DIY Induction Heater Builds and References

[TommyDee]

Box Troll


Alright; we're gonna fuck with the Box Troll above.

We will be doing a comprehensive build of one version of HalfPint, this being the flat version with components on the opposite side. I have made this decision based on several factors in no particular order;
1) The light will be facing up on the IH board.
2) The component side (side with the FETs) has the traces (planes) for the coil. I want to preserve the coil arrangement to some extent. The backside has poor coil connections.
3) I've got something in mind that requires a little real-estate available on the component side - in PCB parlance.

I have a module that came a week or two ago; a recent donor to the cause.

It was apart before I could grab the pix...



There is a trick to getting the blue connectors out without damaging the boards;
Screw the screws of the contact all the way in.
Drill the screw holes at 4mm or 5/32" to remove a little of the blue plastic. Hand drill works if you can get a grip.
Now remove the screws from the connector. They won't be captive anymore.
The 2-pin blue housing will slide off now. The 3-pin will need the inductors remove, which need the capacitors removed.
In the end, remove everything not nailed down and clean all the holes.



For some of you who have worked with this board for a while will notice a layout change. The long trace from the center of the 3-pin connector is no longer there, nor is the unused hole that use to be attached to it. If you didn't know of this trace or haven't noticed, no worries.



Please be aware of what all that stippled diamond hatching is about. It is your safety net. It is not connected to anything. Keep it that way!



We're going to make two modifications to the board right off the bat. First we make for a good solid ground pad, and second we fuck MOSFET Switches out of the circuit. This device can function stand-alone with a switch and only two high current leads coming in from any source. That is where this project aims to go.

First we make available a proper ground connection. A hobby knife makes quick work of removing the epoxy by just chipping and scritching it away. Small bits and steady scraping with the sharp edged tip.



When you put on the 3.25x Dollar General glasses, it'll look like this;



And when were done with that we make sure the whole trace is tinned well with solder;



Now we clean the board with ISO and flip it over.

We're going to make an incidental yet critical cut in the very circuit design. Before we do I want to show what we are doing. Please pay very little attention to all the words, but do focus on the momentary switch just South-East of the credits. I added -that-.



...And I moved the + symbol to the junction of the two Inductors. Notice how little the +positive power is connected to just the inductors and the FETS. FETs are N.O. devices [normally open]. These are, anyway. And look at all the components that are energized downstream of the switch. All low power as far as the switch is concerned. The only concern is that it remain momentary for safety, please. My only warning.

So we are splicing in a Fire switch according to the schematic. You can have as many Fire switches as you want. Wire the switches such that this trace cut is bridged. Many places to make the cut but this is convenient.



Literally cutting a trace! LOL Okay. The top half is positive high power input. The bottom is the switch output to power the "Gate Circuit" that we spliced the switch into.

The secret to this circuit is that the IH replicates the MOSFET switch already. The IH -is- an oscillating MOSFET switch!

I said I would make a stand alone module based on this HalfPint Build. The input wires from this build can come from a MOSFET switch if you want; You would drive that MOSFET switch with an On/Off switch of low power. You will still have redundant protection. The difference is that the fire button is no longer tethered to the MOSFET switch. It is now part of the IH circuit - proper. Only 2 heavy wires going in from wherethefuckever power source you supply.

Stay tuned, this is going to get exciting.

 

[TommyDee]

To continue, we need to cut the 2 inductors in half. Yet, we still need 3 connection, knowing that one pair are a common electrical connection. Please refer back to the diagram to see what we are connecting. That should give you an idea of what HalfPint is all about, and how it got it's name.

We start with finding dead-center of the inductor. We will be adding 'center-tap' to effectively making it two coils with a common connection. This common connection is the positive input for the IH.

I divided both sides up using permanent markets. Where ever the center is, in this case at 15+, is where we need to add a wire.



