[AMPS] Amp Drive Power(Long)
Michael J. Tubby B.Sc. (Hons) G8TIC
mike.tubby@thorcom.co.uk
Wed, 17 Oct 2001 12:28:05 +0100
----- Original Message -----
From: "Jim Reid" <jimr.reid@verizon.net>
To: "AMPS" <amps@contesting.com>
Sent: Wednesday, October 17, 2001 12:19 AM
Subject: [AMPS] Amp Drive Power(Long)
>
> Back in July there was a lengthy discussion here about
> an amp's input circuit. Now I have been finding that
> I have no way to really know the amp drive power from
> my exciter rig. I thought I did by having a VFD peak reading
> watt meter to set the drive, then switch that into the amp.
>
> However, W1HIS, Chuck (who has the same sort of Henry
> amp that I have) and I have been doing a lot of comparative
> tests on our amps. He has pointed out issues which
> impact this; I believe he is correct! His points:
>
> "1. Unless the load VSWR is exactly 1.0, the forward power
> indicated by any directional-coupler-wattmeter exceeds the
> net power delivered to the load.
>
> 2. Because the Henry amp's input is a _nonlinear_ load, for
> which voltage is not proportional to current!, all (simple) bets
> are off!! The load is nonlinear because the grid part of the
> cathode/grid circuit rectifies.
>
> 3. How a nonlinear load appears to a wattmeter depends strongly
> on the voltage and current waveforms, which are nonsinusoidal
> and which change depending on the length of the transmission
> line and any filters, tuned circuits, impedance-matching, etc., etc.,
> between the exciter and the load (the Henry's cath/grid circuit). These
> nonsinusoidal waveforms have harmonic content, and the harmonic components
> typically are reflected from both ends of the line and
> form standing waves. Thus, waveshapes are different at different
> points on the line. Garden-variety wattmeters do not sample the
> entire waveforms and do not compute power, really; they use
> diode rectifiers, which tend to respond to peak rather than rms
> voltage, and so on; and they infer approximately what the
> power _would_ be, were only the waveforms sinusoidal."
>
> Next, Chuck ran an experiment with his Icom driven amp:
>
> "Transceiver: Icom IC-775DSP, which is all solid-state. Adjusted
> its output so that the VFD indicated 99 W when the Henry was
> bypassed. The Icom's internal antenna tuner was switched OUT.
>
> [Below I repeat the experiment with the Icom's internal
> antenna tuner switched IN. Whether the Icom's internal antenna
> tuner is switched in or out turns out to make a big difference,
> although it "shouldn't", because in each case the Icom was
> adjusted to deliver 99 W to the dummy load as indicated by the
> VFD.]
>
> 1.6 Indicated PA voltage....4380 or so as read on
> amp meter (voltage as selected using the "CW" bias position)
>
> 1.7 VFD Pout reading into Bird dummy load....1639 watts
>
> But with the Icom's internal antenna tuner switched IN and the
> Icom's power readjusted for 99 W to the dummy load as
> indicated by the VFD, then VFD Pout reading into my Bird dummy
> load is NOW 1896 watts!
>
> Obviously (IMO) a nonsinusoidal reflected wave, very rich in
> harmonics, returns from the Henry's cathode (input) circuit; and
> these harmonics are absorbed or reflected by the Icom
> completely differently depending on whether the Icom's
> antenna tuner is switched in. So the voltage and current
> waveforms delivered to the Henry are very different.
>
> 1.8 VFD indicated vswr....1.20 here, regardless of Icom
> tuner in/out.
>
> 1.9 Bird meter on the amp Controller Unit.......2000 watts
> when my VFD indicates 1896 W.
>
> 1.10 Now, lower input drive power to bring indicated output
> power down to 1500 watts steady CW key down....VFD reads
> either 1486 or 1529, evidently allowed values
> for the digital readout in this area.
>
> "Drive" power, VFD indicated with amp by-passed....88 Wtts."
>
> Chuck then summarizes his and my comparative measurments:
>
> "1. With the amp biased for CW, I get less or more power than
> you do, with the same 99 W of VFD-indicated drive power,
> depending on whether the antenna tuner _inside_ my exciter is
> switched out or in, respectively. IMO this proves that the Henry's
> grid circuit generates and sends a nonsinusoidal reflection back
> to the exciter.
>
> 2. Neither your nor my VFD indicates VSWR = 1.0 (as it should)
> when looking into a 50-ohm Bird termination.
>
> 3. The wattmeter on your amp controller reads higher than your
> VFD; my controller panel meter is also optimistic, but not by
> so much.
>
> 4. I always see less grid current and more plate current than you,
> for roughly the same RF power output. Probably nothing to worry
> about. Our tubes may be biased differently.
>
> 5. My filament voltage(7.4) is lower, perhaps 'cuz I changed the fil.
> transformer primary tap (to the top) to lower the filament voltage.
> Per manual, filament voltage is supposed to be 7.50 plus/minus
> 5%, in other words 7.125 to 7.875 ? [KH7M filament at 7.7 or so
> per meter on amp controller unit].
>
> 6. My TUNE dial readings are the same as yours, but my LOAD
> dial readings are much higher than yours; indicates higher load
> C, lighter loading.
>
> 7. Both of us seem to be getting reasonable amounts of output
> power to the Bird loads (1600 to 1900 watts with "CW" bias and
> 2400-2500 W with SSB bias) with about 100 watts in. The key
> word being "about".
>
> 8. You might be able to get a valid measurement of the drive
> power required by the Henry, by using an "antenna tuner,"
> ahead of the linear amp and inserting a drive-power meter
> between the exciter and said tuner. "
>
> Is there any real benefit to bringing the line between the
> transceiver and the amp input "flat" using either a tuner,
> or perhaps putting in a 3dB or so pad?? Suppose there is
> no real particular benefit to know the actual drive power
> in any case! Not sure where to obtain such a pad to
> handle drive powers up to 150 watts or so anyway such
> that half of that power is dissipated and about 70 or so
> "real" drive power is used. Would the output signal be
> cleaner with such an input tuner, or the use of a pad
> to bring the line flat, free of the non-linearities?
>
> Complete detailed measurement data on each of our
> 3CX3000A7 amps can be supplied if you might be
> interested.
>
> 73, Jim KH7M
>
>
Interesting tests but do you beleive the output of the Icom 775
to be 50R+0j without the tuner... bet it isn't and therefore you
are actually looking at the complex network of:
- output of radio
- length of 50R (ish) coax
- input to Henry
- change of impedance over drive cycle
Can you do the same tests at 1/4 power and get the same (relative)
results? Probably not because the "aparrent" impedance of the
input of the Henry will change as will the "aparrent" output of the
radio.
If you can get some more-or-less comparative results - like basically
you see the same behaviour when switching the tuner in/out then it
should be interesting to see what happens with a 6dB 100W rated
attenuator (a) at the back of the rig, and (b) at the input to the amp.
Its precisely these sort of things which cause me to prefer Tetrode
amps which are grid driven (esp. at VHF) - even then I like a 6dB
pad on the input of the amp.
For your amp you might like to try a 1/2-wave length of coax between
rig and radio on the band of interest [test] (taking velocity factor into
account) to eliminate the "transform effect" of the coax from the
equation.
Mike G8TIC/M0VHF
>
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com