You can try to lighter the shaft and change the magnet sensor to a light sensor ( diode). Tomatoes are disgusting!

The magnet might make the spinning part unbalanced so either balance it or make a valve that is driven from the output shaft.

A tip, instead of sanding the parts when it's a rotary friction use an abrasive compound like a 600 or 1000 grit lapping paste and then work up to a polishing grit. Just put it on the surfaces making contact and spin it normally assembled for a few minutes. Clean and replace compound every few minutes and don't worry about parts staying rigidly against each others, lateral movement will make for a better removal and polish. Might take 15 20 minutes but you can potentially end up with thousandths of an inch tolerances or even more precise with technique. Then you can get away with no lubricant as long as the parts don't start moving out of tolerance from wear elsewhere.

With a hall sensor you can monitor the position and speed of each rotation, allowing you to fine tune for the optimal point to open the solenoids and for how long. You can even use a potentiometer to shorten the duration of each pulse so you can control the engine speed

Video idea: Can you make rocket with a thrust vector control similar to spacex falcon 1? With a couple of linear actuators and an arduino controller it should be doable

Since the rotor is symmetrical, put magnets on both sides, that way each valve is triggered twice a rotation. In theory this should produce twice the power and your motor will run smoother

You can use a timing belt to have the valves time properly and use a third hole to eliminate any lag in the system

at 10:35 , we can see that an other "combustion" on the other side of the rotating piece is possible just before the magnetic switch turns on. and this maybe by adding a magnet on the other side. so it will also balance correctly the piece and help the engine to run better

To mate moving parts with close tolerances, you can use grinding compound and lap the parts (grind against each other in place).

Read the position of the shaft for your valve timing instead of magnets on the piston, might also need a small flywheel to smooth it out.

I noticed the weight of the magnets and the timing greatly helped in the torque of the rotation. well done mate.

flywheel, output shaft, timing sensor on output shaft, looks like you could reduce blowby between the chambers aswell. rotor balance is important. im sure you're thinking of all these things anyway, its only early days so far. extremely cool i look forward to the next developments

I think you could do two things relatively easily to improve the engine’s power - first use an encoder on the output shaft to monitor the rotor position and use an MCU to control the air or ignition in each cylinder. That way you can also easily and repeatably adjust timing for maximum power. Second, use inlets for all three chambers to maximize power generation. I think the timing method is the most important change for maximizing power though.

Hi, I would like to suggest a theme for a future video. I enjoy watching your videos and find them informative. My suggestions would be to do a video comparing the different engines you have built in terms of durability, energy efficiency, power or torque output, etc. Thanks and I look forward to your future videos.

To reduce the wobble in the piston, make sure to have a way to use the bearings on both sides of the motor housing.

Something about this video feels like a step up in quality. Great work!!!

Something to keep in mind is that the co2 goes through a non adiabatic expansion in the Chamber which causes cooling in the system, causing a drop in efficiency after each activation. Switching back to high pressure air will give you better efficiency for longer durations

I think that in actuality at how fast rotary engines spin I think the imbalance of the magnets on that 1 side will eventually cause issues maybe only in high horsepower applications though because Mazda's rotors weren't even all equal from the factory

I think a small optical switch would work great, basically have a transparent disc with 1 or 2 black marks that trigger the valve, or the opposite depending on what kind of switch you use. magnets work too. the advantage of a separate disk is that you can turn it to adjust the timing.

I worked on in-flight optical sorters used in recycleing plants, they used high flow MAC solenoid valves that would DEFINITELY flow enough for this purpose, assuming your compressor and supply line are up to the task.

For making clear resin parts clear, and to lubricate them, I've found it very handy to sand then up to 800 grit sand paper, then buff the parts with Dow Corning 111. DC111 is sold as an oring lubricant, but being silicon based keeps it from degrading the 3D prints and greatly reduces friction. It also has a high surface tension, so it stays in place very well. If you don't want a lubricant on the surface, another great way to make clear parts clear is with Novas 123. It's a multi step plastic polish that also works great on aged headlight lenses.

You can actually use steel because it has less friction and easier to seal. and place another pair of magnets in the opposite end of the rotar to get twice the pulses in each round which means more power, and you can make some chambers in the casing that has some oiling mechanism to reduce friction as much as you can.

Since you use compressed air so much in your videos I think an interesting idea for a future video would be a 3D printed air compressor! You could power it with an electric motor or for some extra fun, a gas turbine!

Hey man, another great vid. All I would do to make it more efficient would be balancing the rotor and just decreasing the over all size of the engine itself

Hi integza, with the magnet on only one side of the rotor, you're only firing the "cylinder" at a 50% duty cycle - it could be firing for the other side of the rotor too. with a magnet on each side of the rotor and with gas hooked up to all 3 "cylinders" (lobes?) you would be able to get better torque and more power! Edit: like this comment so he sees!

I use a read switch about 15 years ago on a two-stroke lawnmower To run on compress air. Good Job bro.

Hey! I think you should use 2 sets of magnets for activating the valve. Twice the pulses per rotation. Also as other users have suggested, it wouldn't be a bad idea to have the valves triggered by the crankshaft rotating - less weight, faster spinning

I would grab a high pressure nitrogen tank(s)(used in paintball). Small one could fit in a large RC car/boat. Going to a light sensing unit would help with the unbalance... Also wonder if you could power it with a steam engine =))

You should definitely try to make a turboprop engine. Would be a good in-between for the combustion engines and the jet engines you already made...

First off, I love it that you are using the engine design. I would get the weight of the magnets you are using and counter it out on the opposite side (either more magnets or plastic). The main shaft needs to have reduced slop/give. It looks as if you already have a bearing in place so adding something to ride in there to reduce vibration would help.

You need to improve upon your cam shaft, and move the magnets outside of the piston housing. Then the magnets should be placed along the camshaft and you can adjust the timing from there. This will also help you balance the entire system which will increase efficiency. The goal needs to be balancing the entire rotating structure.

I'm sure this has already been suggested, but a second pair of magnets would balance the rotational assembly while giving you two power pulses per rotation. Smoother power delivery, and likely reduce the gear lash from asymmetrical pulses! Great video, interested to see where you go with it.

Great video! The pilot valve has a small port restricting flow. Try a bigger port 3/2 valve as the valve you are using is just for pilot signals and not flow. Also try adding a small air lubricator to the supply airflow, the droplets of oil will assist in sealing.

As you already have the Compressor, you could try mounting several valves in parallel to decrease the pressure loss, all connected to one motor inlet. That way you could try to get longer (and cheaper) runtines. Also you could try to paint a part of the rotor a light color and use a microcontroller to trigger on the passing-by of the spot, to decrease weight on the rotor.

It might be useful to use 3 Hall effect sensors and valves so that only one section of the motor fires at a time, I think this could increase efficiency and make the “fuel source” last longer and still have a performance increase over just having the original single input. Keep doing interesting things man 👍🏽

Idea: the valves can simplified and be self activating! Just a hole on the bottom plate at the right place and the spinning part covers it and uncovers it at the right time. Have to be tuned/placed at a specific spot for the desired rpm - but mechanically simple.

Hi thanks for your videos, but they are a bit advanced for me☺. Could you do a video for beginners, an entry level 3d printed engine?

I notice on most of your high heat builds you normally use rather small contacts between components. I believe increasing the length of the screw holes (and screw) as well adding a washer and tightening it enough to leave an impression when cold may cause the hot plastic to form around the head instead of pushing them apart.

That is some great engineering Integza

There are some issues with the current solution of your engine (sadly), BUT i can write up a few things you can change! For starters your still only using half its potential power, if you put magnets on the other side of the rotor as well, then you'll get twice the output, i would also suggest making gasket to get a better use of the compressed air. Another thing you can do is make an ECU (electronic control unit), this would allow for control of the timings of the valves just like in a real car, this allows you to control how fast you want the engine to spin as well as how long each valve is open and closed, this can be done by adding an raspberry pi to see what Eletro magnet switch is on at what point and check how long it takes for it to reach the next one. doing this could also allow for data output into graphs to allow you to see the power spikes along with its rotation speed. This is one of the things i would of done on many of your old projects as you have a tendency to make them all run as fast as they can without any control. Not only does this shorten the life span of the engine but it also doesn't help if you cant control the engine, as it would just be a glorified spinner with no way to control it. Most engines need to start with low speed with high torque and get higher and higher speeds as time goes on with less and less torque as its already up to speed. this would help combat the friction you keep getting stuck with and it will also help when you put a load onto the output ether to generate power or to spin something. hope this helps. I have also send you an email so hopefully you'll see this i used to study this at college as well as do it as my job but in the last two years i have moved into game creation and i love 3D printing my projects so that's how i came across your channel. (although i only have a half broken 3D printer that is very small, lol). I would love to have a chat with you about this sometime and give you a hand so let me know.

