--- On Wed, 6/5/13, Bill Turner <dezrat1242@yahoo.com> wrote:
> I'm thinking of building a solid state 1500 watt amp using some of
> MicroSemi's high voltage MOSFETs. This is a new kind of project for
> me and I would appreciate any comments you might have which would
> help me avoid any blunders.
<snip>
Hi Bill,
That's a cute part, and yes, you may design for a 50 ohm impedance but ...
There are a few factors you should consider.
The input drive is not trivial. Perhaps consider an extension to a bridge
configuration of the scheme shown in: "A quasi-complementary class-D HF power
amplifier" by F. H. Raab and D. J. Rupp RF Design, pp. 103-110, 1992
The output capacitance of the device(s) is(are) considerable, not constant,
and that is exacerbated by the high voltage swings. See US patents 6072362 &
7092269 for some ideas about this problem. Note that these are fixed frequency
solutions.
A transmission line transformer consisting of a short length of RG-402 with
loaded with some ferrite beads is probably the best choice for output isolation
rather than a commercial balun.
Doubtful if this approach may prove suitable for a (SSB/AM) Linear amplifier.
All of these schemes are intended for Class D.
The cooling problem, on which many others have commented. Two of those
devices in close proximity will only make things worse. It may be easier to go
with four discrete devices to distribute the heat. IXYS DEI parts may be a
better starting point.
I would recommend a Power Factor Correcting (PFC) front end for the power
supply. See the old Advanced Power Technology (APT) (since acquired by
Microsemi) application note APT9303 for a starting point. Incidently, 2000uF
per KW is a good rule of thumb for these 300-400 volt power supplies.
APT application notes APT9502 & APT9701 may also provide some ideas.
Finally, there's a book on the net that you may find interesting:
http://caxapa.ru/thumbs/406149/SW_HF_PA.pdf
73 & Gooooooood luck!
Marv WC6W
http://qsl.net/wc6w/
*
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|