Date: Wed, 6 Mar 2013 11:18:18 +1030
From: "Leigh Turner" <invertech@frontierisp.net.au>
To: <amps@contesting.com>
Subject: Re: [Amps] Expert 2K / lack of specs / pragmatism prevails
When homebrewing big SS amps I've found designing 'em big and de-rating them
a couple of dB below sat Po yields plenty adequate IMD performance for clean
ham band SSB use. Certainly commensurate with the IMD of modern transceiver
PA stages and their associated ALC Tx power control loops.
Inefficiency is the inevitable trade-off one makes when wanting to run the
SS amp clean.
Leigh
VK5KLT
## leigh, there would be no loss in eff, IF you used bigger devices, or more
of them.
The inefficiency starts up when you design a 1 kw amp,,and then operate it at
500w.
Its cleaner at 500w, but its inefficient. Reduce the voltage a bit, like 41%,
and eff
is back on track, but imd suffers due to lower voltage used.
## Instead, used devices rated for 2 kw output, design the amp for 1 kw output,
run the highest voltage possible and eff will be as high as it gets. IMD
will be a lot lower,
and devices will run cooler. Trade off is you require more devices or bigger
ones. heck,
the cooling can be based on 1 kw output, and also the pwr supply.
## enough of these freescale devices may be the ticket. loads of power
AND low imd for a change. Just cuz you don’t hear imd when listening to
an S-9 signal on a noisy band doesn’t mean there is no imd. listen to the
same station, but at 40 over S-9, on a quiet band, and presto, imd is easy to
spot. If IMD is only 30 db below pep ( arrl method), IMD will be 10 db
over S-9. IMD really needs to be down 60 db below pep, when listening
off freq.
## Its’ the IMD 7-9-11-13-15th that causes all the problems. Class A and
some type
of sliding bias scheme would work or the eq of EBS. Or just operate the
devices at
well below their max ratings, like 1/3 their rating.
## You don’t see this lousy imd in the commercial /marine/aviation world.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|