Class E SSTC problems
Andrew L., Fri Apr 27 2007, 11:51AMOk so I built this circuit:
But I modofied it just a litle bit, instead of C6 being connected in series with the primary coil I connected it paralel to the primary coil. And usualy with this circuit I blow a mosfet prety fast. is this because the mosfet is rated for only 500V, and when the mosfet turns off the high voltage spike feeding back from the primary coil exceedes the voltage rating of the mosfet or is there something else?
And usualy when I blow a mosfet the gate driver also blows, why is that?
So do you have any suggestions on how I could improve the circuit so that the mosfet and gate driver wouldn't blow.
Re: Class E SSTC problems
r3d, Fri Apr 27 2007, 12:26PM
Try to connect the cap as it is in the schematic?
r3d, Fri Apr 27 2007, 12:26PM
Try to connect the cap as it is in the schematic?
Re: Class E SSTC problems
Zum Beispiel, Fri Apr 27 2007, 12:50PM
I don't know much about class E, but I'm fairly sure the cap should be tuned so that it's at 0V when the mosfet switches back on. If it has a lot of charge, it will be shorted by the mosfet. Do that, say a million times a second, and you have a blown mosfet.
Oh, btw, what is your supply voltage? Try running it of a lower supply voltage and see if that helps.
And definately put the cap across the Fet.
Zum Beispiel, Fri Apr 27 2007, 12:50PM
I don't know much about class E, but I'm fairly sure the cap should be tuned so that it's at 0V when the mosfet switches back on. If it has a lot of charge, it will be shorted by the mosfet. Do that, say a million times a second, and you have a blown mosfet.
Oh, btw, what is your supply voltage? Try running it of a lower supply voltage and see if that helps.
And definately put the cap across the Fet.
Re: Class E SSTC problems
Andrew L., Fri Apr 27 2007, 01:13PM
Why is it so important that I put the capacitor across the fet?
Andrew L., Fri Apr 27 2007, 01:13PM
Why is it so important that I put the capacitor across the fet?
Re: Class E SSTC problems
HV Enthusiast, Fri Apr 27 2007, 01:33PM
The shunt capacitance there works with the L and load (R) to cause a damped oscillation and make the switch close precisly when voltage and current are zero. The value of this capacitance, as well as the load will determine how the damping works.
For low Vds, the MOSFET has a pretty large Cds and doesn't need much external capacitance, but as you increase the Vds, the Cds of the MOSFET does decrease, so you need to add more.
The capacitance doesn't necessarily have to be across the MOSET, but in practice, is generally is placed there.
Also, your C6 doesn't have to be polypropylene. Usually, it may only require a small few hundred pF from a ceramic capacitor - of course depending on your load and power level. I run my Class-E supply at about 300-400W and use only a few hundred pF across my power stage MOSFET.
HV Enthusiast, Fri Apr 27 2007, 01:33PM
Andrew L. wrote ...
Why is it so important that I put the capacitor across the fet?
Why is it so important that I put the capacitor across the fet?
The shunt capacitance there works with the L and load (R) to cause a damped oscillation and make the switch close precisly when voltage and current are zero. The value of this capacitance, as well as the load will determine how the damping works.
For low Vds, the MOSFET has a pretty large Cds and doesn't need much external capacitance, but as you increase the Vds, the Cds of the MOSFET does decrease, so you need to add more.
The capacitance doesn't necessarily have to be across the MOSET, but in practice, is generally is placed there.
Also, your C6 doesn't have to be polypropylene. Usually, it may only require a small few hundred pF from a ceramic capacitor - of course depending on your load and power level. I run my Class-E supply at about 300-400W and use only a few hundred pF across my power stage MOSFET.
Re: Class E SSTC problems
ragnar, Fri Apr 27 2007, 02:46PM
The capacitor across the MOSFET catches the EMF from the primary; the shunt capacitor and the primary are a resonant circuit. Place the shunt capacitor as close as you can to the MOSFET drain and source.
If this is anything like your last coil, your shunt cap should probably be less than 1nF.
When tuning this kind of coil, it really is important to use a signal generator or a variable frequency circuit to drive the MOSFET. Slightly-under or slightly-over F(res) may give the most efficient operation. It's also important to be able to see how your drain waveform changes with varying frequency.
Also critically important is that the MOSFET duty cycle is LESS than 50%. 35-40% is a good figure to start with.
Do you have an oscilloscope? After you set it up with a variable-frequency drive, perhaps show us some photographs of your drain waveform, and we can help. Tweaking the duty cycle is another variable to play around with, but remember to keep it away from 50%!
ragnar, Fri Apr 27 2007, 02:46PM
The capacitor across the MOSFET catches the EMF from the primary; the shunt capacitor and the primary are a resonant circuit. Place the shunt capacitor as close as you can to the MOSFET drain and source.
If this is anything like your last coil, your shunt cap should probably be less than 1nF.
When tuning this kind of coil, it really is important to use a signal generator or a variable frequency circuit to drive the MOSFET. Slightly-under or slightly-over F(res) may give the most efficient operation. It's also important to be able to see how your drain waveform changes with varying frequency.
Also critically important is that the MOSFET duty cycle is LESS than 50%. 35-40% is a good figure to start with.
