a voltage generatore. So you can have a constant metal flux