Author Topic: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"  (Read 14619 times)

Offline petespaco

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +8/-0
    • View Profile
    • French Creek Valley Activities
Back on topic for the 12- 48 volt induction heaters, for a bit----
Here are two recent questions from one of my "pancake coil" youtube videos that I answered.  Hopefully I got the answers correct:

1st Recent question:
Hi, Is Coil Voltage also 48V or more than that?
If Yes, is there difference between 2500W, 1800W and 1000W?

Answer:
Yes, the voltage across the work coil is a lot higher than that of the power supply.
I just measured the voltage across a 12 turn 2 inch ID work coil on my 1000 watt heater with a multimeter and it is 220 volts rms.  I also looked at it with my old Tectronix 422 scope and then remembered why I don't pay much attention to the voltage across the work coil.  When I connected the 10X scope probe and turned on the power, I got a 300 volt peak to peak signal (at about 65kHz) that was approximately sinusoidal.  But, after about 30 seconds, the scope probe became too hot to touch, the insulation on the ground lead melted, and I almost burned it up.  The scope probe circuit's inductance (whatever it was) tried to act as a parallel work coil, but with only tiny conductors!!!
  I don't think these readings would change very much for the 1800 watt or 2500 watt heaters, but I have not measured them.  I really don't need to do that for my applications.
  For what it's worth, the experts say that the voltage across the Mosfets will equal
  the power supply input voltage X pi.

2nd Question:(In reply to my earlier answer)
I want to measure frequency with my multimeter it's up to 250V. I guess can't do with it.

Answer:
If your multimeter reads rms AC voltage, it should be okay.
 If you live in the USA, check the voltage at a  power outlet in your home.  If it measures approximately 120 volts, then your meter is reading rms.  If this is true (which is  quite normal for a multimeter) then you will be able to read the voltage across the work coil.
As always, use caution when measuring voltages this high.

(By the way, I find the posts about 220 volt AC input circuits interesting, but I that other readers realize the real dangers of working with power line voltages and currents.)

Pete Stanaitis
---------------

Offline rikkitikkitavi

  • High Voltage Enthusiast
  • *
  • Posts: 21
  • Karma: +1/-0
    • View Profile
Very few multimeters ca measure voltage in the 10 kHz range, due to bandwidth limitation. Not really necessary with Trms, as the resonance creates a fairly sinusoidal wave. Watch maximum input voltage too, typically they are still high impedance at a few ten kHz, but better check it.
Same goes for scope probes. Normally they catake full voltage rating up to a few hundred kHz

I wonder if your probe heating was due to the high magnetic field created eddy currents in your probe. In that case a multimeter probe, which is constructed of much thicker wires would go quickly?

And this all depends on how the field is, probe alignment etc.
A man can not have to many variacs

Offline Mads Barnkob

  • Administrator
  • High Voltage Expert
  • *****
  • Posts: 1118
  • Karma: +17/-0
  • Denmark
    • View Profile
    • Kaizer Power Electronics
The problem with the melting oscilloscope probe ground lead is that you are grounding your inverter through your oscilloscope! To measure on inverter outputs you always need a differential probe or an isolation transformer for the mains feeding the inverter and keep it floating off ground. I always recommend this great video:
/>
http://www.kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics

Offline Lane

  • High Voltage Enthusiast
  • *
  • Posts: 5
  • Karma: +0/-0
    • View Profile
Bert911: Thanks for the tip about the 3M coating material. I'm still not sure if a coating will contaminate my work in some way; but I'll keep looking into that, and will try something as I continue this process.

Pete: You made some great observations about the coil voltage; I think I had seen the 3 x Pi voltage reference somewhere before. I would have been scared to connect to my scope, and I'm really glad you did it for the sake of science (as others have pointed out). One question I still have; is it safe to touch the coil when running? I avoid it of course; but I've always been curious. You don't have to try that yourself of course...

Yes; working with 220V circuits is something to be careful with; but it's commonplace outside the US. Here in the US one has to be even more careful because it's tapping two legs. There doesn't appear to be anything strange about this board though (nothing extra is going to go hot when connected). There is no physical power switch, and the input section is isolated from the rest of the circuit. I'm planning to put it on a two pole breaker for a power switch as a safety measure. After inspecting the schematic I posted previously, and comparing to this 3,500W unit's board; I noticed only a few minor differences. The capacitors on the output are larger, some components have a higher current rating, and a few other minor things to note. All the board component labels are different; but otherwise the values and connections are almost all identical. At some point I'll try to relabel the schematic based on this circuit for reference purposes.

