Can you 3D print a Transformer? (Experiment) || How to make a mains Transformer!
ฝัง
- เผยแพร่เมื่อ 7 ก.พ. 2025
- $5 discount for the first 30 to order on JLCPCB with code “JLCScott"
jlcpcb.com/?re...
Previous video: • Make your own "Powerwa...
Transformer Basics video: • Is it easy to create y...
3D printing a motor video: • Is 3D Printing a motor...
Facebook: / greatscottlab
Twitter: / greatscottlab
Support me for more videos: www.patreon.co...
Useful websites which were shown in the video:
www.sauter-sho...
www.roehrenkram...
www.emt-net.de...
www.robkalmeij...
www.jogis-roehr...
In this video I will be showing you how to calculate and build a basic mains transformer with electrical steel sheets. Along the way you will learn about the number of turns and the thickness of the primary and secondary winding and also how the size of the transformer influences those values. Once the testing of the "normal" transformer is done, I will then continue by 3D printing one with ferromagnetic filament in order to find out whether it can be an alternative material. Let's get started!
Thanks to JLCPCB for sponsoring this video
Visit jlcpcb.com to get professional PCBs for low prices
Music:
2011 Lookalike by Bartlebeats
Killing Time, Kevin MacLeod
(incompetech.com)
Building hint: every 100-or-so winds, put a little tab tape over the wire with the wind-count written on it. If you lose count, you only have to go back to the latest tab of tape.
omg why didn't i see this earlier i would have saved a solid hour of my life 😂
Best idea bro
Great tip! - could even write on each tab tape to make sure.
I'm so disappointed, expected after 8:50 a "let's hook it up anyway!". Send it to Mehdi.
Or andi aka photonicinduction, even if it woud work
i hate that he is so serious he should have done it outside or something or at least at the end to show what would happen
@@BadMax02_VR tbf I wouldn't have done it either. Not if I had put so much time into it xD
Until it pops
@@alexreeve I want FLAMES.
Aww, I'm disappointed, wheres my hammer.
These two videos about transformers contains tons of very useful info, thanks a lot for that!
You're welcome :-)
Indeed, i agree. Especially these two i was glad he made.
One of the best things about this channel is seeing the use of the Electronics equipment. Learning so much so quickly.
This is one of my most favorite projects/Videos of yours so far!! (And I've watched almost every one!)
Fascinating, and really well explained.
I expected the 3D printed core won't work due to much less flux density but I'm curious with a 3D printed toroidal transformer because toroidal transformer uses core material with less flux density and doesn't require layer isolation like traditional transformer does.
Maybe I will try that out in the future.
The individual windings of the core-iron acualy are isolated or relay on the oxide
@@greatscottlab How bout R core transformer too?
Not always isolated. Ive seen transformers before where the E and I pieces are welded. There is a weld line along one side.
@@simontay4851 like with microwave transformers. they are all welded. ( well the ones that i did open)
Thanks for giving all these formulas in one video. It is a great reference, for people who lost their school manuals and are too lazy or cheap to buy new ones.
You also clarified why the primary need x turns. As electrical engineers, I learned all that, but the magnetism part was not my favorite at that time.
5:10 Create 2 secondary coils and connect them in parallel!
Haha that was exactly my idea as well at that time. But since I would not create perfect windings, the voltage of the 2 coils would be slightly different. When connecting them in parallel, it can create a compensation current which can be pretty bad in worst cases. That is why I did not show it in the video.
@@greatscottlab Is it not enough to have the same number of windings?
@@greatscottlab hello sir
If slightly different voltage on each secondary winding is a problem, use 0.1 ohm balancing resistors.
GreatScott! And what about winding the wires in parallel? You cut the necessary length twice (or more times), then you wind them at the same time. That’s what I’ll have to do with my SMPS transformer.
Your magnetics videos are very interesting, informative and important. The hands-on experimentation which answers real-life questions is priceless ! Thank you for your patience, dedication to-the-art, and sound scientific approach..
I want you to know you are the reason I understand and enjoy electrical engineering!
*Let´s get started*
Those words always light up my day
I'm a soon to be graduate electric engineer, with specialization in power transmition.
