[Amps] GG Class C

Gary Schafer garyschafer at comcast.net
Fri Feb 17 11:19:01 EST 2006


Hi Jeff,

That's a really long post!  My thoughts on modulating a GG class C amp. The
amount of driver power that doesn't need to be modulated is that amount that
is used up as actual grid drive in the final, that which does not get fed
through to the antenna. The amount that does get through needs to be
modulated.

The reason that you still have feed thru power in class C is that the driver
is in series with the final so even if the conduction time is less than 180
degrees what is there is in series.

>From what I understand about the input network on a GG stage is it needs to
be a pi network or at least a capacitor at the cathode in order to properly
conduct the rf currents to ground.

As to the feed through power not happening during plate cutoff time,
consider that there is no driver plate current at that time either. It is
dependent on the flywheel effect of the tank circuits to make up the rest of
the cycle just like any other amplifier.

73
Gary  K4FMX

> -----Original Message-----
> From: amps-bounces at contesting.com [mailto:amps-bounces at contesting.com] On
> Behalf Of Xmitters at aol.com
> Sent: Friday, February 17, 2006 3:48 AM
> To: amps at contesting.com
> Subject: Re: [Amps] GG Class C
> 
> In a message dated 2/16/06 10:01:12 PM Central Standard Time,
> amps-request at contesting.com writes:
> 
> << s it possible and/or feasible to put a grounded grid linear into Class
>  C?  Reason I ask is my old Tempo/Henry 6n2 is used almost exclusively
>  for WSJT modes and they all do fine on Class C since they send single,
>  sequential tones.  Amp is configured as push-pull on 2m and parallel on
>  6m.  Current biasing arrangement is single 10v, 50w zener in cathode
>  circuit.  I would think the efficiency would be much better in class C
>  if that is possible and would seem to help the 3CX400A7/8874's along.
>  Searches and old handbooks haven't been much help.  Plenty of info out
>  there on tetrodes and biasing but not gg triodes.
> 
>  Thanks - Bruce, N5SIX >>
> 
> Bruce:
> 
> This is certainly possible to do. There are several high power short wave
> transmitters that I am familiar with that have a GG class C final (that is
> plate
> modulated). There is an old radio Handbook construction article that shows
> 4-400A grounded, both for DC and AC. Given the schematics of these 250 kW
> short
> wave transmitters I referred to, the PA tubes used therein are nothing
> special
> in the sense that they can operate at other common amplifier modes as
> well. I
> cannot see why virtually any tube could not be used in a GG configuration
> as
> long as the grid was grounded for RF but not necessarily DC. the 250 kW
> Continental Electronics 419C uses two 8388 triodes and here, the grids are
> at RF
> ground but not DC ground. The bias supply connects from control grid to
> ground.
> The general Electric 250 kW BT-250A, uses two ML-7482 triodes, and is
> similarly
> configured in that the control grid is grounded at RF and not DC. BTW,
> both of
> these transmitters have modulated driver stages.
> 
> The physical layout of your ham rig may or may not allow GG operation
> especially if the grid goes to ground at RF _and_ DC, In which case,
> grounded grid is
> only applicable of the tube specs allow it. Of course, you can raise the
> cathode above ground by the class C bias amount, but this gets cumbersome
> with a
> tetrode. Remember too that according to the Care and feeding of Power grid
> Tubes, there is a formula that relates cathode inductance to its apparent
> DC
> resistance effects seen from the perspective of the control grid. I'm not
> sure how
> that concept would change in a GG configuration. My guess is the amplifier
> would be hard to drive because of all of the inductances from all of the
> isolation circuitry for the power supplies.
> 
> Along the lines of the GG class C, I have done some mathematical analyses
> on
> the 250 kW short wave rigs I mentioned and it appears that, unlike class
> B,
> the driver looks like it does not have to be modulated to 100% for the PA
> to
> make it to 100% modulation. I'm wondering if there is anyone here that
> might be
> able to look over my calculations and tell me if I'm right. My
> calculations
> suggest that if the envelope modulation of the driver falls below plate
> current
> cutoff, the PA will be able to go to 100% negative modulation even though
> the
> driver might not be. The positive modulation ability of the PA of course
> being
> dependent on, among other things, the amount of RF drive it gets at 100%
> positive crest. I also find that if the IPA and PA are separated by a
> simple tapped
> inductor IPA tank, that the PA will be severely overdriven at 100% peak
> positive crest. Therefore it would appear that the driver should not be
> driven to
> 100% positive to keep the drive to the PA below the point where the PA
> grid is
> overdriven. True the grid is grounded, but some of that RF cathode current
> from the IPA lands on the grid and most of it is passed on to the PA to
> add to
> the output. But, with the IPA 100% modulated, the operating line is pushed
> right
> off the tube constant current chart at peak positive modulation crest.
> 
> I discovered that the problem could greatly be reduced by putting a 90
> degree
> PI network between the IPA and the PA (because of its voltage to current
> conversion capability) the overdrive of the PA was better controlled. The
> retired
> transmitter TWR engineers on Bonaire that used the 419C on a daily basis,
> mentioned that its modulation was very much dependent on how everything
> was tuned.
> And when I say tuning, it goes very much beyond the usual simple
> precautions
> of making sure the PA is properly loaded.
> 
> All of the books that I have seen discuss grounded grid amplifiers, but
> only
> class B GG amplifiers. The only mention about the extent to which the
> driver
> has to be modulated was in Bill Orr's Radio Handbook where he mentioned
> the
> driver modulation being only 70% but with no detailed explanation as to
> why. The
> GE transmitter manual mentions the PI network between the IPA and PA, but
> there is no mention about its phase shift or what exactly it is doing.
> 
> It does not make sense to me that he driver would have to go to 100%
> negative
> modulation unless the modulated PA was class B, but why would you run it
> class B? Yes, the driver output is in series with the PA, but the PA tube
> is not
> conducting over 180 degrees, but less than that in class C. Flywheel
> effect
> would not be able to make up that difference between 180 degrees
> conduction angle
> for class B and the 120 degree class C conduction angle. A thevinin
> voltage
> source is replaced with a short when its output voltage is zero in the
> case of
> a linear circuit, which would then allow the driver signal to pass
> straight
> through, but is that theory only true in a linear case? When the tube is
> cut off
> it's not passing current, period. It is therefore acting like a switch and
> it
> is "open" when not passing current. True? Not true? If true then how would
> that driver signal make it through to the antenna if the PA is biased
> beyond
> cutoff?
> 
> What are your thoughts on GG class C and plate modulation?
> 
> I would be interested in corresponding with any of you that would have
> some
> insights on plate modulated class C grounded grid RF amplifiers. The GG
> amplifier was of course used in the short wave transmitters because this
> configuration does not need neutralization; a significant advantage for
> something high
> power and tunable from 3.2 to 22 MHz.
> 
> Jeff Glass, BSEE
> WB9ETG
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps




More information about the Amps mailing list