I had to pick up some 100k thermistor fuses to connect to the boards temperature sensor ports. I'm planning to run them outside the furnace insulation so they don't see anything close to internal temperatures. I'm hoping thats enough to keep the controller from shutting down. Being power adjustable; I'm planning to start testing at the lowest setting once I'm ready to juice it up. Just waiting for a meter capable of measuring coil inductance. I also don't know what the output voltage for the fans is on the board; if it's 18v I'm going to have to order some new fans for long term use.

I went through a few theoretical coil designs. Two 20' spools of 1/4" copper tubing would create almost exactly a 120µH coil  if connected end to end without any shaping. A 50' spool makes some other designs possible as well. I got a 1kg and 3kg graphite crucible to work with. Also a larger 6kg A shape silicon carbide crucible which I haven't tested; but read should work with induction heating. 

The 48v 2,500W unit showed up here the other day, but I haven't had a chance to wire and mount it.

Has there been any testing of the limits in terms of coil frequency with 48V 1,000W or 2,500W units? Should I grab a 1,000W board to sacrifice and push those limits? Since I'm already working with larger coils, it's not much trouble (and smaller coils are easy). Looked at running solid copper for the 120µH coil; but even with furnace insulation I wasn't sure it would stay cool enough. An 1,800C infrared pyrometer showed up; but I haven't been successful measuring metal temperatures in the small crucibles yet. I get almost no heat detected through the ceramic insulation. Not sure about the emissivity setting though (and my graphite is almost powder so it barely glows at the moment).

Offline petespaco

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +8/-0
    • View Profile
    • French Creek Valley Activities
Quote
One question I still have; is it safe to touch the coil when running? I avoid it of course; but I've always been curious. You don't have to try that yourself of course...

I have touched the coil many times, but I would not recommend it.  After all, were see a few hundred volts across it, right? But, I am not dead.
I do touch the coil with "one hand in my pocket" from time to time just to verify that the water is cooling the work coil properly.  Occasionally, I have touched both coil leads at the same time WITH THE SAME HAND and I do feel an uncomfortable "tingle".
  I still think it is a DUMB idea, particularly since I already do have a temperature meter measuring water temperature on my radiators.
  I would NEVER attempt that if I had a board that was powered by 220 volt mains, even though the output ckt is "isolated".

Thanks for the scope "heads up", Mads.
  I should have realized that. 

Pete Stanaitis
---------------

Offline petespaco

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +8/-0
    • View Profile
    • French Creek Valley Activities
Quote
Has there been any testing of the limits in terms of coil frequency with 48V 1,000W or 2,500W units?

Yes.
I have not found a lower frequency  limit yet, but lower seems to be better for the Mosfets.  Things still work well down to about 26 kHz.
It's the upper limit that I have spent most of my tine evaluating.

See my data and my opinions here:
https://spaco.org/Blacksmithing/ZVSInductionHeater/WorkCoilsForZVSInductionHeater.htm
--And click on the "Work coil, Observed" spreadsheet.

Pete Stanaitis
---------------


Offline rikkitikkitavi

  • High Voltage Enthusiast
  • *
  • Posts: 21
  • Karma: +1/-0
    • View Profile
The problem with the melting oscilloscope probe ground lead is that you are grounding your inverter through your oscilloscope! To measure on inverter outputs you always need a differential probe or an isolation transformer for the mains feeding the inverter and keep it floating off ground. I always recommend this great video:
/>

Exactly how is the inverter connected to mains PE/GND, which the scope is? If it is galvanic ally connected to mains it it could be lethal to touch the coil.
I have understood that the inverter I'd riven by a 48V supply, but that might be wrong.

The 48V power supply might have its 48V GND connected to the mains PE, but if it is an old server supply propably not, only the 12V are done so. The 48V are typically totally floating.

Apart from that, yes the video is extremely instructive. It is very easy to blow a scope when connecting uninsulated to the mains or without a doff probe. A normal transformer have quite a lot of capacitive coupling, something to be aware of when measuring fex high side switching nodes in half/full bridges even over isolation.

Safest is of course to put the GUT (Gadget Under Test) at isolation, if you isolate the scope averything, including scope GND is floating and can ruin your day...
Or use a diff probe. In the end they might cost about the same.
« Last Edit: August 23, 2019, 05:06:19 PM by rikkitikkitavi »
A man can not have to many variacs

Offline petespaco

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +8/-0
    • View Profile
    • French Creek Valley Activities
Just to belabor the point about scoping the coil voltage:
I now remember that I was using my little DSO-112 battery powered scope early on when I had been taking readings.