Video idea: I would love to see a drag race between all of your different engines on a car platform for example, and test how each engine performs at their absolute limit! Seeing the engine working is something, but seeing it powering a vehicle is a whole other level of entertainment

Just wanted to say that I'm a constant viewer of your channel and I love your work

It sounds so good with all three chambers

Looking at the slow-mo with the three valves it looks like the motion is jittery and not consistent, I suggest using a counterweight to try and smooth out the motion by preserving momentum. At the beginning of the video, it looks like on the original engine they have one so this could help.

Improvement idea: Add a “warming system” instead of a cooling System as normal. This should help it move more freely since there will be less thermal contraction in the bearing and the stop dry ice from forming in the engine on long runs. Additionally it should reduce efficiency drop off since having the chamber cold means that it takes more of the compressed air to achieve the same pressure.

Efficient engines are all about achieving high output, low friction, superb balance between the moving parts and must be highly durable. I think you can do an aluminum version and put a timing and ignition system in place because right now the power is from external source. Keep doing great! you got a new subscriber.

Also if you could somehow get a consistent measurement of the rotations you could make an electronic valve that opens, and closes rapidly for the same effect as the magnetic censor.

I would deff put another set of magnets on the opposite side of the rotor for balance and double fire rate. Maybe put a regulator on the input pressure side and then stack up on cylinders for extra run time.

I love this video and concept I feel you should have magnets on both sides of the rotor to double the pulse of pressure per rotation and even out the weight on the rotor. Another thought is some sort of center gear so the rotor doesn't wiggle around the block of course Friction is a problem here though. Also maybe some divots in the side of the rotor to catch more air to give it a more forceful rotation kind of how a HEMI piston works.

Make a base plate for the sensors and another for the magnets to get rid of imbalance. Connect the air to a compressor that suits the engine needs and use an accumulator to stimulate a turbo changer or supercharger and see how its behaviour is while running it on boost from the accumulator.

My new favorite channel, thank you

I’m sure this has been said multiple times but clearance and timing are probably the most important - Source, trust me bro. Also make sure you are sensing the centre thing on both sides, in the first run with one valve you were allowing 1 burst of air for every rotation but because it’s symmetrical you should be able to do 2 bursts per rotation per side. I also think the air compressor would be better for overall run time now that there is 3 valves and you are using the electronic valves

The valve is kind of redundant by angling the three inputs, no? Or to be more efficient you could time the pulses to fire based on desired rpm (and input pressure) and eliminate the (relatively) heavy magnets.

For the valve, look into how hydraulic valves work. They usually use a low pressure pilot to activate a higher pressure in the main supply.

I would suggest making the teeth of the housing for the rotating portion smaller and more frequent which will decrease friction, also I’ve seen many valid points about adding sensors and balancing the inner rotating portion. Another good idea would be using a valve that collects gas which emits from the engine and replenishes it into the engine.

Upgrade idea: Do an external Hall effect sensor (trigger wheel) similar to ignition on a car, which should allow adjustment of timing more easily. The timing should allow it to be more efficient.

Hi Joel, I think this might help you with the engine. Starting with the timing, the magnetic sensor idea rocks! but placing the magnets directly into the rotor might cause vibrations and that could be producing that extra bit of friction, so why dont you try to mount them on the motor shaft? Like if it was a flywheel, that way the magnets will spin allways on a perfect circle and you would get "some" momentum to keep the rotor spinning (like an actual flywheel), the only thing is that you would have to design an extended casing to hold the sensors. I think it might work properly. And for the friction...maybe if you use that CNC machine of yours and try to machine the rotor in aluminum, the weight might help it loose some friction, as if it has more mass, in theory will go faster and as being heavier, has more energy to overcome the friction, but with the powder graphite as well. Hope some of this helps you Joel. Greatings from Spain Neighbour 👋

I don't have any ideas for a future video, just wanted to say that I love your videos and your style of humour.

