deflection beam mixer tube LO injection idea

[K1ZJH]

I'm thinking of trying a few 6JH8 balanced mixers in my HBR project. I'd like to use balanced LO injection to the deflection electrodes (especially for the first mixer!) to improve the signal feed through rejection, since I do have a local AM'er on 1600kHz next town over

Since the HBRs use ECO LO oscillators, I was thinking the easiest way to achieve this is using trifiliar wound baluns--with either the primary going in parallel with the plate feedback coupling coil (not comfortable with this!)  or even better, using a cathode follower stage to drive the balun.  The two secondary windings would be used to provide 180 degree   out of phase LO injection to the deflection plates.  The LO frequency range is limited in this sets, so I think a balun approach might be feasible?

Any pitfalls that anyone might see to this approach???

Pete


out of phase LO injection to the

[WA1GFZ]

BB transformers are hard to do at high Z so I would think a triode phase splitter might be easier. 

[w3jn]

I used a simple toroid transformer (not trifilar wound, just a primary and a center tapped secondary) on my 7360, Pete.  The 7360 needs about 5V P-P, not sure what the optimum level would be for the 6JH8.

My oscillator/buffer was a 6J6, cathode coupled and the buffer was grounded grid (just a grid leak resistor, IIRC).  A 2.5 mH choke on the cathode provided sufficient coupling and isolation.  It provided *just* enough LO injection, I would have been happier with more, and I don't necessarily recommend that design.  It worked well enough but I tore  it apart (hence the RX isn't working now) to re-do the LO.  THat was several years ago...

I think a triode phase splitter might be just the ticket.

[W1VD]

Interesting idea Frank - probably the easiest way to go...

Pete, FWIW some info on the SS-1R...

The HFO split phase injection coil is untuned - two pieces of concentric miniature miniductor stock. It's flat across the HF LO range and provides a small voltage step up.

The VFO (6.0 - 6.5 MHz) injection coil is a standard looking ferrite slug/phenolic assembly in a 1" X 1" vertical mount 'IF' can. The factory apparently added the two output windings to the coil and it's tuned by capacitors from each deflection plate to ground. Unfortunately, tuning causes the LO output to drop a couple V p-p at each end of the range - not good!

The mixers have a balance control - fixed 25 VDC on one deflection plate and adjustable on the other producing a deep smooth null.

   

[w3jn]

There's really not a reason to used a sheet beam tube in the balanced configuration unless you need to eliminate IF bleedthru, or LO radiation.  Squires did because of the 1500 KC IF, and limited front end selectivity.  My HB used a 1600 KC IF and I needed to use a balanced configuration due to AM BCB bleedthru otherwise.

Nulling the deflection plate knocked the BCB signal out completely; the balanced plate transformer had less of an effect.

[K1ZJH]

There's really not a reason to used a sheet beam tube in the balanced configuration unless you need to eliminate IF bleedthru, or LO radiation.  Squires did because of the 1500 KC IF, and limited front end selectivity.  My HB used a 1600 KC IF and I needed to use a balanced configuration due to AM BCB bleedthru otherwise.

Nulling the deflection plate knocked the BCB signal out completely; the balanced plate transformer had less of an effect.

I have a local AM station on 1600 kHz, so I'm opting for the balanced first mixer....

Frank, the cathode follower should give a low impedance source for a balun.  I can whip a test lashup
on the bench and see how well it works and if it worth pursuing.  I was thinking the RF voltages would be better balanced using the balun vs. an active phase splitter, but thanks for the suggestion.

Pete

[WA1GFZ]

Pete, I think a trifilar transformer would work fine if loaded on the secondary with say 100 ohm 1 watt resistors on each phase. This would reflect 50 ohms back at the follower. The follower would need to swing 20 volts peak to peak or just over a watt of power. A 1/2 to 3/4 inch core with type 61 ui=125 with a dozen or so trifilar turns should cover the HF band. lower frequency operation I would use type 43. Type 43 will work ok at hf but it gets a bit more core loss above 15 MHz. I'm not sure if the deflection plates want to see a low resistance to ground so may need to ac couple the load resistors.

[W1VD]

Pete, if you have the time (and inclination) for further experimentation after trying Franks idea...try a conventional wound toroid with a somewhat higher turns ratio - single primary and split secondary. Interleave as many of the the primary and secondary turns as possible to minimize leakage inductance. Terminate with the appropriate resistance across the deflection plates. DC balance of the deflection plates should compensate for minor LO amplitude imbalance (if any). No guarantees expressed or implied.