[AMRadio] Pi-Net vs Link Couple

W5OMR/Geoff w5omr at satx.rr.com
Thu Sep 22 17:14:16 EDT 2005

I need to better proof-read my posts, before hitting the 'send' button..

W5OMR/Geoff wrote:
<>Donald Chester wrote:

> What does 1500 watt pep have to do with it?

Well... you know.  I certainly don't want to Advertise ;-)

> My Gates BC1-T uses a pair of 833A triodes in parallel, with a 
> pi-network followed by a T network and another L netork, to couple the 
> final to the antenna.  The grid tank uses a tapped coil, with the tap 
> grounded to produce the out of phase voltage.  Both the adjustment of 
> the tap and the neutralising cap will affect neutralisation.

My rig is a pair of 250THs in the final, modulated by a pair.  Like what 
you prefer, the final has plug-in coils for the grid input tank, as well 
as the final output tank.  The problem I have with that rig however, is 
that there seems to be some interaction between the grid circuit and the 
final circuit, even though the plug-in grid coil and link are 
'underneath' the chassis, that the final coil and link sits on top of.  
When the rig was built back in the mid 50's, most everything available 
was steel, so that's what was used as a chassis.  (there, that makes 
more sense ;-))

A suggestion of physically raising the final coil further up to get out 
of the field of the grid tank could be realized, if i wanted to give up 
the front-panel control of the Faraday [-shielding-] sheilded link.  As 
a result, neutralization is as close as possible/

> The problem with grid neutralisation, sometimes called Rice 
> neutralisation, is that it doesn't hold very well over a wide 
> frequency range, especially if you attempt to switch over several 
> amateur bands.  My Gates stays neutralised over the 160m band from 1.8 
> to 2.0, but I never have tried to use it on any other band.

I've got plug-in coils for that rig, from 80m thru 10m, but I'm more 
than a little leary of running a pair of 250TH's in Class C on 10m ;-)

I wonder though Don, while I've got your .. uhm 'ear' (eyes?)  The link 
could use a little tuning to take a bit more heavier load in the final.  
I've modified  one of my 80m B&W HDVL plug-in coils and removed two 
turns on each side (4 turns, total), to raise the plate current a bit, 
because the Q is a bit sharp in that circuit.  A breif excursion 
off-resonance, w/1.5kWDC on the plates, draws around 550mA.  At 
Resonance (with 125mA of grid drive) the thing dips to around 200mA.  
That said, what value of series capacitance would you suggest/recommend 
to tune the link?

> Plate neutralisation, using the same kind of tank circuit as a 
> pushpull final, and single ended grid tank, works better over a wider 
> freq range, because the plate-to-ground capacitance is usually much 
> less than the grid-to-ground capacitance, and capacitance across one 
> side of the tank circuit upsets the balance of the circuit.  Also, 
> grid  loading effects cause some additional unbalance, even if the 
> capacitance is perfectly balanced out with additional fixed 
> capacitors.  The pushpull circuit works best of all, since it is 
> inherently a balanced bridge circuit, and theoretically works equally 
> well over an extremely wide frequency range.  Limitations lie in the 
> precision of the balance of the split stator tank capacitors from 
> minimum to maximum capacitance.

In my rig, the neutralizing caps come from the crossed grid-input lines, 
and big silver disks mounted on screws on the back end of the B&W 
bread-slicer butterfly tuning capacitor are what nulls out the 
differences in capacitances in the two tubes.  However, I know it is 
possible to use a single tube in the final, like K5SWK's single 833, 
modulated by a pair.  I -think- Otis is using a big capacitor in place 
of the 2nd tube and he's able to maintain a balanced tank circuit.  I 
had him on the phone earlier, but we didn't talk about rigs.
Looking forward to your reply and info.

That should read better.


More information about the AMRadio mailing list

This page last updated 22 Feb 2018.