Building and Testing a New Magnet Motor on a Dyno!

[@kANGaming]

What other things should we build or test in Scrap Mechanic?

[@Andwo0]

Build Obama

[@mpompa1369]

I am curious if it is possible to create a levitation engine , like those in UFOs using the magnets you used in this video .I once tried to make one but I couldn't make anything exciting happen but I hope you will.

[@CraigParsons-n6j]

@kANGaming I saw you first video of this magnet mod i'm getting nostalgia from this.

[@justinpatterson5291]

Can you build a contra-rotating gyrocopter. Like a mono engine gyro had a kid with a v22 osprey?

[@Iamdumb123-t8x]

@@Andwo0# toilet gang

[@Ghostly-m4m]

Kan maybe replace the encoder wheel with lighter material or friction less mesh blocks to see the results. Just curious

[@NSG-kc6zl]

Using mesh blocks won’t make a difference since a creation doesn’t have any friction with itself

[@dm-5]

Just use something that doesn't have a full block hitbox and therefore causes all sorts of collisions. Same for the tremel gear

[@Ben_Bingo]

@Mistner nah hes right, no frictional force gets calculated on the same creation, only collisions

[@coffeedude1468]

A heavier block would work better cuz it would act like a flywheel

[@Ben_Bingo]

Hey just a quick heads up, it's almost always better to use an L shape instead of an I shape on the encoder wheel for anything like magnet motors or piston engines, wedges in the encoder wheel can also hurt performance at speed too. Also what you said about there being a delay between the sensors and the magnets being a thing is very true, its actually the main reason you need that offset before the encoder, true for piston engines and magnet motors as well.

[@TheBaldrickk]

Movement > sensor > magnet state > impulse Looks like there's 4 ticks of latency there.

[@_miobrot_603]

hi ben!

[@PasiFourmyle]

I really hope Kan can find a way to make these scrap mechanic mechanism videos a regular series kind of thing.

[@arcturuslight_]

Brushless dc motors in RC stuff most of the time are actually synchronous ac motors with a controller converting dc to ac. Also they are often wired inside out - the stator with coils is put in the center, and the rotor with permanent magnets is the outside cylinder, and the shaft is connected to the outside part.

[@kANGaming]

I have a few of those types as well where the rotor is on the outside but my 1/10 cars all seem to use internal brushless with the rotor on the inside, stator on the outside with the coils in it.

[@johnhancock8552]

Do you know of or have ever worked with magnetic shaft couplers? Very cool to see them work.

[@Pystro]

With the "physics going on" in your gearbox, I wonder if you could build a magnetic gearbox.

[@bigbug314]

If you have about 600 rpm, a round takes 4 ticks so with the one tick delay you should have about 90° of difference so it's about what you got (of course it's a bit random but it's logical to have a big offset). Also when applying force with a mod it takes 2 ticks before de velocity changes for some reason, so maybe the sensors where activating the magnets for the next turn

[@egornick9206]

I have the same issue with my vanilla Gatling gun

[@flazerrazer2992]

@@egornick9206 why would you need a gatling gun?

[@egornick9206]

@@flazerrazer2992 big red robot

[@braydonfisher9273]

I would recommend having south polarized magnets with the north ones (static on the body of the motor) that are activated with the opposing north magnets

[@amahashadow]

Yes, my thought as well, it would stabilize the shaft and avoid the tangential force moment that's created with the push-pull configuration

[@braydonfisher9273]

@@amahashadow it also opens up the possibility of a bidirectional design

[@tyreni]

Your mix of real-world engineering and Scrap Mechanic possibilities are extremely satisifying. Keep it up Kan, four years later and still cranking out the hits.

[@MrMattkid]

On your last test with the tremal gear, you had the extra bearing coupler in there. If you remove that you might see better results, but idk how much better.

[@TheMostUt]

One thing you might try is putting two separate plus shaped encoder discs. This would allow you to connect both to a controller and change the total overlap distance. You can paint one block on each, and offset them by 90degrees to get two blocks worth of timing. You can do this with a single plus, but what you cannot do is get, say 100 degrees of timing. That you can accomplish with multiple encoder wheels. I would love to watch an hour long video of you adjusting the timing duration and position.