Separate the coils near this center wind. You need to scrape the insulation from the wire's 'face' and along its edges. I used the little screwdriver because it was sharp enough to remove the epoxy from the wire. You want to make copper available along much of this lengh.



Add solder to the scraped area and make sure is is clean and well-wetted.



We will be structuring this inductor to the capacitor to maintain the full winding provided. Loosing winds on HalfPint is not recommended for radio interference purposes. Therefore I am adding wire, 14 gauge from the hardware store to be exact, in order to make a robust structure in the end.

We start by shaping the wires as we want them to be. Relaxed wires are always preferable.



Next, size, tin, and bend the add-on wire and use the 3rd hand to tack solder these to the inductor.



Get set up to where you can connect these two connections to the capacitor. In the end, it should look something like this...



Now we need to prep the 3rd leg. I've chosen some 18 gauge magnet wire to make both the positive connection 'loop' an the coil connection. This is the only thing that will be connected here. If you want to run a switch wire from under the induction, solder a spare wire before fitting the last sub-assembly. Please note that you have two options. I chose to push the capacitor further inboard. That was another new layout compromise - they separated the two pads for the capacitors. Now they hang out over the edge.

Get these wires set up so they are ready to solder to the tinned wire on the inductor. Everything should fit comfortable without forcing things into place.




Once the sub-assembly is soldered in place, congratulations, you've build a HalfPint.

 

[TommyDee]

(Mods, ran into an image limit)

Now one other thing... power is added. One could put the 2-pin connector back and use the negative as input, and the positive input as a Fire-switch return line. Positive has to be added to the loop we made or also within the inductor. Lots of options here. This was mine -



Now this is only one of many ways to get here. It is a bit artisan but that is what we DIY peeps strive to attain. Have fun in your frustration. You have 2 coils, fuck one up. You have 2 caps - break one if you can. But above all, know that these connections have to be solid and robust. Solder should look wet all the time even when cooled and hard. Those traces of copper on the circuit board are no different than the wires you add to them. They're just flat wires glued to fiberglass.

In the mangled words of Dynatec Chris, the Mini ZVS was too hot! I happened to agree.

 

[TommyDee]

I had a question posed as to what is soldered to what.

The capacitor and the coil are a subassembly;

I used 14 gauge solid house wire to extend the legs of the coil and spliced them into the legs of the capacitor.

Set aside and prepare the board...



This wire is the positive input. There is a loop on the other side...



...to which we solder the positive wire...



As to the capacitor, we chose which hole set we wanted to use when the bent up the positive wire to connector to the coil - to refresh;



should look like this -



...at the bottom of the image, that solder junction should be 'tight' before you even touch heat to it. Take the time to tweak the wires exactly there you want then to get a quick solid tac-solder joint. Once the wire is tweaked into place, solder the legs of the capacitor (which already has the choke/inductor attached). After soldering the capacitor, make the center-tap connection on the inductor to the fixed positive wire.

This modifications makes a pair of inductors share a common core. It also separates the power circuit from the control circuit where we made the slice in the trace. This inductor center-tap trace is the power circuit and the positive input lead on the 2-pin connector is now the IH 'fire' connection (positive output from the momentary switch)

Clear as mud?

 

[CaleidosCope]

...to which we solder the positive wire...

View attachment 2803



Clear as mud?
[/QUOTE]


Almost... just one question: Where does the epoxy-wire go to? Switch?

 

[CaleidosCope]

This inductor center-tap trace is the power circuit and the positive input lead on the 2-pin connector is now the IH 'fire' connection (positive output from the momentary switch)

... I thing you named it... just give my poor brain a picture...please!

 

[mecrobs]

Almost... just one question: Where does the epoxy-wire go to? Switch?

If you're talking about the red 18g wire running across the top of the board yeah I believe that is the switch, you can see another shorty wire coming off the positive "loop" junction. I think those are for the switch. You got some options as to how/where you can run those. I might run both mine on the backside. This post sums it up pretty clearly.

