1GBT Induction Heater Circuit Design help needed.

[Mosaic]

Both are correct, if u look at the ouputs both the collector or emitter of the transistors can be used, either common emitter or emitter follower. Look for the TL494 app note for more explanation

 

[Les Jones]

Here is a link to a data sheet on the TL494 that does give the formula to calculate the frequency. **broken link removed**
The formula is near the bottom of page 14.
Looking at the schematic on entry 18 before looking for a data sheet, from the component values and frequency it looked like the components connected to pins 5 and 6 set the frequency.

 

[alec_t]

In your post #18 schematic you show one gate driven via 2.2k (2.2R?) and the other gate driven via 1uF. Surely that's not correct?

 

[Les Jones]

Gary, It looks like you have just added part of the " instructables " driver circuit to your original circuit. In the instructables circuit the 1 uF cap and the 2.2R resistor are effectively in series driving the primary of a transformer. I do not believe you can just connect the two part circuits together the way you have. Also if my understanding of the way the original circuit works the two schottky diodes provide feedback to make the circuit oscillate. If this is so why do you need to retain them in your new design ?

 

[Mosaic]

Gary, as you have difficulty understanding the spec sheets and circuit function, you should not join bits and pieces of other designs because the number of connections match.
Use the internet, choose a pilot project, with less dangerous power levels.
1) Execute the project;
2) Obtain functionality;
3) Understand the basic function and the component working conditions;
4) Intelligently tweak things slowly and note the results;
5) This enhances your understanding of circuit function and how the components will respond;
6) To meet your established goals; start doing 'upgrades' of the high power section calculating for power dissipation and loading in EACH component before replacing anything. It's a balancing act among function, power handling, cost, size, operating conditions, efficiency, reliability, stability. This list can grow if you want to institute ROHS materials, EMC, BOM reduction et al. Don't rush this.

Gary, as you can see, reading and understanding a spec. sheet or app note is key to a lot of these parameters. If the language and diagrams and charts used are difficult for you it means you have some fundamentals you need to digest first.
You must be familiar with theorems, units and formulae based on names like Ohm, Faraday, Maxwell, Thevenin, Kirchoff, Watt, Joule, Coulomb, Ampere, Volt, Henry, Farad,.........etc.
Thermodynamics basics for assessing forced air or water cooling is also useful.
Then you must be familiar with the materials and methods components are made with in order to spec them. eg. Silicon, gallium arsenide, SiC, carbon, carbon film, ferrite, iron, mylar, ceramic; epitaxial, trench, trench/epitaxy, wire wound, sintered, laminated,.. this list is sizeable......

For instance, you can purchase a silicon bipolar transistor with similar specs in terms of main specs such as Vce, Vbe and such. BUT one may be made for audio work (epitaxial) and one is a good switcher (BISS). You must be able to determine this from the spec. sheet, it is critical. If you have good fundamentals you can establish/change operating conditions to make viable part substitutions being aware of their specs.

Now there are tools to help you. For quicker development use a simulator such as LTSPICE (free) or PROTEUS ISIS or Multisim blue (free from Mouser). In a sim. you can 'try' out ideas and easily 'measure' outcomes and then determine if your parts can handle the loads placed on them. A LOT of the calculations are done for you and results are displayed graphically. Sims aren't perfect but they can automate lot of the time consuming design work.

In my experience about 50% of the time you can go from a sim. schematic to a working prototype with no tweaking. Usually the tweaking comes from having taken short cuts like using ideal components (inductors) rather than real world models. If you take the time to acquire/make the real world models your sim is quite realistic.

Take it one step at a time and sidestep the frustration dude. You have bitten off too much IMHO.

 

[Tony Stewart]

Since IGBT's have an effective Darlington output stage, the Vce threshold is ~1.5V and rises with Rce* Ic, which in this case is a rise from 1.5 to 2.1=Vce @ 240A or an Rce of 0.34mΩ , But the additional Vce threshold of 1.5 makes IGBT's an inefficient choice for low Vce and low Ic vs. MOSFETs at the same cost. Because if you average 240A that drop adds an effective 1.5V/240A or 6.3 mΩ to the total Rce. So it only makes sense to us these a thousands of Amps.