[@madmort2081]

Love these new videos!! Maybe you could go back to the submarine you built and use the air generator made by fant to make it more immersive?

[@mystifoxtech]

The wedges on the encoder wheel are actually colliding with the sensors proving that Brick Gang is superior.

[@mystifoxtech]

you should also probably have the magnets on the outside on both north and south

[@mystifoxtech]

the encoder wheel should be separated from the sensors by one block

[@complexcanvas3825]

Oo more dyno. Gonna be a great day

[@swampcooler8332]

Im not convinced that it was the bearings decreasing performance, i thint it might have been the bar. Try using pipe peices instead of blocks to see if the curved pipe peices that were holding the pistons were colliding with the bar

[@elektro860]

you could utilize the xo-meter from the modpack to get the angle of the shaft and from that find out what magnet / piston should enable

[@shielderer945]

try replacing the concrete that are phasing through the pistons with pipes, less collision maybe will make it faster

[@deltacx1059]

1:48 except brushless "DC" is actually AC hence the need for the 3 phase speed controller. And the motor you have in game if say is closer to a synchronous AC motor of some sort.

[@aeresys]

The timing problem with the encoder should be a 1 logic tick delay, and just rotating the encoder isn't really gonna fix it. To make it perfect you'd need to measure the current rotation speed and increase the angle of the encoder accordingly since the delay is constant but the speed changes

[@lesto12321]

BLDC motor are 99% of the time 3 phase motor (so you may argue are really AC and the ESC does the magic). Anyway, if you look up BLDC 3 phase you will find tables on how generate the phases; and you can also "wind" it (in you case "signal it") to be a Star or Delta (torque vs speed). You changing magnet (coil, or strator) should be multiple of the number of phases, and the ratio of strator to coil is generally 4:3 (or 3:4, i always mix it up). Probably the best implementation is to have sensor stepping a state machine, and each step activate one of the state that control all 3 coil; IIRC is only 6 step, and each step change only one phase, so it should be easy logic (not sure how to implement it in scrap mechanic tho)

[@Error_Era]

Finally a channel that actually makes me confused not many things can do that you earned this sub

[@adamhbiggs]

Awesome stuff. the sensors for the encoder wheel might need to have the corners at a 45 degree angle (so that the center of the sensor is aligned with the circle the encoder wheel.) . also might need more balancing on the shaft.

[@kontakt4321]

Take a brushed RC "mod" motor with adjustable timing, clamp the can in a sturdy mount, give it 2s power with the screws loose, and advance the timing. RPM will pick up, and you can find a max RPM point doing the same as you're doing in this video. If you then lock the end bell, and power down the motor, the timing you have set will be so extreme that the motor won't want to start without help.

[@lilwashu]

your stabilizer post shouldve been pipes as well, im pretty sure your pipes you have on the stabilizers are collisioning with the squared edges of the concrete

[@nebsfisher]

Aside from seeing a vehcle powered by them, I would be curious about useing just two pairs of segments opposed from each other powering the magnets to see if the forces are balanced better?

[@levibruner7553]

I like how kAN after the video on the piston trials was like "Imma grind piston engines in creative mode until I find the one to beat this challenge" and now he's just messing around with a dyno. Also, some tips from a person who passed the first torque trial: 1. Make a four stroke engine with one piston above and one below each segment (this makes it more stable and powerful and reduces the number of bearings absorbing torque) 2. Lock the bearings between segments using blocks to prevent the bearings being over torqued 3. For the encoder wheel, put 4 sensors on the bar connected to the giant wheel facing a C of blocks with wedges for the ends of the C (wedges act as half a block, meaning your timing is 50-50 for an almost non-hitching engine) 4. You might be able to get a 3 stroke engine with each segment containing 3 pistons, but that makes things a little more complicated. Once you pass the normal torque trial, the wheel is upped in weight. Instead of cardboard, the outer rim is made of tier 3 wood. In the insane trial, you have the heaviest wheel in all existence: tier 3 concrete. Proper gearing and maybe a more clever engine design may be needed (like an auto rachet engine).

[@DenserThanLead]

Magnet motors are good for rpm and torque, im not speaking from experience but ive seen some doing crazy amounts.

