1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

New Induction Heater circuit with no center tap.

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by gary350, Nov 1, 2016.

  1. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    First picture is the new circuit that I found online, I don't see how this can work?

    Second picture is the circuit that works 35 amps, 735 watts.

    I want to test this new circuit with no center tap on the 8T coil to see how it works or IF it works, compared to the circuit with a CT. It seems to me when each mosfet comes ON there are 2 ways for current to flow, through the 8T coil and around the 8T coil.

    The amp meter shows 35 amps through the TESTED circuit with a 3/8" steel rod for a load it takes 7 seconds to turn red hot. I built a separate 10 vdc gate power supply. I have built new .1uh chokes. With no center tap the circuit will osc through all 8 turns of the 8T coil instead of 4 turns of the coil. Mosfets in both circuits are P55NF06L.

    I want to do away with the CT so the coil can be wound with copper tubing and be water cooled.

    [​IMG]

    [​IMG]

    [​IMG]

    [​IMG]
     
    Last edited: Nov 1, 2016
  2. chemelec

    chemelec Well-Known Member

    Joined:
    Apr 4, 2004
    Messages:
    2,123
    Likes:
    77
    Location:
    Hedley, B.C. Canada
    What Gauge of wire are you using for the Secondary on that Microwave Transformer?
    Looks Small for that high of Current.

    And are you getting the Same amount of Heating in the Same Time with Both of these circuits?
     
  3. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA

    The secondary winding on the MOT is #4 stranded copper. The power supply is built with #12 copper. This is a test circuit it never runs longer than 45 seconds each time.

    The old choke coil is #16 copper.

    The new choke coils are #12 copper.
     
    Last edited: Nov 1, 2016
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. Tony Stewart

    Tony Stewart Well-Known Member Most Helpful Member

    Joined:
    Aug 31, 2012
    Messages:
    3,152
    Likes:
    281
    Location:
    Richmond Hill , ON Canada near Toronto

    The most efficient , depends on the losses in each component, including;

    yours

    0.1H DCR=?
    MOSFET 38 mΩ x2
    0.47uF ESR =? /8
    ESR series choke
    ZL(f) of lead wires or ESL due to aspect ratio L/thickness ratio * mm length
    SRF/f ratio = ? depends on winding spacing and reduces current and power

    vs

    0.1uH DCR=? x2
    MOSFET=18mΩ x1
    0.47uF ESR=? /4
    Zl(f) of lead wires
    SRF/f ratio = ?

    Just my thoughts.
    I wonder about ripple current/(rated ripple current ) or temp rise of each cap, which depends on ESR*I^2 rms and thermal insulation of cap

    Temp rise in any part is due to DCR or ESR is an indicator of losses, where caps shud be <<85'C

    I think you can use water cooled 1/8" plumbing with large ferrite torroids choke to water/glycol circulation side and direct RF feed to coil, and still retain CT feed. Ferrite selection is critical, high Mu large mass with several in series of increasing size to prevent saturation, just like caps of different sizes to prevent high ESR yet low Zc(f)
     
    Last edited: Nov 2, 2016
  6. Tony Stewart

    Tony Stewart Well-Known Member Most Helpful Member

    Joined:
    Aug 31, 2012
    Messages:
    3,152
    Likes:
    281
    Location:
    Richmond Hill , ON Canada near Toronto
    also important is heat rise in diode clamps with ESR/Lcoil * f ? for duty cycle of diode pulse and junction capacitance rises with lower ESR, making active Q clamp replacements instead of diodes perhaps more efficient due to ESR*pF time constant and effect on RF current losses.

    Also scoping transformer coil current with a 50mV shunt and coaxial feed with 50ohm term at scope will indicate transformer saturation effects and need for series current limiter R, that you have in photo. This is where multiple L in series helps if they stay in linear region, just like multiple caps in parallel reduces Z and different values i.e. 100uF//o.47uF xn. Beware of RF gain of transition current spike at series resonance.

    All this needs is good RF test/measurement coax 50Ω methods to isolate all loss currents with current probes or differential A-B coax 50R measurement methods or use a 75Ω RF splitter with F connectors to scope and Term to measure forward and reverse losses as a directional coupler across current shunt. There are several ways to instrument this to determine all losses and mismatched impedances in the network. But it gets complex....

    This RF induction coil is the basic heater of all Perkin Elmer Mass Spectrometers where a sample material is mixed in Argon in a vacuum tube and accelerated and heated by a 100W 8000'K RF induction coil and the material in gas burns up in a flame then the optical spectrum of the gas molecules determines the mass elements and ratios of the material sample. Except these instruments cost beaucoup bucks.
     