Pete Stanaitis
---------------

Offline petespaco

  • High Voltage Technician
  • ***
  • Posts: 122
  • Karma: +8/-0
    • View Profile
    • French Creek Valley Activities
Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
« Reply #228 on: September 16, 2019, 01:50:19 AM »
The making of induction heater work coils the way I do it and why-
  That's the focus of my newest ZVS 1000 to 2500 watt  12 to 48 volt induction heater video.
It is here:
/>
I'd be glad to entertain any comments, questions or criticisms.
  As they say somewhere:  "There's more than one way to skin a cat".
It's just the anyone who doesn't do it my way is wrong. <G>

Pete Stanaitis
---------------

Offline davekni

  • High Voltage Enthusiast
  • *
  • Posts: 11
  • Karma: +0/-0
  • Physicist, engineer (electronic), and hobbiest
    • View Profile
Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
« Reply #229 on: September 16, 2019, 04:18:26 AM »
The LM2596HV chip is almost certainly counterfeit from a Chinese source.  I'd purchased some similar little switching regulator boards, and the "LM2596HV" parts on them are definitely not legitimate.  Haven't tested voltage, but their regulation loop becomes unstable at higher current (by 2A at least) and the switching frequency is not within LM2596HV spec. range (lower).  Many FETs I've bought on EBay are also counterfeit, especially Alpha-Omega Semiconductor parts.  Sometimes they meet DC parameters, but are high on capacitance.  I'm now using normal electronics distributors for most parts.

If anyone is interested in building their own Royer oscillators, I've designed a couple improved (in my opinion) circuits.  In particular, these avoid the trade-off of power dissipation of the gate-drive resistors vs. gate drive strength that exists in the versions being discussed here.

My designs still share the issue of drawing a spike of current during power-up.  That shows up in Spice simulation and in real life.  A beefy slow supply isn't always a good solution either.  The power current peak builds up in the inductors feeding the resonant circuit.  When oscillation does start, that current momentarily makes the oscillation voltage much higher than during normal operation, which can fry the FETs.  I've added TVS diodes to most of my builds, although they aren't shown in schematics below.

My favorite circuit, shown here in a version intended for running off full-wave-rectified 120VAC (wall-plug power here in USA)"
* royer5a.pdf

And here is a very slight tweak to the above.  Power-coupling inductors L1 and L2 become a center-tapped transformer, and L4 is added as the power source inductor.  Power-line common-mode-chokes work well for the L1/L2 transformer.  The new L4 now has twice the frequency, so can be smaller.  This is my preferred implementation.  (Except for situations where the working coil can be center-tapped, in which case L1 and L2 can be eliminated.):
* royer5b.pdf

Finally, here is my previous Royer oscillator design - a smaller tweak to the circuits in the commercial units.  It just adds two PFETs to disable the gate-drive resistors once the gate is pulled high, allowing lower-ohm resistors (for faster gate rise-times) without excess power dissipation.  This version happens to be for 19V (laptop power supply).  It works at higher voltage too.  A friend in Germany built one for rectified 220V using 1200V SiC FETs.
* royer4a.pdf

I've played a little with induction heating, but used these more for driving home-wound inverter transformers to get a few kV (8kV peak for the one I'm playing with at the moment for DIY plasma balls - using the circuit on 19.5V from a Dell laptop computer supply.



David Knierim

Offline T3sl4co1l

  • High Voltage Experimenter
  • **
  • Posts: 70
  • Karma: +8/-0
    • View Profile
Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
« Reply #230 on: September 16, 2019, 06:59:12 AM »
Good stuff, Pete.

Pro tip, literal: the way we did coils with interesting geometry, was to use square or rectangular tubing.  This can bend very sharply without kinking or tearing, and zero-radius corners can be mitered (cut out a wedge in three sides, leaving a thin web along the back; solder or preferably braze the seam closed).

The hard part is finding the stuff.  McMaster does sell it, in short lengths (and McMaster typical prices).  Never saw where we got the stuff from, but it was by 25'+ rolls, probably not cheap either way.

Swaged fittings, and using ~telescopic sizes for couplings and adapters, is an excellent method, of course.

I highly recommend brazed joints.  These are super easy in copper: you can use a self-fluxing (phosphorus-bearing) rod like Harris Dynaflow, no paste flux needed.  A plumbing torch will do for small tubing like this, or a somewhat bigger torch for bigger stuff of course.  The filler is considerably stronger than the base metal, no worries about strength; actually if anything, that's a liability as the joints are inflexible and break rather than bend -- plan your joints accordingly.