[@JoachimVampire]

just to say, a house-sized electric generator can basically rip itself out of the floor WITH the correct instalation if it's electrically attached to an already powered grid but with the wrong timing (you have to align them )

[@OnyxBMW]

On your trammel gear, on that first step, you can directly connect the rotor to the drive gear of the trammel itself, there's no need for the piston there. To me it almost looked out of balance because it was only 1 piston, and as with the other test, bearings harm performance in general. I'm curious if you can also delete 1 side of those pass-through bearings, and instead make a U that passes through the support post that only connects to what's being driven on the other side, removing a bearing. If bearings are weird, it might help in high-RPM settings to remove as many of those as you can.

[@erykd.1360]

The reason why the speed keep jumping up and down is because those curved pipe pieces attached to the pistons are hitting that concrete part in the middle, instead of concrete you should have used pipe pieces instead

[@Elijah1573]

This and also maybe spacing the encoder wheel by 1 block between the magnets would help too

[@Orwrath2010]

You could just try a much simpler solution where you have the north on either side of the shaft then have south right next to them so one top one bottom if that makes sense

[@tommew450]

that has to be teh smoothest anyone has ever explained brushless vs brushed

[@termisher5676]

I wander if they will add that you can hide connection points with right click and shift and show them up again when pointing directly at them physicly seeing them as if you wanted to place block on it to enable connections You know purely visual to not make mess on the machine

[@AsuraXranker]

I got 7406 still going up torque using driver impactor

[@PasiFourmyle]

"Magnets attract each other....... that's a problem" 🤣

[@justinreher1500]

Kan adjust the magnet encoder while connected to the dyno to get the max torque while its loaded down it changes in real time and you can adjust min rpm on the dyno say 850 rpm

[@theredstormer8078]

I wanna know what happens if you make a piston engine except instead of pistons it has other energy sources like thrusters or arms connected to controllers.

[@jaredphipps3757]

Magnet pistons? It would be pretty large since you would need low frictions blocks for cylinders but it would be neat to see if you can time it.

[@Zirnike]

Could you be triggering multiple sensors at the same time if you go to fast? That could glitch things. Maybe make the encoder wheel larger but with the same number of sensors, just spread out?

[@ryotaokura5117]

Maybe add some spacers between the encoder and the sensors so that it wouldn't bang so much giving inconsistent rpm?

[@llama_lordie8776]

I wonder if the -15 degrees that looked like it should be best, might work for better torque at slower speeds

[@benmcbride3721]

Adding S poles on the outside between the N poles should eliminate the cogging/ skipping effect also add in magnets on the spindle so you have plus sign and a total of 8 magnets timed to repel/attract sould drastically increase your rpm and hp. The center shaft/spindle might have problems with bearing strength while rotating.

[@vict632]

You should try to have the outside magents be driven by something else! Like just add a looping signal that triggers them, so it drags around the shaft. Might be worse but idk worth a try 🤔

[@Defhrone]

if only I had this game is school. I kinda did for a school project, attempted to make a frictionless electromotor. An interresting you could try in this game is instead of changing the magnetic force from the static part of your engine changing it on the axle itself. I've just run simulations in my head when I did that project 20 years ago. And in my 'simulation' the axle method would either spin super fast or/and tear itself apart. Something fun to test in a game, not so much in a time sensitive school project

[@FFmotorcompany]

If you use just 8 magnets total and have a variable timing you can tune it up to 20000RPM+. I made my motor as light and friction less as possible and it even has some torque at high speeds unders 7000RPM it's worthless bud at high speeds it's amazing, the same dyno you use measures up to 1800HP for sore reason...

[@kennythegamer1]

It looks like the pipes connected to the pistons on the stabilizer parts of your trammel gear might be hitting the concrete that's nearby; that might also be what was happening with the motor that had the stabilizer. If you replace the concrete with pipes, they should be small enough not to get hit by those other pipes.

