Hi Gary,
I agree completely with your numbers. It's not hard to make the input switch
at a lower RF level, but I didn't do that because it would have complicated
the RF sensing circuit (which now is just a simple diode rectifier). I asked
local hams to listen to the amplifier signal on a dead band, and nobody
could hear clicks or distortion products from bias switching.. Keep in mind
that the IMD products you mentioned are on a 10W signal, and only last about
2-3 mS. So while the 10W is only 20-25 db or so down from the full amplifier
output, the IMD products are down at the 1W or lower level. That said, I
agree completely that my EBS circuit reflects a concession to
practicalities. However, I don't think it contributes to noticeable
interference on the ham bands.
73,
Jim W8ZR
> -----Original Message-----
> From: Gary Schafer [mailto:garyschafer@largeriver.net]
> Sent: Thursday, October 30, 2014 12:14 PM
> To: 'Jim Garland'; amps@contesting.com
> Subject: RE: [Amps] Reducing heat in tetrode amps with EBS
>
> At first glance switching the bias at 1/2 watt input seems like a low
value
> but if you have a 13 db gain amp that puts the output power at 10 watts
> during bias switching. That is only 20 db down from 1000 watts output.
Your
> amp is operating in a very non linear region while delivering power out
that
> is down only 20 db and that does not even consider transients generated
> during switching.
>
> This would generate IM products during that time. Granted the IM products
> would only exist for a short time but they will still cause interference.
>
> 73
> Gary K4FMX
>
> > -----Original Message-----
> > From: Amps [mailto:amps-bounces@contesting.com] On Behalf Of Jim Garland
> > Sent: Thursday, October 30, 2014 12:51 PM
> > To: amps@contesting.com
> > Subject: [Amps] Reducing heat in tetrode amps with EBS
> >
> > Some people don't like tri-state EBS to cut down on heat in tetrode
> > amps, but I found it works very well in my homebrew 3-tube GU74B
> > amplfier
> > and is easy to implement. The three bias levels are -120V(cutoff), -
> > 62V(CW),
> > and -50V(SSB). In the SSB mode, an RF detector at the amplifier input
> > switches the bias from -62V to -50V when the RF input reaches 1/2 W.
> > Between words and pauses in speech, the bias is at -62V, where the
> > quescent
> > plate current is only about 50mA per tube. The criticism of this scheme
> > is
> > the possibility of clicks on the transmitted signal when the bias
> > swtiches,
> > but I've had no reports of that. I set the time delay for the circuit to
> > be
> > long enough that it doesn't trip between syllables, but does trip
> > between
> > words.
> > The EBS circuit is implemented on a 3x5 inch PCB which holds the
> > bias regulator circuitry. A small solid state relay (in a 4-pin DIP)
> > toggles
> > the bias between the two settings. The tubes run at 2600V plate voltage
> > and
> > 325V screen voltage.
> > 73,
> > Jim W8ZR
> >
> > > -----Original Message-----
> > > From: Amps [mailto:amps-bounces@contesting.com] On Behalf Of Paul
> > Baldock
> > > Sent: Thursday, October 30, 2014 10:48 AM
> > > To: Jim Thomson; amps@contesting.com
> > > Subject: Re: [Amps] Alpha 8410 using 4CX1500B - why?
> > >
> > > One nice thing about home brewing a grid driven tetrode amp with the
> > > GU74B is that you don't need a tuned input, just a 50 Ohm resistor.
> > > With 2 tubes, 50-60W drive gives you 1500W out. With suitable cooling
> > > the extra heat is not an issue. The NOS tubes are still available for
> > > around $150.
> > >
> > > The biggest issue with GU74B is the failure due to grid to cathode
> > > shorts, and even that can often be fixed (at least for a while) by
> > > zapping the short away.
> > >
> > > - Paul
> > >
> > > At 08:06 AM 10/30/2014, Jim Thomson wrote:
> > > >Date: Thu, 30 Oct 2014 10:51:07 +1030
> > > >From: "Leigh Turner" <invertech@frontierisp.net.au>
> > > >To: <amps@contesting.com>
> > > >Subject: Re: [Amps] Alpha 8410 using 4CX1500B - why?
> > > >
> > > >
> > > >Indeed Peter.
> > > >
> > > >And high plate bias current and associated low efficiency is part and
> > parcel
> > > >of achieving high linearity and low intermod from AB1 tetrodes.
> > > >
> > > >Low plate idle current and low IMD are mutually exclusive objectives.
> > There
> > > >are no free lunches here :-(
> > > >
> > > >At least in an SSB linear amplifier one can significantly offset the
> > impact
> > > >of high idle current with a competently implemented EBS system. When
> > EBS
> > is
> > > >done properly there are no audible or spectral artefacts stemming
> > from
> > the
> > > >tri-state transitional bias switching.
> > > >73
> > > >
> > > >Leigh
> > > >VK5KLT
> > > >
> > > >## Unless your tri-state EBS is being used on PTT, it buys you
> > > >nothing when using VOX.
> > > >Emtron EBS wont switch fast enough between syllables, much less
> > words.
> > > >No free lunch there..and ditto with ameritron, et all. On a GG amp,
> > > >you can run the idle current
> > > >a lot lower on ssb mode. On CW mode, you can easily reduce the idle
> > > >current to just above zero.
> > > >
> > > >## 2 x 4CX-1000s vs a single 8877 ? No brainer there
> > > >folks. The 8877 wins hands down.
> > > >2 expensive breech block sockets, 2 tubes, 2 chimneys, 2 parasitic
> > > >suppressors, regulated screen supply,
> > > >zero watts grid diss, no thanks. The list goes on and on. Then to
> > > >top is all off, the 4CX-1000 is no longer available.
> > > >Its yet another of alphas silly tube lineups.
> > > >
> > > >Jim VE7RF
> > > >
> > > >
> > > >
> > > >_______________________________________________
> > > >Amps mailing list
> > > >Amps@contesting.com
> > > >http://lists.contesting.com/mailman/listinfo/amps
> > >
> > > _______________________________________________
> > > Amps mailing list
> > > Amps@contesting.com
> > > http://lists.contesting.com/mailman/listinfo/amps
> >
> > _______________________________________________
> > Amps mailing list
> > Amps@contesting.com
> > http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|