
Resources:
Build & demo video: Click here
Coil form tool Download STL file
PCB Gerber files Download gerbers
Schematic Download Schematic
Details:
Voltage input: 12~48VDC
Current input: up to 30A (@48VDC)
Peak power input: 1,400watts
The Circuit

This induction heater is based on the popular "Royer induction heater" schematic. I believe this is the original post/author? Click here
Changes made to the design include:
Higher rated MOSFETs
Bigger Zener diodes
RC filter on the Gates
Flyback Diode (D8)
Gate resistors to limit ringing (R8, R9)
In short, all of the changes help to improve the durability of the induction heater. It has been reported with the original design, often the MOSFET gate's would be damaged from flyback after extended use. To be fair, I haven't tested the circuit without my additions so I can't confirm or deny this. Gate resistors have to be added as ringing was quite a problem without them. The ringing would cause the MOSFETs to become quite hot & this can lead to premature failure. With the addition of a pair of 18R gate resistors. The ringing was reduced to an acceptable level.
Components:
2x Wakefield-Vette, 694-50 Heatsink Click here
2x IRFP4668PBF MOSFETs Click Here
10x WIMA, MKP1J034706B00KB00, 470nF, 630VAC Capacitor Click here
2x 2.2uF Ceramic 100v Capacitor Click here
2x FR307 Fast Diode Click here
2x 47R, 5w Metal Oxide resistor Click here
2x 470R, 5w Metal Oxide resistor Click here
2x 12v, 5w Zener Diode Click here
1x SB5H100 Schottky Diode Click here
2x 100uH, 15A inductor Click here (see below for details)
2x 10k, 1/2w resistor
1x 4.7k, 1/2w resistor
1x 5mm LED of your choice (power indicator LED)
The inductors

For the inductors, I opted to wind my own. Options for this size are somewhat limited & it was cheaper to make my own anyhow. If you don't wish to make your own, then alternatively you can buy them from banggood Click here
Toroid details:
Size: 42x22x17mm
Material: Iron Powder, HY2
Color Code: Yellow/White
For the wire, I used 1.25mm (16AWG) enamel insulated copper wire. Each toroid used approx 1.6m (63inch) of wire. This length of wire yields 30~32 turns around the toroid giving an approx inductance of 100uH.
Tinning the high current traces

Located on the underside of the PCB is several exposed traces that should be tinned with PLENTY of solder after all the components have been installed. These traces carry very high current that pulses through the work coil. Failure to adequately bolster these traces with copious amounts of solder will result in the solder melting off the PCB, which will result in the trace burning out, which in turn blows up one of the MOSFETs (not that I'd have done anything like that :P )
I'd recommend laying down some fairly heavy copper wire on top of the traces & solder it in place just to be on the safe side.
The Coil

This is the work coil. I made my coil from 3/8" copper tubing (the type used for aircon, fridges.etc) It's pretty inexpensive & easy to find at most hardware stores or aircon shops. The inside diameter of my coil is 70mm to accommodate almost anything I'd need to heat.
The coils consist of 6 & 1/2 turns. The amount of turns does play a role in determining the resonating frequency (and also power output/consumption). If you know what you are doing & are willing to experiment with different coils, then go for it. Otherwise, I'd suggest making a coil with 6~8 turns.
If you watched my build video, then you'll have seen me 3D print a form tool, used to wrap the copper around to form the coil into a spiral. Strictly speaking, It's not essential however, it does produce a nice uniform coil. You can download the STL file Here
Performance