Take for example;
Manufacturer Part Number IXFN360N15T2
Drain to Source Voltage (Vdss) 150V
Current - Continuous Drain (Id) @ 25°C 310A
Rds On (Max) @ Id, Vgs 4 mOhm @ 60A, 10V
Vgs(th) (Max) @ Id 5V @ 8mA
Gate Charge (Qg) @ Vgs 715nC @ 10V
Input Capacitance (Ciss) @ Vds 47500pF @ 25V
Power - Max 1070W




It also has less Ciss since RdsOn is higher yet than your Rce of 0.34mΩ But overall 4mΩ is less than Req.=6.3mΩ@240A
Thus 2/3 rd the temperature rise and only $34 ea. in stock (10pc) cheap than EBay surplus IGBT's

This is just FYI for future designs.

 

[Tony Stewart]

Ω'S Law and Q factor at resonance.

BTW you are getting closer to a solution. LC impedance is ~0.7Ω @40kHz ( from mental calculations)

If ESR of your wire, Caps and IGBT adds up to 1 mΩ ( by careful choices) and LC impedance is 700 mΩ , you get a Q of 700, which would amplify the magnetic field x 7 with no load, so be careful and plan on sensing no load in your design to prevent over voltage and high B fields which might erase credit cards.

 

[tcmtech]

the additional Vce threshold of 1.5 makes IGBT's an inefficient choice for low Vce and low Ic vs. MOSFETs at the same cost.

Sort of true but when doing prototyping their ability to take severe abuse more than makes up for a bit of extra loss in lower voltage and current applications.

Personally I use them all the time for any circuit that is going to be working in an abusive electrical applications at more than a few 10's of volts and amps average power levels. To be honest I have found in a number of refit applications switching out power mosfets for IGBT's actually resulted in significantly lower heatsink temps. The forward drop voltage may have been a bit higher but the lower switching losses and robustness more than made up for it in power savings.

 

[Tony Stewart]

Sort of true but when doing prototyping their ability to take severe abuse more than makes up for a bit of extra loss in lower voltage and current applications.

Personally I use them all the time for any circuit that is going to be working in an abusive electrical applications at more than a few 10's of volts and amps average power levels. To be honest I have found in a number of refit applications switching out power mosfets for IGBT's actually resulted in significantly lower heatsink temps. The forward drop voltage may have been a bit higher but the lower switching losses and robustness more than made up for it in power savings.

I agree the IGBT's have much lower ESR than MOSFET in simlar range at slight cost penalty.
In this case 1/10 the RdsOn means if you have any type of non-linear load, the power dissipation wont rise rapidly like in MOSFETs which are a bit cheaper.

 

[DerStrom8]

**broken link removed**

Yep, that's an IGBT. What you see in the diagram is the symbol for an IGBT but it's equivalent circuit is a BJT with the base connected to a MOSFET.

 

[Mosaic]

Perhaps the OP could take a look at this:
https://www.eevblog.com/2015/05/27/...ed:+ElectronicsEngineeringVideoBlog+(EEVblog)

 

[gary350]

Do you know the difference between a Specialist and a Generalist?

A Specialist learns more and more about less and less until he knows everything about nothing.

A Generalist learns less and less about more and more until he knows nothing about everything.

When I graduated from Electronics 1970 most of these parts we have today were not even invented yet. This is a re-education process for me to learn how these new parts work. I worked 2 years as an electronic tech repairing circuit boards. Being a parts changer is a lot different than being a design engineer. Its not an easter egg hunt, find the bad parts and replace them. Not many people know how to find the bad parts. Not many people know how to find substitutes for poor designs and unengineered parts. That 1970s stuff is all stone age technology now. I was making $3.50 per hour, then I was offered a job in Research and Development for a different company starting pay $1 per hour. LOL. NO WAY, I am not taking a 70% pay reduction. I went into industrial electronics for 40 years. I designed and built my own controls and control panels and learned to program several different industrial programmable controlers and we built custom specialize equipment. I been out of hobby electronic for 40 years. I know the basics I just dont know these new parts. There is an old saying, if you dont use it you loose it. Hobby electronics should be fun. Often it makes better sense to buy your own TV don't build your own, buy your own volt meter don't build your own, buy your own computer Mother Board don't build your own. Learn to think outside the box.