(Mods, ran into an image limit)

Now one other thing... power is added. One could put the 2-pin connector back and use the negative as input, and the positive input as a Fire-switch return line. Positive has to be added to the loop we made or also within the inductor. Lots of options here. This was mine -

View attachment 2787

Now this is only one of many ways to get here. It is a bit artisan but that is what we DIY peeps strive to attain. Have fun in your frustration. You have 2 coils, fuck one up. You have 2 caps - break one if you can. But above all, know that these connections have to be solid and robust. Solder should look wet all the time even when cooled and hard. Those traces of copper on the circuit board are no different than the wires you add to them. They're just flat wires glued to fiberglass.

In the mangled words of Dynatec Chris, the Mini ZVS was too hot! I happened to agree.

 

[orten]

Is the coil soldered like in a normal zvs mini?
Did you tested the output in wattage? Which electrical supply are you using with this mod?

 

[HeavyEyeLids]

@TommyDee My inductor can only be split with 14 turns on each side. Should I stick with the 14 turns or should I take some wire from my second inductor and do two 15t winds with two of the ends being combined at the +12v input? Does this "Half pint" mean the driver will pull/require less power? I'm not sure how I want to power this. I want a 2s with 18350s but I've not seen too much discussion on how good of an experience is. What are the heat up times and how many heating cycles can a 2s 18350 supply? Also can I use a 2s board intended for 18650s on 18350s?

 

[TommyDee]

... I thing you named it... just give my poor brain a picture...please!

This is a picture I did for a Reddit post. This is the trace that sends power to the drive circuit.



The magnet wire in the image is part of the switch setup I was working with. I need to clean that up a little. That wasn't the purpose of the build. There are other hints in the images of what I was up to considering where this coil will sit on this particular build.

@orten - The circuit is identical with reduced values. I get a lower idle current by about 30% for the same work coil. Readings while heating the VC are running on average 5 watts lower. Realistically, you can limit this to about 80 watts but it'll push a 96 watt supply without issue.

@HeavyEyeLids - If you are thinking 2S for power, you might want to prototype this. I think the low end of 2S, 6V, is insufficient for this type of circuit. I'd love to be corrected in this regard. I know what is 'should' take but I have not done significant testing in this regard.
heatup times are highly coil-configuration dependent. I aim to maintain the 28-30 inches of original coil wire. Rather than fold turns out of the way, I turn them along the outer perimeter. This makes for a 5-8 second heatup time at full charge and it lags from there as voltage decreases. I run a 16mm ID coil ideal for 7740 5/8" glass. I run 9-turns on the ID to match the length I want to heat. I wrap a remaining 3 turns on the outside. This is a heater that will pull around 60-70 watts on a 3S li-ion pack settling in at around 10.5 volts at the module. if you lower losses, you might opt for LiPo. This really is user preference. However, the performance of the heater must be evaluated through the full voltage range of 12v-9v or 8v - 6v all the while trying to maintain a power deliver of around 50-80 watts when heating the VC.

 

[orten]

One last thing, where are conneted these two 18 gauge (i guess) magnetic wire? Can you make a larger photo?
Is that the wire that you have used for the coil and the other end is connected to the normal positive?

 

[TommyDee]

Sorry for the confusion. You are correct, those are 18 gauge magnet wires I had hooked up for a coil switch. I have since removed them.

What I did was to isolate the pads at the 3-pin connector. I moved the wires to the bottom;



Note that I moved the cut of the trace between the two horizontal wires.

I should add a word of caution here: the stippled pattern trace under the black solder mask is un-terminated. If the wire insulation or the black epoxy insulation on the board are breached, you will energize the plane. Solid wire is very forgiving in robust structures. Please be aware of the limitations exploiting these properties. There are often very small blessings keeping you from misfortune. Due diligence is the watchword.

Time to make a new frame again



Welcome to the forum @orten .