Tim

High Voltage Forum

Re: Help for people buying the "12-48 Volt 1800/2500 Watt ZVS induction Heater"
« Reply #230 on: September 16, 2019, 06:59:12 AM »

 


* Recent Topics and Posts

post Re: Problems with my first DRSSTC
[Dual Resonant Solid State Tesla coils]
Phoenix
Today at 07:37:55 AM
post Re: Large Ferrite Core with IGBT-Brick Driver
[Transformer (ferrite core)]
Phoenix
Today at 06:43:39 AM
post Re: Problems with my first DRSSTC
[Dual Resonant Solid State Tesla coils]
davekni
Today at 01:45:37 AM
post Re: Breakout rod material and size vs streamer resuls
[Dual Resonant Solid State Tesla coils]
davekni
Today at 12:18:00 AM
post Re: My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
davekni
September 22, 2019, 10:16:26 PM
post Re: My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
AstRii
September 22, 2019, 05:40:52 PM
post Re: My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
davekni
September 22, 2019, 06:43:46 AM
post Re: My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
AstRii
September 22, 2019, 12:36:23 AM
post Re: My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
Hydron
September 22, 2019, 12:03:43 AM
post My DRSSTC does not want to resonate at correct frequency
[Dual Resonant Solid State Tesla coils]
AstRii
September 21, 2019, 09:01:48 PM
post Re: The new toy
[Laboratories ]
DashApple
September 21, 2019, 05:54:03 PM
post Re: The new toy
[Laboratories ]
MRMILSTAR
September 21, 2019, 05:38:14 PM
post Re: The new toy
[Laboratories ]
DashApple
September 21, 2019, 04:55:44 PM
post Re: The new toy
[Laboratories ]
MRMILSTAR
September 21, 2019, 04:03:48 PM
post Re: The new toy
[Laboratories ]
T3sl4co1l
September 21, 2019, 03:57:22 PM
post Re: The new toy
[Laboratories ]
Hydron
September 21, 2019, 09:15:47 AM
post Re: The new toy
[Laboratories ]
Hydron
September 21, 2019, 09:11:18 AM
post Re: The new toy
[Laboratories ]
AlexanderHun
September 21, 2019, 09:08:50 AM
post Re: The new toy
[Laboratories ]
AlexanderHun
September 21, 2019, 08:59:41 AM
post Re: Blast from the past!
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Hydron
September 20, 2019, 11:32:35 PM
post Re: Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Hydron
September 20, 2019, 11:28:05 PM
post Re: The new toy
[Laboratories ]
Hydron
September 20, 2019, 11:20:21 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
Hydron
September 20, 2019, 11:10:54 PM
post Re: The new toy
[Laboratories ]
AlexanderHun
September 20, 2019, 09:49:00 PM
post Re: The new toy
[Laboratories ]
DashApple
September 20, 2019, 08:17:36 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
September 20, 2019, 08:06:32 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 20, 2019, 05:43:13 PM
post Re: Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Twospoons
September 20, 2019, 01:32:30 PM
post Re: Adventures in induction heating
[Electronic circuits]
Coyote
September 20, 2019, 10:56:19 AM
post Re: Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
kamelryttarn
September 20, 2019, 10:02:28 AM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
davekni
September 20, 2019, 07:05:19 AM
post Re: Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Hydron
September 19, 2019, 11:16:21 PM
post Re: The new toy
[Laboratories ]
AlexanderHun
September 19, 2019, 09:49:30 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
September 19, 2019, 09:49:15 PM
post Re: The new toy
[Laboratories ]
Mads Barnkob
September 19, 2019, 09:13:29 PM
post Re: Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Mads Barnkob
September 19, 2019, 09:08:20 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 19, 2019, 07:16:20 PM
post Help choosing a Hantek USB oscilloscope
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
kamelryttarn
September 19, 2019, 04:19:58 PM
post The new toy
[Laboratories ]
AlexanderHun
September 19, 2019, 09:01:46 AM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 18, 2019, 07:27:46 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 18, 2019, 06:56:20 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 18, 2019, 05:08:06 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
Mads Barnkob
September 18, 2019, 01:35:58 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 18, 2019, 05:54:23 AM
post Re: Blast from the past!
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
klugesmith
September 18, 2019, 05:05:55 AM
post Re: UD2.7 Gate Signals
[Dual Resonant Solid State Tesla coils]
coilinator
September 17, 2019, 08:24:59 PM
post Re: Blast from the past!
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
MRMILSTAR
September 17, 2019, 08:19:45 PM
post Re: Blast from the past!
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
Mads Barnkob
September 17, 2019, 06:53:52 PM
post Blast from the past!
[Computers, Microcontrollers, Programmable Logic, Interfaces and Displays]
MRMILSTAR
September 17, 2019, 05:52:00 PM
post Re: CM300 coil build questions
[Dual Resonant Solid State Tesla coils]
acobaugh
September 17, 2019, 05:00:47 PM

Sitemap 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30