When You are using MOSFETS in resonant circuits (like induction heaters, Tesla Coil Drivers and so on) You must be prepared for the voltage at the drain of the power MOSFET or IGBT exceeds the supply voltage by 3 times. Its OK when You're using some step down transformer or Your mains are 110V.
Problems begin when You are trying to avoid the usage of heavy and bulky transformer and Your mains are 220v. Indeed, the rectified supply voltage will then be 220 x sqrt(2) = 311 volts. And the peak voltage at the drain (collector) will then be 311 x 3 = 933 volts.
Now if You look into catalogues of electronics suppliers You will get, that MOSFETs, rated to 1 kV and above, are rare and usually expensive things. Especially when we speak about several tens of amperes. Most of the MOSFETs available are rated to 500..600Volts. Why not to refer to IGBT's then?
Because they have very limited capabilities to increase the operation frequency. You will stick to tens of kilohertz. Or 100-200 kHz in the very best case.
So what to do if You are going to use the comparatively low voltage MOSFETs and expect the peak output voltage to be 2 times higher than their rating? There exist some methods of series connection of transistors and "full-bridge" and "half-bridge" topologies are among them.
Indeed the half- or full- bridge topologies allow to use the 500 v rated MOSFETs in circuits being powered directly from 220v mains. However here You usually have another bunch of problems: "a problem of the high side switch control", a problem of "shoot-through currents" and so on. To solve these problems correctly one has to develop a complicated control circuit for driving the bridge. And this circuit will be pretty difficult to tune up for proper operation.
Meanwhile there exist a simple and elegant solution: SELF OSCILLATING HALF BRIDGE CIRCUIT, that makes control of the high side switch easy, and feeding the circuit through a choke to surpress shoot-through currents.
Lets get started!
We will need:
- IRFP460A transistors - 2 pcs
- 10 kOhm x 0.25W resistors - 2 pcs
- 330 kOhm x 0.25W resistors - 2 pcs
- 68 pf high voltage (10kV or above) ceramic capacitors - 4 pcs
- 2 nf moderate voltage (~1..2kV) ceramic capacitors - 4 pcs
- 1.5KE18CA symmetrical surge arrestor (TVS) diodes - 2 pcs
- 400..1000 uF x 450 v aluminum electrolytic cap - 1 pcs
- NTC SD-11 varistor - 1 pcs
- 16-18 AWG enameled copper wire ~ 2 meters
- 26-30 AWG enameled copper wire ~20 meters
- rectifying diodes (like 10A10) - 4 pcs
- (feel free to use a ready diode bridge like KBU-1009 here)
- A FUSE rated to 3..5 Amps - many
- A (reasonably) large heat sink (from a CPU)- 2 pcs
OPtionally You will need:
A pair of moderate voltage but high reactive power rated caps to make the capacitive side of the bridge if You want to use the load connection type common to half bridge circuits.
Step 1: Look at the Scheme:
- The main tank coil should be large enough to endure currents and voltages. There appear voltages in excess of 5 kv across the coil and currents over 10 Amps.
- The main tank capacitors (C3,C4,C5,C6) should endure high voltages and high reactive powers. They should have high Q-factor too. One can use commercial ceramic capacitors or homemade teflon ones.
- Despite the common opinion the 1.5KE18 surge arrestors are really able to protect the MOSFET gates at working frequences up to 10 MHz at least. One can really see the flat-topped sine wave at the gates if bald enough to attach an oscilloscope probe there.
- The gate-to-source capacitors C1 and C2 should have low inductance and must be rated to high reactive power. Certainly one may search for proper types of low voltage capacitors or try to combine them into banks. But if You really want these caps not to huck up Your brains, the simplest way is to use ones with very high voltage rating surplus. That's why it is written that they have to be rated to 1-2 kv, while there can never appear voltages higher than 30 volts between the gate and the source.
- The choke L2 should have low Q factor for not to produce voltage spikes during switching. Wind it over some resistor or use a shunt one instead.
- The U3 termistor is used to limit the inrush current that appears from charging the rectifying capacitor Crect when You plug the scheme to the mains. Feel free to use a halogen lamp rated to 1..2 kW instead.
- THE FUSE IS A MUST FOR ANY DEVICE POWERED FROM THE MAINS. Don't avoid it in any case, except if You are powering the circuit through a protective halogen lamp.
The circuit is able to operate at frequencies of several MHZ (tested up to 12 MHz and simulated up to 30 MHz with IRFP460A) and provide several hundreds of watts of RF power (not measured directly - only by simulation and by indirect signs like heating something or lighting a bulb).
The RF power can be extracted in several different ways:
- If You need high voltage - it can be extracted through a capacitive coupling from one or both sides of the main tank coil.
- If You need a moderate voltage but high current, the power can be extracted in the common for all half-bridge circuits way (as shown at the picture - note the load resistor Rload). Note also that the requirements for the capacitive side capacitors (C7, C8) are high - they must endure voltages above 1 kV and reactive powers above 1 kVA.
