WARNING: This circuit analysis deals with high-voltage, high-current induction heater power stages. Touching one of the HV nodes (600V @ 30A) of these types of circuits would probably be instantly lethal. We take no liability for your actions, so if you don’t know what you’re doing – stay away!
Hello Fellow ‘Schemers,
Today we’re looking into the guts of an induction cooktop. For those of you not familiar with induction cooking, the theory is as follows:
By running a big current through a coil, you can induce an incredibly massive current in a ferromagnetic (say, steel) pot sitting on top of that coil. Since the steel is very conductive, there is essentially no voltage built up across the metal but plenty of current still flows. This heats the pot through resistive heating and eddy losses, transforming the pot itself into the heating surface. Cool stuff. Want to heat a pan? Apply power directly to the atoms of the pan through electromagnetic tomfoolery. Go check out Induction Heating at Wikipedia for a more complete description.
Induction heating has become a meandering side interest here at oschemes, due to our desire to build a sweet induction forge for a metalsmithing buddy. The web has a nice selection of induction heating (IH) projects, including
- Neon John’s Open-Source IH. Pretty fantastic, and he has a kit or prebuilt devices for instant gratification
- Richie Burnett’s IH Discussion and Projects. IH 101 and 102 all in one page. Brilliant description, beautiful theory and plenty of sims and scope shots to illustrate the details. Plus killer glowing metal pics to show that he ain’t screwin around.
- Uzzors2k PLL Based IH. A self-tuning masterpiece. Really great stuff for when you’re ready to expand your mind to orthogonal V/I phase control.
All of those pages have additional links to even more information, so enjoy! We’ve built a couple self-biased Royer Oscillators at a hundred or so watts, and have gotten metal hot enough to burn ourselves. Ow! Dammit, it works! But never any orange-glowing slag or anything else deliciously lethal – that’s next.
So as a starting point for our own designs, we always like to see what’s going on in the industry. Sure the Bosch’s and the Westinghouses of the world can build some cool stuff, but what about the minimalistic approach? High power, low cost! For that, we decided to take a look into a low-priced induction cooktop available around the web. Ours was purchased from Amazon for $70, the Max Burton Model 6000 1800W induction cooktop.
Fig 1 – Yum, Looks like lemon and corpse soup!
As an appliance… Well there’s a reason this model’s on the sale rack. The switching noise (high pitched whine) is earsplitting for those of you that can hear it, and is modulated by the position of the pan. In other words, if you were to wiggle your pancakes or stirfry pan around the noise will cycle from inaudible to teeth-grating and back. Have fun seeking the perfect pan position where the noise is tolerable! It seems to get better after warming up, but it is never silent.
There’s also a weird startup noise that sounds like the device is ramping the frequency (phase?) back and forth to try to check the resonance. This seems like a legitimate technique to us and probably a necessity with jerky customers moving their pancake pans around. But please, designers – continually seek resonance, don’t sweep it 4 or 5 times and then have to resweep whenever the pan moves or is lifted. Ugly!
All bitching aside, this piece of crap can certainly deliver power to a load. It will boil a small pool of water almost instantly (err, after the 3 seconds required to find resonance) and can even heat thin copper such as PCB traces if you trick it by having a pan partially on the coil area. That’s interesting because non-ferrous metals are usually hard to heat at the 20kHz frequency this thing seems to run at, but fun nonetheless. Not sure if it could deliver enough power to reflow a PCB but there’s probably risk of blowing up your components from induced currents, so that kooky idea will have to live on the shelf a while
Let’s look a little further, shall we?
Continued on Next Page… Jump to Page 2
By jpr November 21, 2010 - 6:18 pm
A bunch of the images on page 4 and 5 seem to be broken, please fix.
Otherwise excellent article, thanks
By openschemes December 5, 2010 - 1:50 pm
Thx for the heads up!
By Circuit Analysis of the 1.8kW Induction Hotplate « Openschemes December 9, 2010 - 6:49 pm
[...] dumperoslash on SPMP8k FRMorp – USB IMG dumperfredoagain on FRMorp v1.2 – bugfix!openschemes on 1800W Induction Cooktop TeardownMost [...]