 

[Mosaic]

gary350 Your area of risk is that you want to use industrial scale power via a hobbyist approach with a high chance of hobbyist mistakes causing industrial consequences.
I suggest that you keep the power levels at hobby scale until you have a comfortable grasp of the components, their operating conditions and safety factors. Then scale up.

A little knowledge can be a dangerous thing if it tempts you to take unnecessary risks.

 

[DerStrom8]

gary350 Your area of risk is that you want to use industrial scale power via a hobbyist approach with a high chance of hobbyist mistakes causing industrial consequences.
I suggest that you keep the power levels at hobby scale until you have a comfortable grasp of the components, their operating conditions and safety factors. Then scale up.

A little knowledge can be a dangerous thing if it tempts you to take unnecessary risks.

Mosaic, I get the sense that you're underestimating Gary's level of experience. He's not a newbie to electronics--He understands the basic concepts of electricity (ohms, volts, amps, watts, etc) and how they relate, as well as more complicated ideas. He's only trying to learn about more modern components (IGBTs, for example, were developed in the '80s but didn't become common until the '90s, at which time Gary was working with other types of electronics/repair).

I know where he's coming from--Until the beginning of last year I had no idea what an IGBT was either, and only learned because I was building a solid state Tesla coil that required high-power switching devices.

I don't know if Gary feels this way, but it seems to me like you're patronizing him, which is entirely unnecessary.

 

[gary350]

I won't be building this project and experementing at full power. I am going to use a 1100 watt microwave oven transformer re-wired for 65 volts same at the Lincoln welder. After I get the circuit working I can do upgrades. After testing it may turn out a 1400 watt MOT will do the job and I never need the welder. First test with the welder will be 65 volts 30 amps.

DerStorm8 how did that IGBT work on your Tesla Coil? 15 years ago I learned a Variable Speed Vacuum Fan RQ Spark Gap can be tuned to the correct frequency to get maximum output. Turn the dial and watch the 12 foot sparks get longer and shorter, dial it right to maximum output. Richard Hall said, a rotor is just a fancy toy compared to this. Few people on the TC forum ever took it serious. I have always thought if the spark gap Hz could be tuned to the same Hz as the TC that would be maximum output. I was getting 12 foot arcs from 10kw. I sold all 7 of my TCs, sold the big 10" last summer to the Local University. My 10" PVC pipe actually measured 11.25" diameter with 950 turns of wire 5:1 ratio.

 

[DerStrom8]

DerStorm8 how did that IGBT work on your Tesla Coil? 15 years ago I learned a Variable Speed Vacuum Fan RQ Spark Gap can be tuned to the correct frequency to get maximum output. Turn the dial and watch the 12 foot sparks get longer and shorter, dial it right to maximum output. Richard Hall said, a rotor is just a fancy toy compared to this. Few people on the TC forum ever took it serious. I have always thought if the spark gap Hz could be tuned to the same Hz as the TC that would be maximum output. I was getting 12 foot arcs from 10kw. I sold all 7 of my TCs, sold the big 10" last summer to the Local University. My 10" PVC pipe actually measured 11.25" diameter with 950 turns of wire 5:1 ratio.