For the pneumatic valve instead of the Hall effect sensor try using an absolute encoder or a resolver(preferably with more than 3 lobes). With an angular position sensor not only can you time better the air release, you could characterize the motor, approximate a transfer function of input air and output angular speed, or torque, and develop the control algorithm.

Seems to me, that you can add magnets to the opposite side of the rotor. This will achive two things: 1) Balance the rotor 2) Add power to the system, as the rotor will get twice as much input power The downside is of course, that this will also add weight and friction. To combat this, you might be able to do two additional changes, depending on the activation strength of your hall sensor: 1) Use smaller neodymium magnets 2) Embed the neodymium magnets in the rotor

It would be interesting to see multiple rotary/similar engines running in parralel like the 13b/20b

I saw that a good amount of our view traffic comes from here and I was like hey! I watch that guy! Awesome job on the build !

I love these little engines so much! If it hasn’t been suggested already, maybe you could try and make the smallest functional engine possible?

Seriously, I think this is the most impressive project you've done yet. Not because the concept is mind-blowing (though it is VERY cool), but because it's both the most strictly scientific, and the most successful project yet. Congratulations, man. You did GREAT! Really, seriously awesome.

To improve the design adding magnets to the opposite side so it actuates at both cycles may increase your output and balance the motor. It will increase the fuel draw but improve performance. As for a video idea. Once you get this motor running efficiently use it to power a propeller for a foam airplane and see how far you can get it to fly.

You can use a LED / phototransistor pair (like in a old PC mouse) and a cardboard disk with holes attached to the axle. By changing the hole size, you can control the valve opening time.

Looking forward to the miniaturized working model of the Mazda 787B 😜 Cool video idea!!

You need less weight, less friction and automate the valves to work at high RPM. Good idea is use a thin teflon cover in the walls. Both the rotary part and the walls should have a thin cover. As janky as your setups always are i don't think you can do much of the tolerances but you need to achieve tight tolerances to not lose pressure too. Less weight, well plastic is good so that's why teflon walls. For the automation part, use a very fast micro-controller, a Raspberry Pico should do very well. For the sensor part, magnets are too heavy. you can have metal on metal conductivity. Metal on metal has low friction ad well you only need small pulses of high voltage that's why a fast micro-controller so it is able to detect them at a high polling-rate. So just detect when the rotary part does contact on a desired point to conduct the electricity and then trigger the valve. Try running the positive cable from the rotary thing connected to a tiny metal panel on it's wall. Then on the walls of the engine put tiny metal plates connected to negative (don't forget the resistor so you don't fry things), try different polling rates if it does not work. It's impossible that it doesn't work. Hope the teflon coating does not prevent electricity. If it does, use the graphite lubricant then. Another way is do it all metal and isolate the electrodes from the rest of the engine. Again, tolerances need to be tight so it doesn't fail.

Nice job as always. It looks like your loosing some ground on those open valves being slightly out of timing though. The slowmo helps a lot in seeing that. Might try a vortex bernoulli effect inlet to increase it's speed and angle

i don't know how well the Teflon will handle it, but i suggest liquid lubrication, a very thin motor oil or similar, injected similarly to how a carburetor atomizes gasoline and inects it together with the air. I also imagine you could take inspiration from the valveless designs of 2-stroke engines and the way they move positive and negative pressures around the piston housing to make the engine run better. Your videos are amazing, much love from Germany!

You should make a force gouge with linear rails, springs and a universal bracket to test your jets and rockets. Then make a wall chart to show which jets and rockets did best. We would love to see your progress over all.

the magnet might unbalance the spinning part. But i am SO excited!!! I definitely think you should make it a combo of a steam, jet, and diesel engine to power a mini car or something! Love this!

I'd be really interested to see how the engine would do printed in metal, and maybe cleaned up via CNC. Homemade engines would be a huge boon to lots of people, especially if it could power smaller transportation like a bike or motorcycle.