I do want to add here that this little Box Troll IH is a nice testing vehicle. Compared to the earlier full module, I am still sold on the HalfPint concept. I like the way it delivers and I like the way I can augment my experience with Sip&Dip [tm master piggy] to enhance the draw. A few seconds won't fry the bowl. I find the bake gentle on the terps while still delivering a righteous punch. More manageable heat in my most humble opinion.

 

[orten]

Thanks @TommyDee , tell me if I'm right "you moved the cut between the lower wires, are you using them as switch? If yes, is there a problem for the choice of positive/negative? Also, what's the effect of splitting the plate for the coil? Will you normally solder it the two opposite site?

 

[orten]

So the two connections (red and black) that I draw are for the power supply and on the two orange circles are for the coil, am I right?

 

[TommyDee]

Correct on the diagram and work coil connections. I am making islands for the switch connections.

 

[BirdFisher]

Wow, this is precious, thank you all for those fabulous builds, I want to try the half-pint myself, if only I can repair all the fried modules I have lying around they all seem to fail at some point. What is wrong with them? So OK, the first half dozen, I fried the FETs by underpowering the modules using 5A PSUs that I have a lot of - oh, well - but now I've been using this German Salcar 6A PSU, but my last two good builds just stopped powering yesterday morning..? How can I check the FETs (with a basic multimeter) / what to look for / replace with what to prevent those catastrophic failures?

 

[TommyDee]

FETs short closed. Measure across the Source-Drain junction. Those MOSFET switches? They are pin-compatible. As a HalfPint circuit, they hold up. If you can replace them, they work.

You need close to 8 amps for the full module. A regulated supply is the best way to go since you can control your wattage. Having a power source glitch the inputs of the IH is asking for trouble. For me, they've been pretty solid. The only blown unit I have I put a steel pipe into. Dumbass! But that is how I learned tip #1 above.

HalfPint can do small. Show us what you come up with. We need more builds on this thread.

 

[BirdFisher]

Wow, that was fast! Thanks a lot for taking time to respond to a quasi-noob ; I'm afraid I need a little more guidance, please bear with me :

> FETs short closed. Measure across the Source-Drain junction.
OK (I had to look up the drain & source) what value should I look for? I guess any positive value means that the FET is alive... Am I right? Oh, and can testing a FET while it's working damage it?

> Those MOSFET switches? They are pin-compatible.
Do you mean that I can use any of them, providing they say "MOSFET" and are in this form factor?

> As a HalfPint circuit, they hold up. If you can replace them, they work.
Well, to make the half-pint I first have to repair my "full pint" modules... As per this soldering guide video Pipes very kindly made, I think I can do it. I just need to know what to buy ; And where, but that's another story.

> You need close to 8 amps for the full module.
WHOA ! Do you mean... That even with my 6A PSU I'm still underpowering the module? I do have access to 8A (up to those cheap 14A cheap metal encased ones) PSU... So I should forget about the 6A? Wow. If this is true, thank you for telling me!!

 

[TommyDee]

Wow, that was fast! Thanks a lot for taking time to respond to a quasi-noob ; I'm afraid I need a little more guidance, please bear with me :

> FETs short closed. Measure across the Source-Drain junction.
OK (I had to look up the drain & source) what value should I look for? I guess any positive value means that the FET is alive... Am I right? Oh, and can testing a FET while it's working damage it?

I read open in one direction and something in the 900K range in the other. low ohms under 1K suggests issues.

> Those MOSFET switches? They are pin-compatible.
Do you mean that I can use any of them, providing they say "MOSFET" and are in this form factor?

LOL nono... The popular MOSFET switch offered on eBay for a dollar, that one. It has 2 FETs that are -pin compatible- meaning they fit and function similarly. They are in fact not equivalent. The original FETs are superior.

> As a HalfPint circuit, they hold up. If you can replace them, they work.
Well, to make the half-pint I first have to repair my "full pint" modules... As per this soldering guide video Pipes very kindly made, I think I can do it. I just need to know what to buy ; And where, but that's another story.

Replacing FETs is a challenge even for me. Not ripping up the board trying to remove them is tough enough. Then cleaning all the solder so they sit perfectly flat again so they heatsink against the copper is another story. But yes, if you manage to get through the hurdles, they can be used in a pinch. Certainly worth saving $10 over. But the question is, why are you killing modules? That is -not- the norm in my experience.

> You need close to 8 amps for the full module.
WHOA ! Do you mean... That even with my 6A PSU I'm still underpowering the module? I do have access to 8A (up to those cheap 14A cheap metal encased ones) PSU... So I should forget about the 6A? Wow. If this is true, thank you for telling me!!

It is a loaded question. If you detune the IH, 72 watts is more than enough [12v/6a]. But if the module wants 80 watts, then what? Will you even know it is taxing your power brick? Not without measuring. Therefore I have to recommend an 8 amp supply to be safe and meet the requirements of the power brick. You should never overload them. What's worse, as quality goes up on the bricks, they should actually shut down on an overload. That means your IH would be cutting out and you'd wonder what was wrong. Well, an 12v 8a supply keeps those questions at bay ;]

 

[MikeRotchHertz]

I'd like to start this thread off with inviting all the DIY IH efforts.

I invite developments as well as builds to participate here to share knowledge and understanding.

Myself, I am in 4 for 4 on home-built Mini ZVS 120W modules. Now that I have some magnet wire, efforts are under way to qualify coils.

My aim is to make a small handheld portable that utilizes a cradle to maintain charge.

...this below ain't it

But this is the entry I would like for this first page. This is the Mini ZVS 120 watt induction heater.
Improvements to this module include;
a) The positive power circuit and the control circuit have been separated by cutting the positive trace under the 330nf capacitors.
b) A simple momentary switch, or several in parallel, can be placed anywhere. This is a scavenged switch that is ShoeGoo'd into place. This is wired between the positive input lead and the positive output lead. If you maintain the blue connectors, the positive input lead is now a switch input for the gate circuit; the center pin of the output connector (3-pin) is positive power in put.
c) A 9-turn coil of the original ID and wound closer together raises the resting current about 25%. I need to see what 11-turn coils will change. This one is a little too hot at a 12.6 volt input. I also need to see what axial orientation does to resting current. Oddly enough, orientation is in play, unfortunately.
d) The traces on these boards are woefully undersized. There is one via each on my board to power the source lead of the FET. Therefore I scraped clean the trace between the two S-leads and soldered the 16 gauge negative lead directly to that trace. The via's will power the gate circuit as normal. Same with the positive lead. I wired the 16 gauge positive lead directly to the coil lead.

Result; 5-6 second very hot heat-up on a 19mm coil. It runs about 75 watts at 10.5 volts. I want to raise that 10.5v to 12.6v at this wattage. I'm going to try to do this with the coil alone.

On a cord and a little box, this is a cute little heater.

View attachment 310ioollo

Reference: IH Bits&Pieces thread

L

 

[BirdFisher]

(...) But the question is, why are you killing modules? That is -not- the norm in my experience.

Thank you very very much for trying to help ; I'd really love to know too, seriously. Here is my design, it's simple :

Special OT note : It sports a tuning system to set the height of the heated element, and lock it in this position, like you mentioned in this post If I could source non-magnetic, like aluminium bolts, I could even lose the wooden base.

And here are the PSU bricks I used so far :

At first I used the big grey 60W in the middle, because they're easily available around my neck of the woods. When it works it works (albeit a bit slowly, like 15+s for the 1st click and 10 for the subsequent) but sometimes the blue light flickers, and if you stop right there it's OK, but if you don't notice it and keep the button pressed, one of the FETs go up in smoke ; I killed like four or five modules like that, then I switched to the 6A on top, I used a unit with it (daily, fairly heavy usage) for the bigger part of a year, but one morning it simply refused to turn on I checked another unit, dead too and now I don't even have a visual clue as to what is wrong... Then I got the behemoth brick in the bottom, 8A, to power my last module, which is ironically the first I bough, like less that 3$ on Amazon ; When this one fails, I'm dead in the water. How can I protect it?