By gene September 20, 2011 - 5:06 pm
hello,
i am just interested in doing some PEMF tests on myself, the induction cooking device is a ready made PEMF device that usually retails for thousands of dollars and is much the same as far as i can tell.
but how would one safely make this device work without the metal on top to trigger the coil to work at maximum all the time?
thank you,
gene
By openschemes September 20, 2011 - 8:47 pm
Try our article about Manual Control of the 1.8kW Induction Cooktop. That’s exactly what we’re doing there
By gene October 10, 2011 - 11:55 pm
i read your beautiful article.
but although i am technically minded i am not familiar with electronics; if the coil is tricked to be “ON” when there is no metal on it would it heat up to a self damaging degree?
thank you,
gene
By openschemes October 13, 2011 - 4:40 pm
You will want to run at a low power if there is no load. If there is no load, then the voltage ringing will be higher than if there is a load present because there is no place for the energy to go. If it gets too high, it will pop the power switches. The coil is at no risk of overheating – frankly, if any wiring were to burn out it’d be the PCB but that’s still no risk.
If you are going to do loaded/unloaded operation, please consider a foot pedal to increase the power setting once you have the object to be heated at the coil.
By Smiffy October 22, 2011 - 6:48 pm
Great stuff lads, very interesting (and entertaining!)
I came across this article whilst trying to figure out if an induction cooker with enough oomph can be modified into a small heater for silver/gold melting. The idea I had is to swap out the L~C circuit for something a bit more industrial that I can drop a small crucible into. Temps required would be around 600 to 1500 degrees C, variable as required.
The unit I have in mind is the Buffalo CE208 (3kW) which sounds like it has enough juice, but I need to figure out (with my limited understanding of electronics) if a water-cooled, tuned tubular coil/cap setup can be used as the output stage. Typical coil inductance is a few uH in normal heaters if I’m not mistaken? Not sure about frequency though (65kHz??)
Do you think this is possible without nuking the IGBTs? Or myself? =:O lol
Your input greatly appreciated!
By openschemes October 23, 2011 - 4:52 pm
Hmm – maybe. Metals such as silver and gold are very conductive, meaning that you have to put a TON of power into them in order for them to heat up. Also, the lower frequencies of these stoves may not be great for heating the metal directly. BUT…
If you get yourself a graphite crucible then the graphite can be easily heated by even these stoves, and you probably won’t need 3kW to melt a modest (1oz-ish) amount of silver if the crucible is sucking up the power so easily.
As for the coil design – this coil is about 50 uH whereas a few spirals of copper tubing is barely 1uH. So they are pretty much incompatible. You could use many spirals of copper tubing or find some way to insulate/cool the litz wire coil from this beast. Rewound into a cylinder shape, of course.
The best idea may be to leave the coil as a pancake, use a flat graphite crucible/ingot mold, and place said mold on top of a thin ceramic tile that is NOT touching the work coil. Then direct a fan upwards to try to keep the tile’s heat from melting the coil’s insulation. Or something like that.
By Smiffy October 23, 2011 - 7:26 pm
Thanks for the reply, that’s given me some great ideas & potentially made the job a whole lot cheaper!
I’ve heard of indirect heating but hadn’t really given it much thought until now & that definitely sounds like the way to go. Do you think the work coil could be re-wound as a cylinder with a heat shield between it & the crucible or will this radically change the inductive characteristics of the coil?
I would like eventually to get a small system up & running with a remotely operated ‘hatch’ at the base of the crucible, & the whole lot enclosed in a mini vacuum chamber. Moulding & general metal quality (esp with silver) is a lot better this way as it’s prone to porosity & nasty oxidation bubbles.
Thanks for opening up some very cool possibilities!!
By openschemes October 25, 2011 - 7:51 pm
Glad to help! You can certainly re-wind the work coil as a cylinder. In fact, ripping apart the pancake and winding a coil on PVC was one of our earliest experiments on the Burton stove. Wear gloves, the damn lacquer used to hold the original coil in place is as sharp as a razor blade when you break it up by unwinding the pancake.
The vacuum chamber is a great idea, as you can heat right through the sidewalls and you don’t have to worry about vacuum sealing a bunch of hot power terminals. Are you thinking about melting and casting all in the vacuum? Just trying to imagine how to flip that hot tube upside down to get the melt into the mold. Not our area of expertise, of course.