There are a couple of main questions in there. I will address the spark gap one first:

The frequency at which the spark gap fires is ideally 120Hz (for 60Hz mains) or 100Hz (for 50Hz mains). The idea is that it should fire every half-cycle of the high voltage AC supply to ensure maximum energy transfer between the capacitor and primary coil. Of course, this means that the capacitor also has to be matched with the mains frequency/transformer voltage/transformer current to ensure that it can charge up completely in that half cycle. Everything in a Tesla coil works with everything else to get you a decent output. Adjusting the width of the spark gap to change the frequency is only part of the equation. Adjusting the voltage input to the high voltage transformer changes the output voltage, which also affects the spark gap firing frequency, and how much the capacitor charges up before the spark gap discharges it through the primary.

The solid state Tesla coil I am building that uses IGBTs is not yet complete. However, I can tell you that the IGBTs are arranged in an H-bridge formation and operate as an inverter, switching the 170-340VDC from the power rails to alternate current through the primary coil and capacitor. While the theory of operation remains the same, in reality a solid state Tesla coil (a dual-resonant type, in my case) achieves its goal in a much different way than a spark gap Tesla coil.

 

[Mosaic]

Mosaic, I get the sense that you're underestimating Gary's level of experience. He's not a newbie to electronics--He understands the basic concepts of electricity (ohms, volts, amps, watts, etc) and how they relate, as well as more complicated ideas. He's only trying to learn about more modern components (IGBTs, for example, were developed in the '80s but didn't become common until the '90s, at which time Gary was working with other types of electronics/repair).

I know where he's coming from--Until the beginning of last year I had no idea what an IGBT was either, and only learned because I was building a solid state Tesla coil that required high-power switching devices.

I don't know if Gary feels this way, but it seems to me like you're patronizing him, which is entirely unnecessary.

Well, I do think the OP has inadequate fundamentals since he can't fully understand the TL494 spec sheet and use the very basic oscillator equation. The OP thinks he just needs to bone up on the latest component functions to do the job. Yet he patches entire circuits together piecemeal from different sources, with no analysis evident. I had hoped to reduce his frustration and risk with advice. However, as you are a super moderator I yield to your authority and will desist from my approach.

 

[gary350]

Its been 15 years since I built a TC I had forgotten about firing the spark gap on the AC peaks so caps have maximum charge. When I turned the variac my vacuum fan speeded up and the output arcs got longer and longer to a point where they stopped getting long. If I continued to increase fan speed output arcs started getting shorter. I must have been dialing in on the AC peaks. This worked good on all my TCs 4", 6", 8", 10" except for the tiny 120 watt TC.

 

[DerStrom8]

Well, I do think the OP has inadequate fundamentals since he can't fully understand the TL494 spec sheet and use the very basic oscillator equation. The OP thinks he just needs to bone up on the latest component functions to do the job. Yet he patches entire circuits together piecemeal from different sources, with no analysis evident. I had hoped to reduce his frustration and risk with advice. However, as you are a super moderator I yield to your authority and will desist from my approach.

Mosaic, you know very well that is not what I mean. My point is that you're treating the OP as if he was inferior to you. There is no call for that here. Feel free to keep helping, but please do so without the "Better-than-thou" attitude. My being a moderator has nothing to do with me wanting members to treat others with respect.

Its been 15 years since I built a TC I had forgotten about firing the spark gap on the AC peaks so caps have maximum charge. When I turned the variac my vacuum fan speeded up and the output arcs got longer and longer to a point where they stopped getting long. If I continued to increase fan speed output arcs started getting shorter. I must have been dialing in on the AC peaks. This worked good on all my TCs 4", 6", 8", 10" except for the tiny 120 watt TC.

The fan speed is adjusted to quench the sparks in the spark gap. Faster fan = faster airflow = more quenching => faster spark gap firing frequency. Your "sweet-spot" where the arcs from the topload were longest is likely where your fan was quenching the sparks at just the right time so that they fired at around 120Hz. So yep, I expect you were indeed "dialing in on the AC peaks".

 

[gary350]

I set all the gaps with a feeler guage they were all about .028" that way all the gaps had to fire not just 1 gaps that was adjusted closer than all the rest. The pic of the 10" TC looks a bit strange because I used steel rods as targets left and right side to get a better photo of arcs rather than randon arcs all over the place.