[Lf] [Fwd: LF: RE: Amplifier Topology]

[Andre' Kesteloot]

Talbot Andrew wrote:

>
>
>      Hello all - Mal and others have been asking for
>      details of the high power amplifier I described
>      earlier.  At the moment only the QRP 650W
>      version has been tested and the 2-3kW design
>      concept was a best otline estimate based on my
>      tests of that breadboard.As this sort of design
>      is not for the faint hearted I am not just going
>      to give out circuits that anyone could try
>      building from (at this stage anyway).  Just
>      think of the damage 5000uF of capacitors charged
>      to 340V would do if shorted.  That is nearly 300
>      joules of energy, equivalent to dropping a 5kg
>      sledge hammer from the 6m second floor of a
>      building.  I blew 4 IRF840s on the breadboard
>      due to various (now obvious) reasons, and each
>      time they went short with an almighty band that
>      set ears singing and me into a state of shock !
>      This design has 5 times that stored energy.The
>      outline concept given is enough to get
>      experienced constructors going.  If you need
>      more info than that, then you're not ready to be
>      building such a beast !!!!!Here's the outline
>      again :Drive stages, 2.Fc input, divide by two,
>      BPSK phase modulator or other on-off gating,
>      HP4426/7 driver chip. NO RAMPING OF DRIVE TO TRY
>      TO AVOID CLICKS, don't even attempt it.   Could
>      try using the invert/non invert HP4428 driving
>      the drive transformer in push-pull (double the
>      primary turns) - gives more power to drive 4 FET
>      devices together.4 x IRF450 or IRF460 in full -
>      bridge driven as per the Decca transmitter
>      output stages with BYV29-500 catch diodes.
>      Current spike soak, 1-2uH CT choke on each pair
>      of devices, damped with 20 ohms   Look at hte LF
>      experimenters manual for the Decca Tx
>      design.Tank circuit Ql = 6, caps rated at 3kV,
>      use series parallel to take the current /
>      voltage.  Inductor wound with Litz wire - use
>      Rayners formula to estimate the turns needed and
>      measure from there.Impedances and transformer
>      taps calculated as per standard techniques Vcc =
>      Vpeak.  Vsine peak = Vsq peak * 4/pi.   Pout =
>      Vcc^2 / 2.RlPSU  Mains - 20A bridge rectifier,
>      5000uF capacitor(s)  output 340V DC, (obviously
>      with soft start and surge limiting). Plenty of
>      input filtering, lots of PSU rail decoupling at
>      wide frequency band - all the usual etcs. of PA
>      designOne of my aims over the Winter Solstice
>      enforced holiday season is to build up a website
>      where circuits and software can be placed.
>      However, don't hold your breath, building more
>      interesting RF projects and software will
>      probably get in the way, so  probably won't
>      happen :-(Andy  G4JNT
>
> --
> The Information contained in this E-Mail and any
> subsequent correspondence
> is private and is intended solely for the intended
> recipient(s).
> For those other than the recipient any disclosure,
> copying, distribution,
> or any action taken or omitted to be taken in reliance on
> such information is
> prohibited and may be unlawful.