[@bjrnfrederiksson2505]

Great episode KAN. I loved it. I don't know what you should build now but whatever it you end up with is gonna be hilarious 🤣

[@mihalydozsa2254]

If bearings cause problems I wonder how much would change if you weld the middle only to the dyno and the other side with the sensors are just welded to the outside of the dyno. I believe the piston caused the problems because the tubes were colliding to the pole you used to hold the shaft. If you would use a pipe with smaller collision to hold the pole maybe would be the same as without.

[@Punchin83]

Replace the structure blocks with pipes to reduce collisions.

[@broscolotos]

You may have been getting the higher averages because you weren't resetting the motor before resetting average. It kept the old speed and added the speed of the adjustment from the controller.

[@XWorstYT]

wondering when if you were gonna do the torque piston challenge? anyway love ur vids.

[@mukuldeep3488]

you know you could place the lift on your creation if you place a bearing facing up then the lift gets suspended mid air. Sorry is this is like an old thing but just pointing it out

[@nischiita]

Hey @kANGaming, I just got your "old" video about the explosive engine recommended. Would be cool to see that one on the dyno too! Cheers ✌🏻

[@xpen2007]

That is exactly like a sensored brushless motor. Merely the esc is a bit basic.

[@jdwg5]

TLDR, your motor is an induction motor with a primitive VFD built into it. Your magnet motor is more akin to an AC induction motor. They have windings on the stator providing a rotating magnetic field (same as your device) and a typical “squirrel cage” rotor induction motors have non powered rotors (though typically an induction motor doesn’t have permanent magnets in the rotor, the magnetic field of the rotor is induced by the EM field that the stator generates). Only difference is induction motors don’t have your “encoder” ring built into them, they just run at whatever frequency is provided to them. So I guess your addition of the encoder ring is like adding a VFD to the motor that controls the excitation of the stator windings based on the speed/position of the rotor

[@Poldovico]

If your theory is correct and there is a constant lag between the sensors and the magnets, it makes sense that the faster the engine goes, the more extreme the lag's effect on the optimal angle of the encoder wheel.

[@peptobepto]

get 4 of these things and put them on a car. That would be awesome

[@Jkauppa]

you talk about brushed/brushless ac pulsed motors, try railgun type dc brushless motor, ie two rings with rolling ball bearings, the axle on top of the balls, vertical magnetic rails under

[@Jkauppa]

high current over the bearing, which is 90 degrees to the magnetic field, always, but short distance, so high current, low voltage

[@Jkauppa]

does scrap mechanics understand lorentz force, the cross 90 degree force magnetic field to current line/wire/plasma/liquid

[@Jkauppa]

the true dc brushless electric motor does not have gaps, no pulses

[@SJ16toons]

try putting low friction blocks where the tremmel gear slides

[@theess1275]

you should try ditching the encoder and use a logic clock to alternate the magnents like an ac motor

[@Omlet221]

I believe there’s a trade off with the timing between torque and speed. When you have the motor already spinning fast, you need the timing to be advanced to account for the lag of the rotor reacting to the charge of the stator. The amount of advance for the fastest speed can often be so much that the motor will go backwards when starting from it. When you want more torque you need the timing to be more retarded so the rotor can keep getting pulled along when it’s bogged down.

[@lil_minecraftien]

I agree that magnet motors are good for speed but really bad for torque, I've made 1 that goes to about 15 000 rpm I think and when turning on the dyno it stopped in a few seconds (I don't remember the torque reading)

[@szki272]

I am assuming the magnet motor could be scaled up as in longer and / or larger diameter?

[@tannerparson7030]

Maybe take a look into ICE timings might provide some help

[@wildwyatxbox]

With all this new knowledge of piston engines, I'd like to see testing with them in a vehicle. Seeing if they can go up a steep face, speed, etc

[@toad0net]

i love this stuff

[@nik325007]

Try replacing the support structure to pipes, this wil eliminate the collisions.

[@nerfgodbigguy1405]

kAN, I have an idea but I don't know If it would be executed good because it will be difficult , but you could try and make a hybrid system in scrap mechanic ,would be cool, and great work on those piston engines and this e motor is sick👍

[@mstermincollosal840]

That would actually be cool.