Must of the time i have to pause the video, think about what he just said, diggest it, and then continue. Jajajaja. Eventually i just go back and rewatch it.
Love the vids, always learning something.
Greetings from Venezuela.
4:45 the song name is "Home Stretch" by Gunnar Olsen for those who are wondering. Its really hard to find in the TH-cam music library...
You know, I had always wondered how one could practically test the flux density value of a transformer/inductor using basic test equipment. Still an amateur at electronics so I learn something from every one of your videos. Thank you for the work you do, I find it very interesting.
you need to know the magnetic path length and cross sectional area of the core, voltage, frequency and number of turns then you can calculate B. Path length and cross sec area are located in manufacturers data sheets so your all set. Really just pick a B that you want to operate at and calculate the volts/turn.
Interesting - I was hesitating on buying a reel of this as a "better than air" substrate in a 3d-printed axial motor, but it appears to give a moderate benefit - Thanks.
You should consider it a sensor & mechanism material instead: focusing fields onto a sensor, building an armature when you want fairly precise spacing, armatures for "fluxgate compases", passive components for magnetic latches, "programmable magnets", part or all of a filter for Hall effect sensors, part or all of magnetic logic circuits, cores for "magnetic 1-wire" access cards, etc.
The stuff's just not particularly fit for motor use, but don't write it off without considering it's other possible uses.
Thank you for this video. I have learned transformers through my apprenticeship. Now I'm a Lineman. Never understood transformers to this degree.
Should have called this one "DIY or DIY: transformers"
The Best !!! :-)
Or DIY or 3DP
He only printing the core so shud be called DIY transformer core lol
DIY or BUY, you mean? :P
It might be beneficial to compare the hysteresis loops of these cores. It can be measured with a scope and a variable transformer and a few small other things. It also would be much more spectacular I guess.
The hysteresis curve of the 3d-printed core is a straight line for most practical purposes (yes, i have measured it). Its permeability is so low (about 2) that reaching the saturation in high-duty-cycle devices like transformers and motors is pretty much impossible.
You are the best in the projects. The second Einstein
You could have skipped most of the sheets and made it a solid block though. The particles are dispersed so finely that they wouldn't need laminations. Not sure how many microHz you'd gain from that though. Thinner wire and a round bobbin you can spool stuff onto with a drill could work :)
Especially for very low current requirements.
I suggested something similar. Also suggested printing the coil former out of the same material
Isn't printing specifically in sheets prevent current loss due to eddy current? Wouldn't a huge lump of block cause a massive eddy current to occur internally?
@@sto2779 A bit late, but for those wondering about this, the answer is no. The iron particles are suspended in the plastic and not continuous. The only way you're going to get significant eddy currents from that would be to heat sinter the plastic out (using a kiln at several hundred Celsius for some hours).
Interesting! Due to the high cost of isolation transformers in my part of the world I endeavored to build a 100V to 120V 50Hz isolation transformer using an inexpensive 20A 240V to 120V 60Hz step-down transformer as its basis. While disassembling the xfmr I found out why it was so cheap. It had only one winding (autotransformer design) and used 14 ga aluminum wire which I didn't recognize at first because the wire's coating looked just like copper. The xfmr came apart easily enough but I managed to drop the cheap nylon bobbin two feet onto a rug covered floor and it shattered into a dozen pieces, so now I had to build a bobbin. This I did out of 1.5mm FR4 sheets (basically uncladded PCB stock) and while doing so separated the primary and secondary halves into their own sections and this turned out fantastically. I then looked up as many online transformer calculator sites as I could find, calculated the number of turns I'd need adding taps for 105, 110, 115, 120 and 125 Volts (the original xfmr had a 5-way switch for this purpose), added a few extra turns for good measure, and rewound the core using the original 14 ga aluminum wire. When I was done I had 20 or so turns worth of wire left over but left it off so as to not upset my calculations. I then powered it up... and it worked!... but not at 20A. At best it would give three or four amps while humming badly and it will overheat if I leave it on for more than an hour or so. Not to be deterred, I added a two amp circuit breaker to the output, a temperature controlled fan and a 20uF AC motor capacitor to the secondary for a bit of PF correction and it now will work reliably for several hours, but I dare not leave it on for longer than that. Most of the drop in performance I attribute to the aluminum wire but none of the calculators I found had adjustments for aluminum. The unoptimized 60 Hz core material probably doesn't like being run on 50Hz as well. However, it will serve like this on my electronics bench until I can afford to buy a true isolation transformer or locate a used one.