    Last edited: Nov 2, 2016
  7. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    I just dont see how this circuit can work with no center tap on the 8T coil current takes the path of less resistance most of the current will go through the .1h choke before the 8T RF coil.
     
  8. Tony Stewart

    Tony Stewart Well-Known Member Most Helpful Member

    Joined:
    Aug 31, 2012
    Messages:
    3,152
    Likes:
    281
    Location:
    Richmond Hill , ON Canada near Toronto
    Both ends are inductive biased to V+ instead of centre tap , so conduction path is commutated from ends to ground alternately... with twice the induction , but twice the source loss.

    Energy stored in half cycle is also released by diode clamp in the Q off state, so some circulation boost helps in both depending 1/LI^2t stored energy. with t controlled by ESR/L from diode Pd rating so I think it's ends up being lost in ESR*I^2 with some BEMF induction heat.
     
    Last edited: Nov 2, 2016
  9. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    Still cant bring yourself to setting yourself up with full H bridge circuits made from high capacity switching devices, like 600 volt multi hundred amp IGBT or Mosfet power brick units, driven by dedicated driver IC's (IR2110's) run from purpose built FM/PWM driver circuits (LM3525's) that will give you full range frequency and switching duty cycle control plus won't burn down or blow up the first time you load them up, huh? :rolleyes:
     
    • Agree Agree x 1
  10. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    I bought some IGBTs July 2015 but not ready to use them yet. Today I experimented with winding choke coils and heat sinks. It is interesting when I put a magnet on a choke coil the value drops. Since I am going to be running a high current I am winding chokes with #10 copper wire to learn what the values will be with the number of turns I am able to put on each 1 toroid then 2 toroids then 3 toroids etc. A 3 ounce aluminum heat sink pulling 35 amps on the mosfet heats up from 70 degrees to 130 in 2 minutes with no fan. I can do the math and calculate BTUs. A few more experiments I can do the math to estimate about what I need for a good fan cooled heat sink. Also today I built an 8T water cooled coil and solder on the RC circuit capacitors. Soon as I get this coil connected to my existing circuit I can do some water cooling experiments. DATA is what i need to build a good usable circuit. It has been 18 months since I worked on this I hope it wont be another 18 months before I get around to it again. Wife and I have been going camping every Tuesday but doctor appointments, paying bills, vehicle repair, and many other things kept us home this week. 88 Degrees here today again Wow summer just keep hanging on but a front is moving in and temperatures are going to drop. It will be fun camping in the snow soon. I still need larger heat sinks and fans for the IGBTs. Before I retired I worked in engineering for 40 years we made sure everything was 100% right before spending a lot of time and money on something that won't work. A hobby is fun until it turns to work.
     
    Last edited: Nov 2, 2016
  11. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    So that's a very long 'No' then? o_O

    but......... That's not what your threads show. Ever.:confused:

    Just saying ~$20 in parts and one day and you could have an adjustable wide frequency band full duty cycle range driver unit that would make your poke and guess and burn things down and start over induction heater design process worlds simpler and easier to work with. :rolleyes:
     
  12. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA

    That would be great but I have no circuit drawing for my IBGTs and I don't know how much cooling they need and don't know yet the best way to build a water cooled coil. I have seen pictures of other peoples circuit that covers the whole top of the work bench I want mine built into a small portable case like a portable welder. Once it is built I don't want to be making changes because cooling is too small or something else is not right. I don't need a 100% duty cycle but I do want to be able to use it 30 minutes each time to heat 3" steel pipe so I can put a 5 degree flare on the ends.
     
  13. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    It's a shame no one ever invented a central communications system with endless databases that anyone could look things up on just by typing in some words, like a devices part number, or something in order to find its exact manufactures specs and recommended base circuit diagrams plus any other technical info as well. :rolleyes:

    Not to be rude but, if I
    I'd be GD embarrassed about the excuses you give
    for doing things the way you do and have done for how many years now on this site relating to these induction heaters of yours?

    You say you have the IGBT's so obviously they have a part number you can look up their data sheets with to which those said data sheets will give you the exact numbers you need to calculate the worst case thermal dissipation values you can have given your input voltage, amperage and switching frequency to which from there figuring out how big of heatsink you need can be done. :rolleyes:

    You've built a big ass Telsa coil at one time that worked (if I am getting things right) so obviously as a past engineer you know how to do basic technical math and design so finding the numbers you need and the rest of the data for any other related items and coming up with basic designs that exceed those minimum working requirement values should be like breathing to you. :facepalm:

    So why isn't it? o_O
     
    • Like Like x 2
  14. chemelec

    chemelec Well-Known Member

    Joined:
    Apr 4, 2004
    Messages:
    2,123
    Likes:
    77
    Location:
    Hedley, B.C. Canada
  15. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    Just for a heads up..........
    .
    I'm not trying to be mean or rude but my opinion is you will never make a homemade induction heater of the design you are playing with that can run anywhere near that power level in your remaining life time and here's why.