Have magnets on two opposition sides, it Will help to balance the Middleton pièces and gives more boursts. You definitly need a compressor as well, or a biiiig tank of compressed air

Place a magnet on both ends. This has two benefits - 1. It will keep the rotating part stable. 2. The valve will be triggered twice as often (think of it like 4-stroke Vs 2-stroke engine).

Use oil, and make a tighter fit on top and on bottom, it would prevent air from leaking in neighbour chambers reducing the efficiency. But it would spit out oil from the holes on the bottom, so it's problem to be resolved or just be ignored, it's not big of a deal, if oil coating is in sweet spot, both thin enought to stop spitting, and thick enough to make a seal. Also, don't forget about balance of the rotor.

Here's something I wasn't expecting

You could use steam to power the engine. Modify a pressure cooker with a safety release and pressure gauge to be able to safely see the bars. With this you should be able to supply a near constant feed of high or atleast constantly pressured water vapor for tests.

You would need a big tank for an output oriented application, so anything moving from a car to a plane is impractical. Maybe a boat but it would be huge. You could make a stationary engine and power a generator. Also to balance out the heavy magnets, you could stack more plates together slightly shifted for a smoother power output, bolt a propeller on the shaft and compair the generated thrust to your best jet engine design

I love Prusa 's 3 d printers ,nice pick 👍 Video propose idea: make a homemade fire extinguisher that instead of co2 cartridges as propeller this uses some of the engines you have made so far . Sorry for my english,love from Italy Love your videos

you should race all of the different kinds of engines you have made over the years in a super sick cinematic style video!

Id love to see this in an rc plane! Great work on the vid man.

Like you mentioned, you should try adding some exhaust ports to reduce back pressure. Install 2 magnets on the peanut so the single inlet triggers 2 times per cycle, and have open exhaust ports on the other 2 chambers?

Remove the magnets and place them on the main shaft for timing then use a Pic processor to control the solenoid. Would be easy with a Arduino controller and also be a good chance to learn the Arduino IDE and further advance your awesome maker skills!

I really love you're videos Inetzga! Recently i've been reading about Spinlaunch's, well spinning launchs that throw rockets into the air. Maybe you could use one of your engines to throw a mini-rocket or something into the air?

Full time liquid lubrication system with a mechanic pump driven by the output shaft or some kind of pneumatic pump driven by the air compressor itself(or vacuum perhaps?) This should allow for better seals and higher revolutions with lower chance for mechanical failure.

I think you should add an expansion chamber for the CO2 inlet and use a hall effect sensor with an Arduino or esp to control the valves with smaller magnet and opposing weight to balance it.

Ok…., here is what you want to do step by step: 1: get one bigger tank 2: get rid of your valves (might want one fore a „master switch“ 3: 3D print or mill a hollow control shaft with a casing. (In said shaft you put three holes in a presise angle so every time one of the holes comes over an „exit hole“ ,leading into one of you three inlets, the rotor turnes.) 4: you have a sick, much less expensive and mostly easy to DIY engine. Much love: bilder of an all aluminium and steel pnymatic 5 zylinder radial engine😁🇦🇹

The first thing that comes to mind is to stack them--each rotationally offset from the last--and use a crankshaft to connect them. The more you have, the smoother the output should be.

Super cool! Hope to see you build a sterling engine some day!

add the magnets to both sides, it will help balance it and double the pulses per rotation. Or, use contact circuits to open close the circuit so there is nothing to balance, but I would still add them to both side. To do this you would want to run a contact through the resin to a metal shaft so you could ground the shaft, and then have the positive points be on the casing, so when it rotates to close the circuit on either side, the air valve is actuated.

Metal casting 3d prints is actually pretty easy. I'd love to see a old 3d rocket become metal casted.

to improve the engine, perhaps it would be good to add a second pair of magnets to the other side of the spinny part, also perhaps finding some way to keep a constant intake preasure would let it last longer as it seems that a very high pressure burst isnt doing much more work compared to the lower preasures. to increase efficency it would be good i think to setup the 3 intakes to function individualy of eachother with 3 seterate electrovalves and 3 seperate sensors to control each one.

I notice that before the cartridges are completely empty the mechanism will start to slow down, this might be because the condensation from the rapidly cooling components mixes with the graphite to form more of a mud. Try a liquid lubricant like motor oil and see if that improves.