Do you have an oscilloscope? After you set it up with a variable-frequency drive, perhaps show us some photographs of your drain waveform, and we can help. Tweaking the duty cycle is another variable to play around with, but remember to keep it away from 50%!
Re: Class E SSTC problems
Dr. Shark, Fri Apr 27 2007, 04:24PM
Good to see more people play with Class E! Almost everything has been said, one addition though:
Do you have a resistor between the driver and the gate? If you dont, the driver will discharge into a short and overheat in very little time. I usually omit that resistor for maximum current into the gate and pay for it with a lot of dead TC4422s
Dr. Shark, Fri Apr 27 2007, 04:24PM
Good to see more people play with Class E! Almost everything has been said, one addition though:
Andrew L. wrote ...
And usualy when I blow a mosfet the gate driver also blows, why is that?
So do you have any suggestions on how I could improve the circuit so that the mosfet and gate driver wouldn't blow.
And usualy when I blow a mosfet the gate driver also blows, why is that?
So do you have any suggestions on how I could improve the circuit so that the mosfet and gate driver wouldn't blow.
Do you have a resistor between the driver and the gate? If you dont, the driver will discharge into a short and overheat in very little time. I usually omit that resistor for maximum current into the gate and pay for it with a lot of dead TC4422s
Re: Class E SSTC problems
Andrew L., Fri Apr 27 2007, 05:07PM
Ok I forgot to say that this is my first SSTC
and that I am a total beginner.
Andrew L., Fri Apr 27 2007, 05:07PM
Ok I forgot to say that this is my first SSTC
and that I am a total beginner.BlackPlasma wrote ...
Do you have an oscilloscope?
Yes I do have an oscilloscope, but it is quite old and not working correctly.Do you have an oscilloscope?
joe wrote ...
Do you have a resistor between the driver and the gate?
Yes I do have a resistor between the driver and the gate.Do you have a resistor between the driver and the gate?
Re: Class E SSTC problems
WaveRider, Fri Apr 27 2007, 05:24PM
Andrew,
Class-E requires precise phasing between drain current and voltage waveforms.. I would need to think about your circuit a bit before commenting on whether or not you can get Class-E.
Moreover, usually, you want the generator to look like a current source during the transistor off-periods..
Have you tried any spice simulations of this circuit?
Cheers,
Bill
WaveRider, Fri Apr 27 2007, 05:24PM
Andrew,
Class-E requires precise phasing between drain current and voltage waveforms.. I would need to think about your circuit a bit before commenting on whether or not you can get Class-E.
Moreover, usually, you want the generator to look like a current source during the transistor off-periods..
Have you tried any spice simulations of this circuit?
Cheers,
Bill
Re: Class E SSTC problems
Andrew L., Fri Apr 27 2007, 05:49PM
I don't have spice program.
Andrew L., Fri Apr 27 2007, 05:49PM
I don't have spice program.
Re: Class E SSTC problems
Terry Fritz, Sat Apr 28 2007, 12:40AM
Hi,
Here is a spice program for the "right price"
We used to tune the driver side to the FET too with RLC circuits instead of resistors to get lots of low loss current drive at 13.56MHz to the FET gate.
I would give some patents, but is AFU for "patent search" links at the moment...
Terry
Terry Fritz, Sat Apr 28 2007, 12:40AM
Hi,
Here is a spice program for the "right price"
We used to tune the driver side to the FET too with RLC circuits instead of resistors to get lots of low loss current drive at 13.56MHz to the FET gate.
I would give some patents, but
is AFU for "patent search" links at the moment...Terry
Re: Class E SSTC problems
Ken M., Sat Apr 28 2007, 12:47AM
I found that if you don't mind NOT having the adjustable duty cycle just wire the R2 and R3 block with a sing pot with no diodes and it should work. Also as a side question but somewhat relevant, what would be the MAX voltage applied to the fet, on a circuit like this, and would adding a Cap in paralel to the primary do any good as well as keeping the cap across the D\S of the fet?
Ken M., Sat Apr 28 2007, 12:47AM
I found that if you don't mind NOT having the adjustable duty cycle just wire the R2 and R3 block with a sing pot with no diodes and it should work. Also as a side question but somewhat relevant, what would be the MAX voltage applied to the fet, on a circuit like this, and would adding a Cap in paralel to the primary do any good as well as keeping the cap across the D\S of the fet?
Re: Class E SSTC problems
ragnar, Sat Apr 28 2007, 02:38AM
If you have a proper flywheel / output network connecting the primary, there are many variables you can alter:
Cshunt
Cload (primary)
Cseries (flywheel)
Lseries (flywheel)
Lload (primary)
And then there's coupling. You really need to play around.
ragnar, Sat Apr 28 2007, 02:38AM
If you have a proper flywheel / output network connecting the primary, there are many variables you can alter:
Cshunt
Cload (primary)
Cseries (flywheel)
Lseries (flywheel)
Lload (primary)
And then there's coupling. You really need to play around.
Re: Class E SSTC problems
Andrew L., Sat Apr 28 2007, 08:32AM
For now I'll have to put the project on hold because I ran out gate driver chips.
Andrew L., Sat Apr 28 2007, 08:32AM
For now I'll have to put the project on hold because I ran out gate driver chips.
Print this page