There is a thread started by Lee (Wilcom Labs)
http://part15.us/node/1280
There is a thread started by Lee (Wilcom Labs)
http://part15.us/node/1280
which has evolved into a great resource for those in the part 15 AM community who want to dig into some really meaty theory and experimental data.
One of the last posts in this thread by Ermi Roos is an especially good read for those interested in RF measurement techniques. (I added my $.02 and I hope this also helps).
http://part15.us/node/1280#comment-4659
This entire thread and Ermi's recent post is a really good read for the technologists here.
At well over 50 posts we may set a part 15 us record on this one! Kick in with your comments on Wilcom's thread.
Neil
I hope that Wilcom Labs will, indeed, develop the Part 15 AM tube transmitter he proposed in his thread. I wish him great commercial success.
I am planning one more post about tubes. I stated in a previous post in the Wilcom Labs thread that a really efficient tube transmitter requires a powerful narrow-pulsewidth generator to drive the grid. I am now working on a MOSFET driver for the grid. If it works, I will report the results.
Another open question in Part 15 AM that I am working on is whether or not a power MOSFET can be used as the active component of the final stage of an efficient transmitter. In 1978, Frederick H. Raab published an article about a 1.8 to 54 MHz 16 watt broadband amplifier in "Ham Radio" that used power MOSFETs. He claimed 78.5% efficiency. My own experience with power MOSFETs in Part 15 AM designs, however, indicates that the parasitic diode between the drain and source of the MOSFET severely reduces the efficiency. I already mentioned the disadvantages of this parasitic diode in the Wilcom Labs thread. The problem is that the diode dissipates much of the energy that is stored in the drain inductor during the "on" state of gate drive cycle of the MOSFET. Dissipating the drain circuit energy severely reduces the overall transmitter efficiency. Sometimes a Schottky rectifier diode is connected between drain and source (a "snubber" diode) to prevent current from flowing through the parasitic diode. The snubber diode does not help efficiency, however, because the snubber dissipates the drain circuit energy just as well as the parasitic diode in the MOSFET. I will be looking into the issue of the MOSFET parasitic diode carefully and will post in the Wilcom Labs thread if I come up with a definite conclusion.
Ermi,
My experience with MOSFETs is limited to small signal applications so I am not versed on the diode problem. From your description:
The problem is that the diode dissipates much of the energy that is stored in the drain inductor during the "on" state of gate drive cycle of the MOSFET.
is it correct that the energy storage occurs during the ON time and not the dissipation?
Restated, I assume the dissipation occurs when the gate drive turns the device off (flyback negative voltage) and not during the gate ON cycle. I just thought I would ask since I am trying to visualize the problem and want to be assured that I understand this. Your mention of a damping diode suggests that a model for this would be an external diode connected anode to source and cathode to drain across the MOSFET which conducts when the FET turns off. Maybe a diode in series with the drain and the tank, forward biased when the FET is on, would isolate the FET from the flyback voltage This may be naive, but that's why I ask.
Also, wouldn't a similar problem happen with a class C BJT through the collector to base junction?
You mentioned posting more in the Wilcom thread later, but I thought I would bring this basic MOSFET 101 question up here so as not to clutter your future report in that thread.
Neil
The storage of the energy occurs during the ON time of the MOSFET cycle because current flows through the drain inductor during the ON time (Current in an inductor is analogous to voltage across a capacitor). During the OFF time, the energy stored during the ON time causes the drain inductance and capacitance to initiate ringing. The parasitic diode from drain to source causes the negative excursions of the ringing to be clipped off. This clipping of the ringing causes the stored energy to be dissipated.
The same thing does not happen with ordinary transistors, JFETS and vacuum tubes. What I am describing is called the "turn-off behavior" of power MOSFETS. It is described in detail with scope photos on page 5-41 of the Siliconix Mospower Applications Handbook, ISBN 0-930519-00-0. I have seen clipped, positive-going, pulses on the drains of Power MOSFETs used for part 15 AM transmitters that match the scope photos published in this book.
A circuit to isolate the MOSFET from the tank during the OFF time would be just what is needed to prevent dissipation in the parasitic diode, but I don't know how to make such a circuit.