Keep us posted on your project – we’d be happy to post your writeup here if you end up with a system you’d like to share with others.
By Smiffy October 28, 2011 - 9:19 pm
Once again, thanks for the info & reply! I was wondering about the resin holding the coil together, thought that might be an issue & will try a hair drier to soften it up first. As for the vacuum chamber, I’m planning on incorporating the coil, crucible, trapdoor actuator for pouring/vacuum release timing electronics yada yada inside the chamber. I’m currently looking for a suitable weldable steel box section over 300mm round or square to house it all. It’s a pretty mammoth job, but the results will be well worth the effort.
Hey, how cool is this?…. I managed to get the entire guts of the 3kW Buffalo cooker without the casing etc for £56 here:
http://www.nisbets.co.uk/products/productdetail.asp?productCode=CE208 (under the Spares & Accessories tab) seems to be only available within the UK though?
One issue that I’m perplexed by is how to incorporate a thermistor in direct contact with the outside surface of the coil while avoiding a)vaporizing it by inductive heating! b)shielding it from magnetic interference & getting a false voltage from it. Any ideas? I just wanted a way to monitor the coil for overheats, but if this is as challenging as I think in this kind of environment, I may just run it & pray. At some point I want accurate control of the heat for things like annealing & tempering, so will be incorporating a laser thermometer directed into the crucible area since a pyrometer inside the coil will definitely vanish in short order! lol.
Delivery of the parts Monday, next step is the crucible & insulation. That will probably be a graphite tube with two spaced layers of ceramic tubing to support & insulate it plus fiberglass cloth shielding for the coil turns. etc etc. Will take some pics of the setup as I go.
By openschemes October 29, 2011 - 7:37 am
As far as the resin – it was just broken up by forcefully unwinding the coil while wearing gloves. Heat may soften it, but maybe not as it’s intended to endure heat all the time. Watch out that the outer steel box you plan to hold this project is not too close – even with a cylindrical coil it may pick up some field and steal power away from your melt (while heating itself, of course).
That deal on the 3kW heater really is great! Keep in mind that the Buffalo device may be very different from the Burton we opened up here, but it should be no problem to mod that one for similar manual control. Looking forward to seeing pics!
The Burton heater has a thermistor stuck right in the small central hole in the pancake coil to check overheating of the hob surface. That probably picks up as much stray field as your external thermistor would, meaning that your idea will probably work fine without shielding or anything. The saving grace might be that the turns ratio between the coil and your thermistor is low (thermistor = 1 turn) as well as the fact that the thermistor being outside the coil is not well coupled so the actual current induced in the thermistor is probably nil. If your coil wiring is horizontal, then run the thermistor wiring vertically in order to cut down on the coupling.
By Smiffy November 5, 2011 - 11:31 am
Hey you guys, it’s me again.
Eventually got all the parts & find that the 3kW model is almost identical to the 1.8kW version, with a few components shuffled around on the mainboard. The only real difference is the control board, which has a completely different user interface system – no logic buttons on this model, according to pictures of the complete unit it’s operated with a single rotary control knob (& nothing else) which I don’t have & a 4 pin socket on the control board to attach *something*. Tracing the pinouts on the socket, to the best of my ability probably incorrectly, I arrive at what’s shown in the pic (see link at end)
As simple as that sounds, it leaves me completely stumped over how to fire up the mainboard correctly. What do you reckon that would be? A simple rotary pot, or say a 10 position selector with resistors/something else? As you can probably tell, I’m at the edge of my understanding of electronics here lol, & with hindsight I think I should have just bought a complete machine! Tried contacting Better Co for info but no response.
Lastly, the 3 pin socket on the mainboard (red circle next to the ribbon connector) – that’s for the thermistor right? Do you have a value for that component?
Thanks for your help! (and this noob apologises for pestering)
mainboard: http://i571.photobucket.com/albums/ss155/sirius_gem/MainBoard.jpg
control: http://i571.photobucket.com/albums/ss155/sirius_gem/Control.jpg
By openschemes November 5, 2011 - 5:49 pm
Wow, you lucked the f*ck out in that it is so similar – almost identical, just like you said! Great pics, and you’re definitely not pestering by the way – spawing new misuse of electronic gadgets is why this site is here!
The best suggestion would be to control the device manually using a potentiometer as described in our article on modding the 1.8kW cooktop, but since this mod may take a little more discussion than what we can fit here it might be a good time for us to introduce The Forum!
Why don’t you go over to the forum and start a topic about this 3kW mod. You can either place it in the Site Projects area under a topic like “1.8kW mod for 3kW hob”, or under the Your Projects area with whatever name you choose. We’ll keep an eye out for your post (or you can msg us here) and then we’ll get into the gory details of how to control this beast. Looking forward to it!
By Smiffy November 5, 2011 - 6:45 pm
AHA! Manual control… AWESOME! I bookmarked that at the beginning & for some reason never revisited. So far it sounds remarkably simple, heading over to the forum after I’ve read through again. >:D
By Jeff Clithero December 20, 2011 - 12:23 pm
Any clue as to where to snip to disable the beeper?
By Jeff Clithero December 20, 2011 - 1:14 pm
Ok, found it, next to the ribbon cable that connects from the controls to the main board. It’s a black round cylindrical thing with a little hole in the top. There is nothing to clip, so I stuff some glue into the hole and it’s almost silent now. Now if I can find a quieter fan…
By robert August 6, 2012 - 12:28 am
hi,
Although technically not that advanced, but can someone explain to me if such circuit, designed for 220-50Hz, would work without alterations on 220-60Hz too? Or should I get another cooker? My device is the BT-310K Better Co. Ltd. “designed” for the EU market.
Thank you
By openschemes August 11, 2012 - 9:19 am
It should work fine.
By Bob January 10, 2013 - 9:29 am
Can you take the copper coil and use it remotly from the PCB and all other components? I would like to have the electronics about 5 feet away from the copper coil.
By openschemes January 10, 2013 - 4:20 pm
Probably would be fine, if you could use a big thick piece of litz wire to avoid heating from skin effect. A copper tube would be another option, although an uninsulated copper tube would be a death trap! You might want to try increasing the distance gradually to see if your lines to the coil are getting hot. Of course, don’t touch any bare metal to see if it’s hot – again, it will kill you before you can remove your hand.
By Anton January 23, 2013 - 3:08 pm
Thanks for the instructive article.
I would like to hack my induction cooker for heating the 6 mm stainless steel pipe. It is a part of my research project.
I have tried to replace a flat coil with a self-wound spiral one (tried few of them). In general, it works fine for the inductances 50-200 uH with a big metal rod inside (2 cm diameter). The original flat coil inductance is ~100 uH.
While playing around, I have encountered two problems:
1) neither of the self wound coils reacts on the 6 mm stainless steel pipe, only on big objects.
I have heard that stainless steel is not particularly well suited for inductive heating. Does it make sense trying at all?
2) when tested with a 5 uH coil with a thick metal rod inside (15 uH altogether), the cooker got broken. Investigations have shown that IGBT has been grilled. Without any load this coil was doing no harm, i.e. switching the cooker off.
Thanks for any comment on the two items.
Anton
By openschemes January 23, 2013 - 9:47 pm
1) Real stainless steel isn’t ferromagnetic – ie, a magnet won’t stick to it. So it’s not a great choice for this type of heater – if you’re not opposed to experimentation, you may try heating it red hot using conventional methods, and then switching to induction. Or use an envelope of something that IS susceptible to induction heating at these frequencies – for example a thick graphite tube that holds your desired stainless tube.
2) 15uH is going to be too fast for the IGBT, so they were probably still turning on/off when the resonant peak happened and the big spike is what killed them. It’s happened many times when experimenting on the bench here, those damn IGBT’s get expensive!
By Anton January 24, 2013 - 6:20 am
That explains something…
Both tips will help me get further.
Thanks for the prompt reply.
By Anton January 25, 2013 - 9:56 am
one more question.
How to reduce the operation threshold, when the device won’t switch on anymore because the pot detector sees not enough load.
My object is now an 8 mm ferromagnetic pipe with 1 mm wall thickness. When inserted into a 5 cm air coil, it is unable to trigger the operation, so that my cooker switches off right after pushing the ‘start’ button.
By openschemes January 29, 2013 - 6:07 am
Not sure. Which cooker are you using?
By sunil george February 2, 2013 - 7:27 am
I want to know the values of U3 and R15 on th board. Cockroaches have caused damage to both of them.
By Anton February 5, 2013 - 4:29 am
The cooker is “Tarrington House IC 2009″ with 1800W power (sold under ALASKA brand name).
Here is a link to a photo with a few text labels:
http://www.freeimagehosting.net/bs7le
The pcb design appears very much similar to the cooker described in Fig. 12 of your article (an evil twin).
By Liam February 9, 2013 - 4:44 pm
Hi all,
Stumbling across this article is nearly a god send to me. Ok I am a student Product Designer in my final year in University and I am developing a product for my Final Year Project which will be using induction to heat a part (disk of ferromagnetic material say 50-80mm diameter and 1-7mm thick) of the product above the coil. The Electronics department want me to do and prove a few things before they help so been an amateur/ enthusiast in electronics the internet and expert are my next line of help so that is where you might be able the help me.
I have a few questions I hope you could answer (FYI I look to have the same circuit as Anton posted in the last post:
1) Can the coil be unwinded a few windes or rewinded tighter to a smaller diameter to make the coil diameter 80mm to 100mm but still have an effective current for the material I will require to heat or can I buy a smaller coil and control it with this electronic configuration or do they go hand in hand?
2) Probably answered here somewhere can I turn off the safety sensor that doesn’t allow the system to turn on unless there is a certain amount of metal sensed in the magnetic field?
3) The other thing would anyone have the wiring diagram for the type of circuit in Anton’s images above, as the electronics department said I would need it if I wanted some help later in the project ( hoping to get a mounted circuit made instead of the through hold circuit to make it more compact for my project)
Thanks in advance for any of your help, my student email is 09006782@studentmail.ul.ie if anyone wants to contact/ give me some advice/ help
By openschemes February 14, 2013 - 7:37 pm
Glad it’s a help to you.
1) Yes, the winding can be changed. It’s probably better to rewind it tighter (higher L, lower frequency) because it will make the cooker’s work easier.
2) No idea if this is possible. We built our own board that does not use the manufacturer’s MCU, so it doesn’t have any protection. It should be on the site somewhere, if not we will get someone to dig it up for you.
3) The wiring diagram is not available unless Anton or another owner of the device traces it out. At least we don’t have it.
By Anton February 25, 2013 - 5:33 am
Hi Liam,
I was not able to find a circuit diagram for my PCB. Here is a link to another wiring diagram with some details included:
http://www.speedshare.org/download.php?id=8671285B1
Regards, Anton.
By mithun February 13, 2013 - 12:35 am
can anybody provide me the software program of the micro-controller. plz contact me at mithun_as@hotmail.com
By openschemes February 14, 2013 - 7:33 pm
It is probably not possible to read the software of any cooker micro. You would have to do it yourself.
By Robert February 14, 2013 - 12:09 pm
Hello
I would like to ask you I have same 1800W Induction Cooktop Teardown and I want to use it as induction heater but I need to increase frequency for min 100-200kHz is it possible ???
By Max February 25, 2013 - 6:12 am
Can anybody provide me the software program for microcontroller. PLZ HELP max_83@in.com
By Nelu March 15, 2013 - 11:05 am
I welcome everybody here.! I have a plate the same as the picture Anton. Tarrington House model IC 2011b. I want to increase to about 5000W power IGBT’s by changing to a more powerful (and of course the rectifier bridge will be increased and capacitors …). ?
Current control circuit can support a stronger IGBT. ?
I ask this because I want to experiment with current motherboard to work with another IGBT stronger.
Also, I made a coil higher power formula here:
http://www.circuits.dk/single-layer-air-core-inductor-calculator/
and using information gathered about the original coil.
I want to instantly heat up a pipe diameter of 30-40mm and 300-400mm in length passes through the water.
With thanks to all and especially to Openschemes !
By rahul March 19, 2013 - 8:33 pm
i had a small question
heatsink and fan present in the circuitry
exactly which components do they serve to cool?
can the coil be placed away from the main circuits (using a longer wire) and still be used to heat, without overheating itself?
By openschemes March 19, 2013 - 11:47 pm
The heatsink and fan cool the rectifier bridge and more importantly, the IGBT’s. The coil can be placed away from the main circuits if it is a low impedance connection, hopefully using the same type and size of wire as the main coil.