Sorry if you said it and I missed, is your ferromagnetic filament conductive? If it is not conductive, you can build up a more idealistic transformer by directly printing a pair of solid EE core (without making a layered design).
Yes, thís teaches a real new lesson. This primary number-of-turns-calculation was usefull ánd rather new, that we are glad to learn from you, great Scott. So... Thanx for making this video!
Excellent procedure, result analysis, summary and presentation. Thanks
Great video, I love it when creative people do innovative things! Gets us all thinking.
5:28 - For the first few hundred turns, I was precise and exact. Then I remembered the wise words of Sweet Brown: *_Ain't Nobody Got Time Fo' Dat!_* 😂
I'm a little surprised he didn't try to use a fishing-rod reel. Those are commonly available, and do the same winding style.
Rob's web "the unknown transformer" is a handy guide for finding specs on used transformers. Thanks!
It works, at high frequency. At low frequency, the impedance a.k.a inductive reactance of the primary winding is too low, because the 3d printed core has very low inductance compared to laminated iron core. As frequency increase, the impedance also increase. xl=2πfL
So this could be valid for switch mode power supplies at high frequency? If so what frequency are you thinking?
Great as always! Remember that the printed parts doesnt have the oriented grains, as the metal one does. This impacts the mag flux.
I was thinking about that, just couldn't remember the term. As in interesting, and possibly related idea, the 3D printed core has its own form of grain orientation, i.e. the direction in which the filament is printed. Probably doesn't have much effect, but who knows: printing in a pattern that mimics the shape of a magnetic field might have some real effects. Food for thought, anyway
The limiting factor here is not the maximum flux density of the material. The 3D printed material is basicly iron particles with a lot of non-ferromagnetic stuff (the plastic) in between. The saturation of the individual particles is still the same. The limiting factor here is the high magnetizing current due to the low inductance. This in itself would not be a problem but it causes high resistive losses in the copper windings.
interesting study
Simply amazing!!. Loved how used proper calculation.
You made a nostalgie for me. My first transformer was when I was 8. I have made (with father's help) a semi manual winding machine, making windings turn-touch.
Never went to maximum B area, less heat.
lol I made my first transformer last year (when I was 33) out of a nail and some enamel wire. Real sophisticated technology, that. I had 16vac at 60hz on the primary and it wound up melting the insulation on the wire because I forgot to work out how much power the wire could handle. Which at 2mm diameter, wasn't much.
Ah, the (gladly) forgotten joys of packing the core juuuuust a tiny bit too enthusiastically (ie. "trying to put back in every single E/I that came out of it before re-winding") and slicing straight through the plastic and a few of the inner (typically primary) turns...
Every layer of turns should be isolated with special paper/cardboar or kapton tape or at least covered with a layer of clear lacker! The use of paper/cardboard has one more advantage other than isolation - it makes the winding of the next layer easier. Every layer shoud have roughly the same number of turns.
If you don't have the right size of wire you can make 2 identical winding and connect them in paralel, OR just lay 2 wires next to each other, thus doubling the cross sectional area.
at schooll they teached me to use beeswax to reduce, even inhibit noise and noticeable vibration, melt it over the corners and then hit it with a torch, so wax gets into the small empty spaces. Or any heat resistant barnish will sufice, nuts tend to loosen, so its a good measure to avoid or delay those issues. Indistrially they just weld them toogether.
The wire diameter and number of turns were always a mystery to me. this video explained it very well. thanks a lot...
Now I feel really smart for always buying or salvaging all of the transformers I use for my projects 😎 I've always wanted to build one, but it seemed way to complex, guess that feeling was right 😁
Great videos! I'll be watching these a few times and taking notes. Very useful formulas. Thank you!
I'll see you next time as well GreatScott!
Great video Great Scott ! Just like always !
for the same behaviour permeability of the core is in charge . great work
The ferrous filament composition is much more like the composition a high frequency switching core material; as your experimentation demonstrated. I think you could have better luck printing a solid core for a switching power supply. The ferrous grains isolated by the plastic in the filament would provide the eddy current protection at the higher frequency of the switching power supply.