With a 48V power input and a 140mm PC fan blowing air over the heatsinks & caps. Temperatures were quite acceptable at 80c (176f) for the MOSFETs, the capacitors were about the same. And the inductors never felt even slightly warm. I think for higher input voltages (above 48VDC) You'd have to start considering spacing the caps further apart for better cooling & depending on how far you push the envelope.... maybe water cooling the MOSFETs & work coil.
Will it melt Aluminum?
The original author demonstrates his induction heater melting an aluminum heatsink. I've attempted to melt alloy in my induction heater and it failed to do so. I'm not exactly sure why he was able to melt alloy and I was not but, I do wonder if he is tuning up his power input to achieve this? He seems to indicate this in his post. It's also possible different alloy's melt easier than others so perhaps this is something I'll explore later.
What would it take to make this induction heater more powerful?
In truth, the components used could probably handle 60V (just monitor the peak voltage across the caps & make sure to stay within their max voltage rating). However, at higher power levels, you certainly need to bolster the high current traces on the PCB to prevent failure. Keeping the caps & MOSFETs within temperate limits could also be a challenge. Perhaps water-cooling would be a viable solution? In any case, it's certainly possible for this circuit to handle higher voltage than 48VDC if appropriate measures are implemented.
Hello! Thank you for your work and info on this project. I'm a fan of your channel. I went to download the Gerber files, but the link is the same as the Coil Form Tool. Is there a better link to use?
Sorry about that. The link has been fixed. Let me know if you have any further trouble 😄
No further troubles! The download went smoothly. Thank you! Keep up the great work! I'm learning a lot from your videos and very much appreciate the quality and the humor you put into them.
Do you have any of the Toroid Cores you would be willing to sell, or a link to where I might be able to buy the ones you used?
you can get the inductors from banggood (100uH 15a 27mm) for 1,8Euro
or as bundle with 3 or 5 pieces (than its cheaper)
Thanks for that! Thanks for the project schematic, I just upgraded my CNC machine and plan on using shrink fit tooling, and needed an induction heater to put the cutting bits into the tool holder. The commercial devices are prohibitively expensive, so this will work out very nicely, very much appreciated.
Hy. I can adapt the coil to heat a 100mm diameter stainless steel pipe. If so then can you help me please. Thx
ASI COMO ESTA EL DISEÑO DE LA BOBINA PUEDES CALENTAR EL ACERO DE 100mm, si cambias el diámetro de la bobina a menor, aumentaras la impedancia de bobina y cambiaras la frecuencia de resonancia, asi como esta el diseño origanal, te va a funcionar genial!
I'd suggest 5 turns of 3/8" copper tubing. To efficiently heat the SS pipe, I'd recommend making the coil a relatively close fit. Maybe 110mm internal diameter. To avoid shorting the SS pipe on the workcoil, you can insulate the workcoil with fibreglass sleeving
Thank you. I will try to do it!
我做了3个 就一个启振的 , 另两个死活不启动
Hi, what is the current through the coil at 48V and no load? Trying to figure out the magnetic field it generates.
Annoyingly I don't have the equipment to measure it accurately however, assuming I have done the math correct, theoretically around 200~240amps
How did you calculate 200-240A? Isn't your frequency around 20 kHz?
What was your approximate build cost for this, if you don't mind?
Also- I get that coil as designed will heat steel- how exactly do you determine size of coil for different metals, like you did above for stainless tube? How did you calculate diameter and turns in the first place?
I've wanted one of these FOREVER- this looks actually buildable- thank you so much for posting all this!
It cost me about $150NZD (approx. $100USD) for everything incl the coil. When it comes to coil design, the math is enough to give most people a headache (me included) Google has proven to be most useful for this topic. Generally speaking the larger the coil, fewer turns are needed as the overall length of the coil is longer compared to a smaller diameter coil. The coil plays a major role in determining the maximum power output & power consumption. It's design can be altered to some degree with little risk but, it's safer to have more turns, than not enough.
Glad you found the build useful. I've started building a cost effective PSU for this heater & I'm working on a more powerful induction heater as well. Plenty more to come :)
My hackerspace is now privately designing a 10 kw induction furnace to melt iron and steel, subbed you for the awesome response here. More than any other electrical item, I would jump into learning differential calculus all over again if it meant understanding these types of heaters well. I have great need for this, for heat shrink tooling (machinist by day), remote blacksmithing (I have a commission waiting and this would make blacksmithing super quick) and wonderful for heat treating my own tooling a lot quicker.
Please keep posting more of this and how to make them even more powerful! I suck at electrical stuff- but think I can make this thanks to you. Have you considered making your board out of talc block, and instead of traces, using solid wire laud in channels in the block? Would stand heat well...
Keep in mind induction heating requires a ferromagnetic susceptor (whatever is placed in the coil) If you want to heat something that is not ferromagnetic, make yourself a ferromagnetic tube or sheath to fit around around your part. It should fit fairly well since you are depending upon radiation to heat your part with the susceptor. Ferromagnetic means something like iron that is attracted to a magnet.
What is the flyback diode you used? I can't make out the number.
It's a SB5H100
Thanks so much!
@Schematix Ok, maybe a stupid question, but are the 18R gate resistors just an 18omh resistor and what wattage would it be? Does anyone know where I could find some?
So I don't want this to sound bad. I'm genuinely curious.