Add a counter weight on the opposite side of there the magnet is to balance the system. Try to decrease the weight of the internal parts. Try to find some low friction seals so the pressure can build up in the system

My Ideas for you future videos are: • Make your 3D printed compressed air powered wankel rotary engine water powered (it does only need some minor changes) and try revisiting your 3D printed compressed air powered wankel rotary engine car using the water powered wankel engine and also make a 3Dprinted boat entirely powered by the 3D printed wankel engine powered by water ( which doesn't even need a tank and the water can be taken from the pond and it can produce extra thrust if you use a nozzle in the outlet), use a Tesla turbine as a pump to pump water to the engine. In the car have a tank and redirect the outlet water to the tank itself. Hence, There is no need for changing the water. Use a Lily impeller and make a boat so that it would be a amazing thing which has 3 amazing inventions by 3 amazing scientists(Wankel Engine - Felix Wankel, Tesla turbine - NIKOLA TESLA, Lilly Impeller - Jay Harman), please try make this one in your style and I am keen to see your video on this. • Make a aeroplane powered using your aerospike rocket engine and launch it like a rocket and land like a aeroplane. Use your 3D printed solid rocket motor as a booster to launch it(and you can make it like a space shuttle). • Make a 3D printed rocket engine powered by hydrogen and O2 with your Metal Filament(From The Virtual Foundary), and iam keen to see what fuel ratio you will use. And use this engine to make a 3D printed rocket. • Make a Ramjet engine powered missile with your Metal Filament(From The Virtual Foundary) and please destroy some tomatoes🍅 with it (tomatoes are Disgusting 🍅). • Make a 3D printed rocket which utilizes your 3D printed Tesla pulse jet engine which would sound great exept for the thing that your neighbours will be angry at you. •Make a compilation video of you destroying tomatoes 🍅 all of these years( Tomatoes are Disgusting 🍅 ),(please make this one). • Make a 3D printed jet aircraft which houses your vortex cooled rocket engine but with little modifications like using your Metal 3D printing Filament(from The Virtual Foundary)instead of ceramic resin changing oxidizer from O2 to compressed air or normal air like conventional jet engines which is a Feast for our eyes and that would be an interesting topic. I think you need to name it the integzajet engine or vortexjet engine. • Make a 3d printed revolving piston steam engine And make a car which utilizes this. • Make your Tesla earthquake machine with the Metal 3D printing Filament(from virtual foundary) and make it from metal and try to use it to power a Tesla Coil. • Make a Thermo Magnetic Motor which was invented by Nikola Tesla. • Make a pyromagneto Electric generator which was also invented by Nikola Tesla. • Please make a Tesla Alternating current Induction motor with 3d printed parts.( Metal Filament from The Virtual Foundary) I would try many of these if i get a 3D printer but i could not afford it . It would very helpfull if i get a 3D printer. And you have got great 3D modelling skills and you are a professional in it. Please make a tutorial video in 3D modelling mechanical parts in solidworks with parametric equations(I want to learn it please make it). Iam asking this as a request please do it. Wish you a good luck in your rocket engine making courier and wish you for becoming a rocket engine specialist soon, you and Sir Nikola Tesla are my inspiration don't stop doing sir Tesla related inventions.

to improve on this design, I would recommend plugging the cover that has the ball bearing in it to contain the air more, and put in some exhaust ports(about 1/2 to 3/4 the size of the intake) to make the engine use the air pressure more efficiently. Hopefully that would make it so that you would use less "fuel" and allow the engine to run longer.

Love the mad scientist look!

If you want to keep using CO2 cartridges I would highly recommend having a way to keep them somewhat warm so they do not lose power as they get colder and colder

After watching your videos for the last few years I think that it's finally time that you 3D print a moustache. I just can get used to the clean shaven look. Seriously though that build is super cool. An optical sensor and potentially an Arduino to control the timing will probably give you better results. I don't know if it would work with air but a fuel injector solenoid could possibly work for this. Make sure that your solenoid and hoses are rated for the pressure that the CO2 cartridges put out. Most air solenoids are only good up to about 120psi and over that you are risking an explosive ending.