I built my own 220-12v charger when i was in collage using an old transformer laying around lol. It was very easy, i made another with a 3v, 6v,8v, 12v and 24v output for charging Various batteries
I just subscribed (while wondering why I wasn't already).
Anyway, I assume the ferromagnetic filament for the 3D printer has some plastic or resinous content, to make it compatible with a typical printer, seeing as they normally print plastic parts. That means the total amount of iron in a given volume of printed core is less than that of a solid piece of iron, or even that of an equal volume of ferrite or powdered iron core. This would have a similar effect as having fewer iron sheets, with plastic or air (which essentially have the same magnetic properties) between them. And this bears out in your observations: versus an otherwise identical iron core inductor, an air core inductor will have a lower inductance. What you have is somewhere between iron and air (because there's "air" between the iron LOL).
I've heard about other methods of 3D printing metal, though not exactly accessible to the average hobbyist or maker, which could probably print very functional cores.
Now, as to your original question, and the topic of the video, I'd posit that it may I fact be possible, with ferromagnetic filament, to 3D print a mains transformer. Personally, I'd also print the coil former since that would increase the total iron content. But experiments need to be done to determine the size and shape of core that would be required. The end result will be larger than any existing design intended for solid iron sheets.
One experimental direction you could attempt, and might be able to make a second video on, is to take that "90 times bigger" approximation you came up with, and choose a core with 90 times the ratings of this one, but calculate the winding requirements the same as this core. Or find a compromise, such as a core with 10 times the ratings plus 9 times as many windings, or 13 & 7, or any other combination that's close to 90 without going to far under (like 11 & 8 would probably work, but with an overall lower output rating).
Another experiment to reduce the amount of "air" in the core might be trying to print a more solid core. Try, for example, 3 *thick* Es and Is, solid printed (no honeycomb!). Or even two thick halves, something halfway between an E and an L, so the two stubs of the center part of the core meet in the middle, and hold them together with a pair of rectangular printed frames on top and bottom but not passing through the coils. Food for thought.
It occurs to me that a core printed out of this stuff might be much more difficult to saturate, if not impossible... Why this matters? Maybe not much these days, but back in the days of vacuum tubes, this would be quite significant. A single-ended power amp (as opposed to push-pull) had an output transformer that needed to be built so it couldn't saturate, because its primary also carried DC current, carrying the output/power tube's plate supply current and thus being its load. They hadda put an air gap in the core so it couldn't saturate, and then make the whole thing huge to compensate for the loss in flux density resulting from the gap.
But of course, these operated from around 100Hz to 15kHz, and the biggest (rated for 25W) could be expected to see 80mA on the primary plus 56mA across it DC, and 2.5A on the secondary (designed to match the 4kΩ vacuum tube to a 4Ω speaker). But that's an exceptional example. Still, the end result would have been massive and heavy with an iron core. I wonder if it could be better made on a 3D printer?
As your observation, the problem is the filament. It can be easily done making custom filaments. Who knows what the actual percentage of iron is in the PLA, can't be more than 10%, if you saw the inductance of the ferro-PLA, it was only 78uH compared to 3.4H... that is just ridiculously small. If there was at least 50% iron in the PLA the inductance should be at least 1H. The filament scott was using is "proto-pasta ferromagnetic iron filament" and they do not disclose the iron content percentage in the filament (which is stupid). But reading reviews about this ferromagnetic filament the iron content is around 8% (no wonder inductance was 78uH). It can be easily done by making home made PLA filaments doped with higher percentage of pure iron powder, lets say 50% iron and 50% PLA. Better yet using nanocrystalline or amorphous metal powder which is 5 times more magnetically permeable than iron doped to PLA. Buying this proto pasta ferromagnetic filament is really silly for any electrical use. Now that's some real food for thought...
Fantastic topic, minimizing hysterisis and eddy current losses would be a challenging task with normal 3D printers and limited material choices. But of course, it's cheaper to buy prefabricated transformers online :)
Great video again!
I believe the properties of this iron filled PLA are very similar to ferrite which is sintered black iron. It works well in motors as demonstrated by Christoph Laimer in an earlier video series.
