Velice zajímavý směšovač elektronka / fet
Autor: administrator <>, Téma: Schémata, Vydáno dne: 07. 10. 2014


podle W9BRD, text v angličině (z nějaký diskuze na webu)

Novy1 Of late I've been enjoying the use of a hybrid pentode-MOSFET (6SH7-2N7000) mixer that converts 40 meters (7 MHz) to a 1.75-MHz IF with the help of a transistorized local oscillator in the 8-MHz range. (The implementation is used in a highly modified Allied A-2516 receiver--modified such that right now it covers only 40 meters.) I'm slowly working on a detailed writeup for my radio pages, but as that will take as long as it takes and as I think some here might get a kick out of experimenting with it in the meantime, here's the schematic:

Attachment:

rmx_550x516.gif [ 126.63 KiB | Viewed 171 times ]


This mixer is made regenerative by means of the Colpitts divider formed by its grid-to-cathode capacitance and the 2N7000 drain-to-source capacitance. If there's insufficient capacitance in the tube's parallel-tuned grid tuned circuit, the circuit will oscillate. As it is, because the A-2516's stock preselector circuitry tunes 40 meters at the high end of its bottommost tuning range, I connected a 510-ohm resistor between the 6SH7 grid and the tuned circuit to keep the grid circuit from peaking too sharply.

A 2N7000 will need a peak LO voltage of somewhere between 0.8 and 3 V to turn on. The A-2516's local oscillator module (which I run at a regulated 9 V instead of the Zener-regulated 18 V of the stock radio) puts out such a low level that I used the bias circuit at A to pull the 2N7000 gate up far enough to make the LO peaks turn it on. (B shows another, adjustable bias arrangement. The gate-to-common resistance is not critical; you have a lot of latitude in setting up the bias.)

No RF amplifier is necessary with this circuit at 7 MHz; it hears down to the noise floor of my antenna system just fine.

A high-transconductance pentode is not necessary with this circuit, and in fact may well give problems with oscillation that sounds like wideband noise (perhaps superregeneration). This was with a 6AC7 (transconductance at 300 V, 9 mS) as the pentode; experimenting with a lower-transconductance tube than the 6SH7 (4.5 mS at 250 V), I found that a 6SJ7 (1.65 mS) also works well.

This circuit is really cool in that it provides a simple means of interfacing a solid-state, low-voltage local oscillator with a vacuum-tube mixer. For crystal-controlled converter use, you can drive the 2N7000 from a solid-state Pierce oscillator configured to use the 2N7000 gate-to-source capacitance as part or all of the Pierce's necessary anode-to-ground capacitance. Even a 6C4, 6C5, 6J5 or 12J5 Pierce oscillator operating with a plate supply of 12 V will drive the 2N7000 well.

Best regards,

Dave
amateur radio W9BRD