Why am I better off building this than buying a zvs from Amazon / ebay etc?
I'm thinking I know what I'm getting here, I like building. QC is on me.
That said they have economy of scale. I'm thinking they could crank them out cheaper than I can get parts.
They also probably use crap parts.
Is that about it or I'm missing something?
Also meant to say. It's awesome and thanks for putting this together. I looked at a lot of builds and this is the cleanest by far. No idea how it performs compared to others but I feel I can build it.
It's a valid question. I have bought a cheap 1000w ZVS induction heater & still use it to this day. The performance difference between the 1000w unit & my DIY version is night and day. The cheap ZVS induction heater is roughly about 30~40% the output performance compared to my design. And naturally, the components for my unit are cherry picked for quality. The cheap 1000w heaters do have their uses but, in truth their performance figures are overinflated which is something to take into consideration when choosing between "Build it" or "Buy it"
Awesome. I'll build yours and see how it goes. I was hoping you would say that as I wanted to build haha.
Follow up question. I'm off grid and running a 48v (nominal) bank and charger.
I was happy as hell to see you suggest 48v because I can tap right in to it and don't need a power supply etc.
Any reason that's a terrible idea? I can pull a free 200a from the sun at 48v out of my charge controllers. Seems too good haha.
I'm interested in making one of these for knife-making, and was wondering if you know what the impact on the performance would be having a non-circular coil. Knives are made from flat bar stock, 2-4 mm thick and anything up to 70-90 mm wide. Would it be possible to optimise the performance of this unit for flat stock by making a flatter oval-shaped coil?
Replying to the question about the coil being oval shape. I will work just fine as virtually any shape as long as the susceptor (object to be heated) is close to the same shape. Only critical in terms of efficiency.
I am buying this now:
1. https://aliexpress.ru/item/32992323686.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.7594166.52695243.1608725283-59860447.1608725283&_gac=1.45835280.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE
2. https://aliexpress.ru/item/33042599841.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.119631756.52695243.1608725283-59860447.1608725283&_gac=1.16502658.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE
3. https://aliexpress.ru/item/4000086392320.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.119631756.52695243.1608725283-59860447.1608725283&_gac=1.16502658.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE
4. https://aliexpress.ru/item/32895796381.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.108316678.52695243.1608725283-59860447.1608725283&_gac=1.245564080.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE&sku_id=65773737933
5. https://aliexpress.ru/item/4000892118309.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.173662247.52695243.1608725283-59860447.1608725283&_gac=1.216794658.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE&sku_id=10000010457488490
6. https://aliexpress.ru/item/33009200272.html?spm=a2g0o.cart.0.0.51723c00Ta09W7&mp=1&_ga=2.173662247.52695243.1608725283-59860447.1608725283&_gac=1.216794658.1608974643.EAIaIQobChMIw-q5yanr7QIVC9WyCh2B6gpWEAQYASABEgId2fD_BwE
I will blow up this machine or it will ruin me.😁
Good power supplies with voltage regulation. But my wallet is weak. Meanwell RSP-1600-48 or Meanwell RSP-2000-48
Moser will ship the IRFP4668PBF transistors to my country in March. Now they are out of stock. It is sad.
The radiator for the transistor must be disassembled on an electric stove. They seem to have it fixed with glue. And solder it with solder. The transistor will need a low melting point solder. For example: KOKI Solder Paste SS48-M955
Melting point 179-173 degrees Celsius. And then the radiator fins will not fall off.
Or: SE48-M955. SS48-M955. SSA48-M955
Found the power supply unit RSP-1500-48 Mean well, not new. But the price is normal - $ 84. Well, delivery, 15 dollars.
Hello, I am currently building this. I was wondering: what are the specifications for the 18Ohms Gate resistors (R8 & R9)? Cheers and greetings from germany
Johannes
You can see it in the photo at the beginning of the post. He is the smallest. There are no conditions. The current through it is minimal. This can be traced to the datasheet of the transistor.
Some things have changed. I will cool the condensers through a thermal pad.
https://aliexpress.ru/item/32844227002.html?spm=a2g0o.cart.0.0.89993c00a1LO3p&mp=1&_ga=2.102736257.1280038651.1609337131-1866543773.1609132623
2. https://aliexpress.ru/item/4000402051726.html?spm=a2g0o.cart.0.0.89993c00a1LO3p&mp=1&_ga=2.102736257.1280038651.1609337131-1866543773.1609132623&sku_id=10000001655687885
or
2.1 https://aliexpress.ru/item/4001041690247.html?spm=a2g0o.cart.0.0.89993c00a1LO3p&mp=1&_ga=2.102736257.1280038651.1609337131-1866543773.1609132623
or
2.2 https://aliexpress.ru/item/4000523964139.html?spm=a2g0o.cart.0.0.89993c00a1LO3p&mp=1&_ga=2.102736257.1280038651.1609337131-1866543773.1609132623 My
3. https://aliexpress.ru/item/33015186114.html?spm=a2g0s.9042311.0.0.16f833ed1Ry28m&sku_id=67205344112
I changed the pump
https://aliexpress.ru/item/4000570403968.html?spm=a2g0s.9042311.0.0.16f833ed1Ry28m
I'm broke
Who knows - this is a normal algorithm voltage sequence? And what should you turn on first? Start or Main? What beads would be better to put on R8 and R9? http://www.amidoncorp.com/small-ferrite-beads/?sort=featured&page=3 Two beads for each resistor? On resistors R2, R5 and R1, R4 is the same desirable? Who's to say? On one of the switches to put on delay timer with relay?
Hmm. Are there any live ludi here? Or I'm in 28 Days Later
Ok, maybe a stupid question, but are the 18R gate resistors just an 18omh resistor and what wattage would it be? Does anyone know where I could find some?
Anyone?
@lamon_j I think these resistors control the gate on the mosfet, which typically does not draw much current, so the resistors could be low wattage types. It needs the circuit designer to specify though.