Step 2: Get the Parts
Feel free to use You favourite electronic parts suppliers.
Step 3: Wind the Main Coil
Wind the main coil L1. Use 50 mm jig (in diameter) and wind there 8 to 15 turns of 16-18 AWG wire, dependently to the frequency desired. The length of winding should be longer than 80 mm. Otherwise the HV RF power can make a breakdown along the coil.
Step 4: Wind the Choke
wind the choke coil L2. The winding should have diameter ~20-30 mm and length 80-120 mm. Use 26-30 AWG enameled wire. The winding style is "turn-to-turn". No need to be very accurate here, but too many turn overlaps can cause breakdown or overheat.
Step 5: Assemble the Circuit
The schematics was given above. And now You have all the necessary parts. The example of the design is shown at the picture at the beginning of this step. You may choose some different design - the circuit is rather tolerant to stray inductances.
One more note: when assembling it for the first time, use only 2 nf across the gate-source leads of the transistors. Add the further 2 nf only when You are satisfied with proper circuit operation.
Another note: HIGH RF POWER tends to heat the joints strongly. It may be wise not only to solder them, but also to strap them with some wire - to ensure that they will stay tight, even if the solder have molten. The most serious heat dissipates at the joints of L1 terminals (L1 to C4,C5 and L1 to C3,C6 connections) and at the joints between C4 or C3 and the gate of the correspondent MOSFET.
Step 6: Power It Up
First starts better to make with powering the circuit through some variac. It helps to control the power if something tries to go wrong. It is also wise to connect an ammeter (rated to 10 amps) in series with the circuit. Don't use those fancy electronic ammeters - they will show errors due to the strong electromagntetic interferences. They can even die out. Use simple electromagnetic gauges instead.
With the present values of R2/R1 and R4/R3 dividers the circuit should start from the supply voltage ~180..200 V(AC). The idle current @220 V(AC) is 1-1.5 amps.
When You see the current flowing, check for the presence of RF oscillations. One can use some neon bulb. It should light up at 10-20 cm from the coil. Hold it with extreme care - the HV RF spark may hit Your hands and You will get RF burns of electric shock. Better to use some long dielectric handle.
Note that the common dielectric shells of metal instrument handles are usually not protective enough - the capacitive displacement current can still flow through the stray capacity between Your hand and the metal body of the instrument, resulting in RF burns.
Step 7: Balance If Needed
Since the circuit contains a pair of transistor there can be a case, when You are not satisfied with its symmetry. For example, one transistor is hot as a kettle, while the other one stays cool. It may even seem that only one of them works.
To set it right, one may vary C6 capacitor (or C5, but with C6 the reply of the circuit ith smoother). When You decrease C6 capacity the high side transistor (U1) begins to produce more power and more heat and vice versa (lower power and heat on U1 with increasement of C6).
Generally if You are using commercial grade C3, C4, C5, C6 capacitors, and if the layout of the circuit is reasonably symmetrical, the heat dissipation is equal for both transistors +-30%. Its enough if You're not trying to push it to its top limits. So in most cases You can skip the procedure of balancing.
On the contrary, if You are using homemade capacitors, the balancing may be urgent before proceeding to higher powers.
Step 8: Full Power
Now its time to increase C1 and C2 capacitors to push the circuit towards its full power. The higher their capacity is, the more power is produced by the circuit. With IRFP460A it is more or less safe to set C1 and C2 to 4 nf. 6nf seem to be the margin - power is high, but the MOSFETs behave like matches.
As You could see, the circuit is able to light up a noticeable torch type discharge if one of the L1 ends was equipped with a needle. The coil can also heat a 4mm drill bit to the red hot state in 4 seconds.
And note: all these things can be done when powering the circuits directly from the mains. No need of any step-down transformer or other adapter.
Be carefull however: the parts of the circuit are under the voltage of the mains (sorry, Im unable to call 220 v as "high voltage", when speaking seriously) and hereby DANGEROUS! The possibility of sparks, torch and RF arc discharges and RF burns makes the thing to be EVEN MORE DANGEROUS! PLEASE USE WITH CARE!
At my present design the time of continuous operation is limited by C3, C4 C5, C6 capacitors overheat. When the caps begin to smoke and desolder themselves the transistors are barely warm to touch. (Dont't touch the circuit before unplugging from the mains!)
Step 9: Conclusion
The use of the circuit is quite wide. One may use it as a
- inductor heater with unusually high for these devices frequency
- torch discharge producer and plasma generator (driver of RF plasmatron)
- RF power supply for a gas discharge laser
- Wireles energy transfer device
As far as I know the circuit is novel. I've never seen before a self oscillating RF-resonant half bridge based on two transistors of the same polarity (n-MOS). If I'm wrong feel free to notify me and I will gladly include the reference.
Please avoid using this as a radio transmitting device. Its signal is dirty and its power is high. This combination may arise the problems with law in Your country.