[@noahherre1144]

The Engineer of the year: „magnets attract each other that’s a problem“

[@milkam2004]

so I was messing around a bit more and figured out a passive and more effective mid run timing adjuster system that makes the speed of the electric motor increase the longer it runs and I managed to get 86k rpm, 134.8k hp and 24.8k nm torque. however I have noticed that it does this thing where it's rpm skyrockets, settle down to a pretty consistent speed then skyrocket in speed again so it might have infinite speed somehow. I just set the default controller angle to 20 and the one that needs to be activated to 360 then put an "or" logic gate and connect 2 of the timing sensors to it and then connected the logic gate to the controller. idk why this works but it does and now I might have an infinite power electromagnet engine (btw I used the same magnet layout as the other comments that I wrote) Kan if you see this, you should try the layout I made using only 40 magnets. I have to do some more tests

[@adrianadrianadrian4388]

the collision is bad, look in to the friction and i think it will work way better if you use wedges inside

[@steeljawX]

I can't speak on the behalf of the mod parts, but I think you're running into the limitations of the game. I don't think logic can effectively fire off faster than 1 tick in the game. So the redundancy of having 3 magnets on to push/pull the center magnet is what probably saved it and what's probably causing the jerkiness to it. It's jerking it because it's spinning fast enough that there are magnets reacting late because it's taken them 1 tick to turn on, 1 tick to turn off, then 1 more tick to react once more to turn on. But because it's spinning so fast, each outer magnet isn't necessarily firing in the planned unison you designed it for. That's what's saving you. You may have 1 or 2 magnets that are actually on that are push/pulling the center magnet and 1 that's possibly fighting against it, then as it's rotating 2 more activate on time to push/pull it and another one activates to shove the magnet laterally instead of radially. So the problem would be the speed limit of logic in general. If there's a mod color sensor that can react faster than the in game sensors, that might clean it up and make it run even faster. But my theory is that it's taking the logic too long to actually catch up to the triggering of it.

[@Budget_Monkey]

id like to see a rebuild of this where you use more magnets, see how much hp you can get

[@king49games72]

I found some of your old logic tutorials video and was wondering if you could make some more?

[@blueflame_sm]

Advancing the timing wheel speeds up an engine because scrap mechanic has tick delay. Let's say we make any engine and the engine is spinning at 300 rpm. The tick delay is 1/40th of a second, meaning that the engine makes 300/60/40 = 1/8 rotation before the piston or magnet activates (assuming sensors go straight to the piston/magnet). This is the main reason the timing wheel has to be offset, having a proper starting position is only a small part of it. If the offset is just slightly wrong the engine will work against itself, producing less power or potentially spinning the opposite direction. Now note that this calculation is done for 300 rpm, the required offset is actually a linear function of the current rpm. You need different offsets at different speeds. In vanilla SM this is basically unattainable, with the modpack it is possible tho. Let's do another calculation. The highest speed I've had one of my engines go after learning from the guys in the mechanical community is 1250 rpm. At this speed the timing wheel has to be advanced by 1250*360/60/40=187.5 degrees. The timing has to be advanced by more than half a rotation!

[@TexusNoe365247]

You should make a 3 phase induction motor

[@MarkkuS]

I feel like the offset is different with load or no load. Because of the tick delay.

[@WheezyCyclist]

Please put a heavy flywheel on this It will smooth out the inconsistent RPM The way this dyno works it may skew the torque output though But would be nice to see the effect of a really heavy flywheel

[@kadmandu2768]

is there any way to get it to spin up slower at first and work its way up to 1400 rpm rather than going 0-1000 real quick

[@madmothertrucker18]

Here’s something interesting… could you build something like the Tomorrowland people mover from DisneyWorld in Florida in ScrapMechanic? 🤔

[@madmothertrucker18]

Believe it uses magnets to move itself along.

[@xtremexgamer5958]

yo wsg im in the discord, great vid, i do loads of scrap mechanic and have something like 7k hours, and yet im still learning from you! thanks for being everything you are. also an idea u could do next could be a stabiliser such as the one in your hovercraft but better or improved, OR you could make a tutorial on how to make all the basic logic circuts, which im sure would be incredibly useful for people just starting to play, from one useful buddy to another.