 

[TommyDee]

Aha, yes, you nailed it right on the head with the silver supply @BirdFisher . There is a situation that can perturb the oscillation and that can kill a FET. It doesn't take much. How a supply acts at current foldback has everything to do with it. Someone needs to update BeastHoss on this little known issue.

The 6 amp supply was like a faithful puppy. It gave all it could as long as it could but in the end, the internal heat aged it prematurely. Bury it with honors. Above and beyond and all that.

Never regulate current with an under-rated power supply.

Yes, 12V/8A should be sufficient. If you want to regulate that down, there is a nice 200W DC-DC converter that will give you new options, cheap.

I love the 'Cart' concept. That looks great. You can see if the metal screw is coupling or not by making a idle current measurement with and without the adjuster. Stainless steel will be a little better by reducing coupling current by some - aluminum screws is hobby store fodder or a tap and die set on hobby rod stock. Maybe a wooden head on the screw. I love seeing the adjustment implemented. Works a charm doesn't it

 

[BirdFisher]

The 6 amp supply was like a faithful puppy. It gave all it could as long as it could but in the end, the internal heat aged it prematurely. Bury it with honors. Above and beyond and all that.

Just to be sure, it's not the 6A supply that I have to lay to rest under a folded flag, but the (2) modules I killed with it, right?

Never regulate current with an under-rated power supply.

Yes, 12V/8A should be sufficient. If you want to regulate that down, there is a nice 200W DC-DC converter that will give you new options, cheap.

Wow, so even 8A should be just "sufficient"? Can you please post a link to this "nice 200W DC-DC converter" ? Also, would one of these work safely?

I love the 'Cart' concept. That looks great. You can see if the metal screw is coupling or not by making a idle current measurement with and without the adjuster. Stainless steel will be a little better by reducing coupling current by some - aluminum screws is hobby store fodder or a tap and die set on hobby rod stock. Maybe a wooden head on the screw. I love seeing the adjustment implemented. Works a charm doesn't it

Yikes, thanks! I'm really happy with it too ; Wood, nuts & bolts are easy enough to find around here ; I found a brand of food coloring that makes small glass bottles just the right size, and set up a system to cut them at the right length ; The tuning system took a bit of thinking : A wooden head on the screw, that's what it has right now if you look at the pic closely, so it's a bit of a unelegant (glued on) solution, I'd love to use just-the-screw, and will experiment once I get new modules. But first, making sure I stop killing them, it's really heart-breaking, let alone the bank

 

[TommyDee]

The modules died in battle. The 6A gave its all to the end. Less modules must die... -- NO MODULES MUST DIE!

Converter w/ 12V limit suggested - https://www.ebay.com/itm/XL4016-200...lator-Adjustable-Step-Down-Board/153661352000 similar ones with fixed potentiometer will work but will have less sensitivity on the knob/dial.

You are far enough away with the screw to be okay. If the glue bubbles, you know why Love the use of the glass jars. That's excellent.
There really is nothing like these artisan vapes. I thing you were plagued with a properly functioning power supply in the 60 watt supply. Those are rare in the wild.

Keep up the Great Builds @BirdFisher

 

[CaleidosCope]

Just had a "test" for the half pint... the mentioned finger, not released while stunning elsewhere... about 1 minute full heat. Nice colours on the cap. Ti-Tip roasted black. My half-pint LiPo setup passed through. What a mess. Polishing (Rotweiss, Polierpaste) the Ti-Tip was a game changer. Oh... and the cap works like normal (2020M-Cap). My silicone-plug on the end of the glass tube is a little melted. Just a little