Great video. definitely the first ive ever seen of someone making their own transformer. have you done a switch mode power supply yet? it would be interesting to see and you might be able to get a semi functional 3d printed transformer out of it. still a bad idea but it would definitely be interesting to see.
7:04 those screws were satisfying
Can you please make a mains transformer using the Starship Coil design? And let me know what you have to compensate for!
Love your videos! I’ve learned so much from you.
Awesome! Thank you!
Great scott! Great video!
In the previous days I was working on your DIY soldering station, I need a toroidal transformer . But it is quite a bit EXPENSIVE,when I searched in Amazon india(Actually I am an Indian).
So, need DIY TOROIDAL TRANSFORMER
Hmmm.. I've always wondered if you could print transformer materials or anything relating to electric transformers.
Nice, now I get to find out!
Very interesting experiment and thanks for the info to that german transformer site.
try the magnetic filament with the white spray paint as isolation between the 3d printed plates.
Great video thank you.
Ps. I always enjoy your videos.
I did enjoy this video and I always learn something new.
Awesome :-)
if you don't have the right wire thickness, you can try "hobbykinging" it, using multiple wires to add up to the correct cross section or slightly more.
Haha that was exactly my idea as well at that time. But since I would not create perfect windings, the voltage of the multiple coils would be slightly different. When connecting them in parallel, it can create a compensation current which can be pretty bad in worst cases. That is why I did not show it in the video.
So many answers to questions that have bugged me for a long time. Thank you!
You're welcome :-)
I have been thinking about doing the 3D printed transformer bobbins, especially for custom higher frequency transformers (be it the transformer with built-in inverter driver, or just to handle high voltage), rather than using one that's available, as they're still a little bit harder to find. And 3D printed ferromagnetic transformer core may probably do okay with high frequency, beyond 60 Hz, I gotta actually test it on my 3D printed transformer at some point.
when go with higher frequency you will need to decrease number of coil turns
Ooooh please Mr. Scott do a gigantic 3D printed SMPS transformer
Best one!! Awesome,still waiting for DIY toroidal transformer
Can you make a big Video about the Wemos D1 Mini?!!
No
@@greatscottlab why not?
Because
after reading this you cant unhear him inhaling at the end of every sentence
Part of the format along with " _Let's get started!_ "
Him inhaling is nothing compared to "Luis Rossman" - If you don't know who he is, youtube him. Dude is a guru in Advanced Macbook repairs.
@@sto2779 lol of course I know who Luis is
Do the steel sheets coming with the proper coating already?
That ferro printed transformer would probably work fine for a high frequency transformer
Do you plan to try this with a ferrite core transformer ?
for example, replacing the ferrite core with a 3D printed core to see if it works in a switching power supply
I love your channel.
Thank you for your work
Amazing! So helpful and informative.
For Americans
Mains means what comes out of the wall socket or Line voltage
"Mains" actually is commonly used to describe that, among those people that have ever considered talking about it at all. Also used are your "Line" and "Wall", as well as "House" or "Household", the technically vague "120 volt" and "230 volt", and probably a number of others that just don't come to mind at the moment.
In america mains is a bit safer its just 110v ac
In Europe its just dangerous 230v
@@absalomdraconis I know. I was just using those terms for people how had no idea
@@Rainbow__cookie what about amperage? I know here in the US, we only have 120 volts but I always wondered if we have more apms because of that.
Amps*
A very great and informative video
Very interesting video! thanks you for your work. Have you considered make a third video with high frequency transformer with ferromagnetic filament core?
What about replacing the core of a fly back transformer with a 3d printed one? And also 3d printing an E core box and filling it with iron powder? Like the pla core you printed but with zero infill?
Print with highest Z-resolution, and every 19 layers of ferromagnetic filament, print one layer of normal filament, as one piece!!! It will reach a much higher efficiency, i promise!!! Greetings from Cologne, KD.
Can you elaborate exactly how and why your theory would work? Higher resolution would just prevent eddy current loss. But doesn't increase inductance.