[@xtremexgamer5958]

now that i think about it why not try to make a frictionless engine using floating magnets

[@dylanwatts9344]

You assumed the magnets work the same way as real life, which was immediately disproves just minutes in. So you have an average pull between all of the magnets, that's why the offset sped it up. It aligned with the average push/pull

[@Charm_Dragon]

I dont think it was smoother because you removed the extra bearings. I think it was smoother after you removed the shaft support because the shaft was sometimes hitting the shaft support. You used full blocks to support the shaft causing the pipe pieces to be flush with the support. The small wiggle room that bearings cause allowed the pipe pieces to hit the full block support. If the support was instead made out of more pipes, I do not think you would have the collision issue.

[@Charm_Dragon]

(obv. less bearings tends to equal better performance but I don't think the main issue was the bearings.)

[@CaptnApathy]

how about an axial flux motor? see how it compares with this one.

[@Brick8651]

This would be a great alternative for A radial engine

[@thomb.9013]

sensor delay is 1 tick, aka 1/40th of a second.

[@thesunnynationg]

about 60 hp with all the ups and downs

[@DS-tj1xf]

Loving these vids! Keep it up

[@erumaaro6060]

makes sense, the higher the speed the higher the offset due to sensor delay .

[@green7449]

I wanted more gear episodes and I got more gear episodes yet I have no idea what is happening this episode. 😂😂

[@roccov3614]

Yeah, we need to see it in a vehicle. Having numbers is great but, to some people, the numbers don't mean much.

[@Boromoro]

good ol Scrap Mechanic'

[@dominikvarholik7519]

i made like maybe 2 years ago a magnet engine with crude design or whatever but it made like 11k hp had like maybe prolly 2k torque or idk but it spun to around 400k dgs aka 67k rpm

[@dominikvarholik7519]

and what ive figured out is that compared to thruster engines and piston engines timing is just a straight line aka the higher the rpm the more reduced the timing have to be to account for the tick delay but in magnet engines it aint a straight line making my auto timing whych does what i talked about on magnet engines it just doesnt work at all realy what did work is finiking with the timing to get it up to speed and then it would be fine without the auto timing its realy weird idk

[@MrMeow-iq7kq]

Today youtube recommended to me one of the 3 year old videos that had a piston powered vehicle in a regular gameplay mode. Given the recent fiddling with piston engines, I initially mistook it as a recent video because of that. Sheesh, its like things never changed. Then I see magnet engine.... RIP piston engine?

[@milkam2004]

guys, I made a replica of kan's magnet engine, completed the inner and outer ring of magnets, adjusted the controller to 53 degrees and made a 1 block space between the sensors and timing wheel and got 47k rpm, 35k to 39k hp and 13k to 15k nm torque on the same dyno that he used with basically no lag, the potential is limitless, I havent's even added new rows of magnets yet. And no it's not a fluke test because I got similar results after every test, altough some of the tests were outliers and showed only 4k rpm

[@milkam2004]

ok, so it turns out that 4k was the correct rpm without configuring the timing for the new inside magnets. however, I found out that because of the scrap mechanic delay the faster you go the more of a problem it becomes, so I tested adjusting the timing mid spin with a switch can make it go even faster. It actually worked, however the scrap mechanic inconsistency hit and suddenly decreased the performance of the engine by 3 times for no apparent reason as I didn't make any changes to it, however, I could fix this issue by adjusting the controller angle and giving the timing adjutment more stages. anyways using this timing adjustment system I got 18k rpm, over 6k hp and abou 5.5k nm

[@sumoshark7986]

advancing the timing, or delaying changes top speed

[@curseofarosandmore9444]

This really reminds me of a V8 Distributor lol with instead magnets it would be spark

[@bravo320zf]

I would love to see you make a car with an engine like this.

[@rodneypatzer7912]

Agreed 🙂🚗🏎️🛻🚕🚓🚚🚙🚑🚛🚌🚒🚜🚎🚐

[@stuartpratt3662]

@kANGaming what about testing what difference would having the radius of the timing wheel (I'm thinking a higher radius would have more precision)

[@gepardmic6003]

Its about the game tick delay, the game runs on, that count. 60 rpm delay 1 sec. 120 = 2 sec... So in theory high rpm means bad low torque.