    I did the rough calculations on heating a 3" ID .25" wall mild steel tube three inches long to rough red hot (1200 - 1300 F) in under a minute with estimated losses and came up with a theoretical induction heating wattage in the range of ~20,000 watts minimum.

    The only rational way I see being remotely doable is you are going to have to step up to running off of a at minimum 240 VAC 100 amp isolated circuit rectified and filtered to make a stable 340 VDC high current DC power source that feeds a purpose built high capacity H bridge (1200 volt 600 amp IGBT's would be the bare minimum size switching devices I would even try to use) that in itself either drives a high frequency high power impedance matching transformer system (so you can tune it to whatever heating coil you want) or a very specifically built and tuned LC tank heater as a absolute bare minimum design.

    So if heating a 3" pipe with induction heating methods is your goal maybe it's time you either poop or get off the pot as in do some studying to see what you really need to get to where you are wanting to be or give it up and stick with heating nails and small bolts with what you have because what you have now will never do what you want no matter how much you fiddle with it. The necessary power level and how you generate them are way beyond this designs capabilities.

    That said, given everything I have seen you posts over the last few years regarding your induction heater experiment and designs if you really need to heat a 3" pipe for flaring work you would be decades and thousands of dollars ahead to just go buy yourself a common oxy propane torch system and a mid sized big bertha heating head to do the job. :(

    BTW here what a commercial built unit hat can do that power level looks like.
    Picture 040.jpg
    Picture 036.jpg

    Picture 037.jpg

    Picture 041.jpg

    Picture 035.jpg

    It's roughly 3 feet square 6 feet tall and weighs 1000+ pounds not including the coil cooling pumps and systems and takes a 77 KVA 480 VAC input to produce 50 KW peak and 30 KW continuous heating output. :eek:

    If you look closely at the picture of the bottom of the rig that large tan block with the brown HV insulators and 4 red tube on top is just the tuning capacitor for the H bridge LC tank primary circuit and just to the right and above it is the huge liquid cooled ferrite core impedance matching transformer that drives the output. :wideyed:

    If you seriously want to do what you say you are you will need to build something about half this capacity at minimum which will make it about as small and portable as a standard commercial welding power supply. (not the suitcase style either. The big steel box on wheels type) :oops:
     
    Last edited: Nov 3, 2016
    • Like Like x 1
  16. Dr_Doggy

    Dr_Doggy Well-Known Member

    Joined:
    Aug 11, 2007
    Messages:
    1,718
    Likes:
    37
    I like the zvs circuit very much , almost the equivalent to a flasher circuit:
    http://wild-bohemian.com/electronics/images/t-flash2.gif

    my first one was small with 5 amp at 12v, my second one was rated for just under a 15 amp 120v breaker, but still needs some final assembly

    It is very interesting to watch the magnetic responses from metals that are commonly known as non magnetic, such as copper and aluminum.....
    ... with specific coil designs they are even defying gravity (notice the loop at the bottom, his coil is wound CW for the first 2-3 turns then flips and is wound CCW for the rest, almost like magnetic tweezers)!.... also notice the power levels required and time taken:

    http://inductionheatertutorial.com/inductionheater/inductionlevitation.html


     
    • Informative Informative x 1
  17. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    Here are my IBGTs part number CM600HA-24H rated 1200V 600A

    [​IMG]
     
  18. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    So what is the wiring issue with those? o_O

    As long as you don't over voltage the gate and blow it out those things are darn near indestructible at normal household voltage and power levels.
     
  19. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    There are no good circuits for two IGBTs with no center tap.

    [​IMG]
     
  20. tcmtech

    tcmtech Well-Known Member Most Helpful Member

    Joined:
    Feb 22, 2009
    Messages:
    6,691
    Likes:
    431
    Location:
    Sawyer, North Dakota, USA
    Cripes. Seriously?
    Are basic internet searches for real data and information on the operation and application of very devices you are using beyond your grasp entirely? :facepalm:

    Did you even read a single thing that I said in the earlier posts about the power levels you will need to do what you want to do either? :arghh:

    The power levels at the supply voltages you will need are way beyond what you are ever going to be able to produce and control with a simple free running push pull circuit like you are fixated on using. :banghead:
     
    Last edited: Nov 5, 2016
  21. gary350

    gary350 Well-Known Member

    Joined:
    Dec 28, 2005
    Messages:
    1,097
    Likes:
    53
    Location:
    Murfreesboro, TN 37129 USA
    This circuit is center tap too. 65V needs to be changed to 170VDC.

    [​IMG]
     

Share This Page