@@sto2779 But Eddy-current-loss-reduction IS already higher efficiency... Laimert, another German YTer has done similarly, to create a 82%-efficiency motor, maybe I got the idea from that project, but to be honest, I cannot remember at all, it´s 9 months ago. Printing as 1 piece makes better usage of the volume (no air-gaps between the ferromagnetic laminates), so, more ferromagnetic material in the same volume brings higher inductance, would be a guess now, but to be honest, I cannot remember... The higher the resolution, the thiner the isolating layer, the better the volume-usage...
small main transformer always pain to winding manually 😂, high power rating ei core should wound stacked primary and secondary to minimize phase shift, minimize air gap with tighter the core to minimize leakage flux, i usually add more turns primary, to compensate fluctuation main voltage up to 250v.
What about 3D printing the coil former as well ?
It's a more complex shape to 3D print and design.
Would like to know if it's possible, as I have a plastic coil former I managed to crack, and had no luck getting a replacement.
Try to get some polyester drawing film (it normally gets used like drawing paper, but made from plastic). If you can, then look up how to do solvent welding with it, and you'll be able to do much finer-scale coil formers than a 3d printer would be able.
Really, really interesting! 😃
Thanks a lot for this!
Ferromagnetic filament might be good for high frequency and switch mode transformers. Even moreso for RF applications like baluns.
Maybe, but personally I'd look more at using it for sensor applications. Much like an iron nail, you can use it to focus magnetic fields, but unlike a nail you can trivially form it into any shape that you want.
Excellent Video, Thanks.
Yes, i wanna an Optimus Prime.. Haha Sry, I couldn't stop myself.
Don't sweat it ;-)
yeah, 3d printed motor V2 =D
I want to know if this could be used for high frequency isolation xfmrs like for fet drivers or ethernet. Also I wonder if at smps frequencies if it could be used for low power apps like running leds
Can you experiment on performance of a 3d printed transformer for a flyback converter, say outputting at most 20 watts from mains? You may take the advantage of power integration tiny switch family for simplicity
If you can also show how to rebuild a transformer based on known specific input/output voltage and windings would be great. I have a solid graphite core that I need to rewind but don't know how many Henrys it should be
Is it possible to use iron powder and resine to make a 3d printed magnet ?
yes it is possible...
I know you likely don't do requests, but is there any chance you could do a diy or buy of an LCR meter?
Clearly that filament is not a good replacement for a magnetic core, have you tried using it as shielding? I don't expect good results, but custom 3d printed shielding could be interesting.
Actually yes you can print a core, just not with classic FTM 3D Printing, but with SLS/SLM, EAM or LPF (in Short Metal) 3D Printing you can easily print a core.
Is the ferromagnetic plastic conductive? If not then no need to laminate I think.
Not conductive. But I wanted to keep its look close to a real transformer.
What a coincidence!
Today I just got myself a "C shape", really big and powerful transformer, that weights 8 kilos! And you posted a really interesting transformer video, what a great day :)
PS. I don't know, what power it can deliver, so I will have to figure it out :)
There are relations between weight of a transformer and its output power. Google it and you will find it :-)
So what about using the principle of requiring a higher frequency to an advantage and design and run one of the 3D printed versions as a switch mode transformer?
I would be curious as the results from a single solid (or as close as possible) print with 100% infill..
Many of us have had the same thought it seems
In a way it is a success. You can use it as a high frequency transformer and create a modern DC power supply.
He could maybe use it as an audio transformer too.
What about high frequency transformers like the ones used for ATX PSUs? It should work.
It could work on mains if you made a switching circuit to control it 😆 DIY smps with 3D printed transformer.
Winding the primary gets tricky, too few windings and it overheats, too many and you won't get enough voltage. I usually add about 50 extra turns to the primary and it won't overheat.
You don't need to print lamels, print a solid core. The ferric particles are insulated from each other by the plastic in the filament
Your feromagnetic 3 d material isn't conductive so you can make a big solid magnetic core and increase inductance by having no air gaps inside.
3D printing the E and I pieces with magnetic filament doesn't work but you could 3D print the plastic piece that you wound the wire around (what you called the "former").
Why not just increase the cross-sectional area to compensate for the limited Bmax? You get the benefit of a squared term as you increase the radius
GREAT 👍 SCOTT - plz make a video on mosfet based CC discharge capacity tester which could extendable to multiple bays. that would be super great,