gbugh
Member level 5
- Joined
- Mar 15, 2013
- Messages
- 81
- Helped
- 1
- Reputation
- 2
- Reaction score
- 1
- Trophy points
- 1,288
- Location
- Fort Worth, Texas
- Activity points
- 2,234
I've been studying VHF DC-DC converters to see how to make the most efficient and smallest size of an envelop tracking power supply for an RF transmitter power amplifer.
It needs to be a capacitively coupled (or transformer coupled) boost converter so the output voltage can go down to near zero volts. I'm thinking of using a lower frequency like 500KHz to control bursts of VHF switching with push-pull like Class E ZVS type circuits.
I see a few research papers like this about VHF converters:
**broken link removed**
I've also studied step recovery diodes. When the diode gap snaps wider when reverse biased, the charge at the junction makes the typical reverse recovery spike and ringing. Some people utilize the feature to make powerful pulse generators. Certain silicon diodes work better than Schottky diodes when purposely making these spikes.
There are many patents like this that describe the effect:
https://www.google.com/patents/US5774348
It seems like some VHF designs could and possibly already are getting a benefit from these reverse recovery pulses if a MOSFET is switched at close to the recovery ringing frequency and the voltage and current of the diode junction generated spikes can be phased correctly. With the correct phasing the ringing would become more like zero voltage switching. Some MOSFETs seems to have body diodes that would lend themselves to this if they were switched at VHF frequencies.
See typical ringing in figure 11 here, not that I'm suggesting this particular slow switch:
https://www.hpe.ee.ethz.ch/uploads/tx_ethpublications/05543337.pdf
Is it even possible to phase the voltage and current of this ringing so it is a low loss source of more power to be rectified by much faster diodes, so as to increase over all efficiency?
Maybe if the inductor core is switched near the frequency where complex permeability becomes a factor then it would help with optimum phasing of the reverse recovery current and voltage. When magnetic materials are operated at high enough frequency, the magnetic field from the wire coiled around the material couples to both the magnetic materials individual electron spin directions and the electron spins' gyrations around their spin axes. This gets the B and the H curves out of phase with each other. Usually this results in more heat loss but if the frequency is not too high above this ferromagnetic resonance frequency maybe there can be just enough effect to help with the phasing of the reverse recovery current and voltage. I'm just making completely untested conjectures.
George
It needs to be a capacitively coupled (or transformer coupled) boost converter so the output voltage can go down to near zero volts. I'm thinking of using a lower frequency like 500KHz to control bursts of VHF switching with push-pull like Class E ZVS type circuits.
I see a few research papers like this about VHF converters:
**broken link removed**
I've also studied step recovery diodes. When the diode gap snaps wider when reverse biased, the charge at the junction makes the typical reverse recovery spike and ringing. Some people utilize the feature to make powerful pulse generators. Certain silicon diodes work better than Schottky diodes when purposely making these spikes.
There are many patents like this that describe the effect:
https://www.google.com/patents/US5774348
It seems like some VHF designs could and possibly already are getting a benefit from these reverse recovery pulses if a MOSFET is switched at close to the recovery ringing frequency and the voltage and current of the diode junction generated spikes can be phased correctly. With the correct phasing the ringing would become more like zero voltage switching. Some MOSFETs seems to have body diodes that would lend themselves to this if they were switched at VHF frequencies.
See typical ringing in figure 11 here, not that I'm suggesting this particular slow switch:
https://www.hpe.ee.ethz.ch/uploads/tx_ethpublications/05543337.pdf
Is it even possible to phase the voltage and current of this ringing so it is a low loss source of more power to be rectified by much faster diodes, so as to increase over all efficiency?
Maybe if the inductor core is switched near the frequency where complex permeability becomes a factor then it would help with optimum phasing of the reverse recovery current and voltage. When magnetic materials are operated at high enough frequency, the magnetic field from the wire coiled around the material couples to both the magnetic materials individual electron spin directions and the electron spins' gyrations around their spin axes. This gets the B and the H curves out of phase with each other. Usually this results in more heat loss but if the frequency is not too high above this ferromagnetic resonance frequency maybe there can be just enough effect to help with the phasing of the reverse recovery current and voltage. I'm just making completely untested conjectures.
George
Last edited:
