Sunday, April 22, 2012

Two-Penny "Mystery" Tubes

I took a chance on a little box of QF721 "mystery" sub-miniature tubes some weeks ago; paying the 1953 equivalent of two and a half cents per tube (including postage!). 

As Raytheon used the "QF" prefix to indicate an experimental or special-order tube, the specifications for this device may never have been released to the public. However, the listing photo suggested these were pentodes in a T2x3 glass bulb. I also assumed that it used a directly-heated cathode (filament) that was so common in Raytheon's other lines of sub-miniature tubes.   

Upon the arrival of my "mystery" tubes I noticed they were stamped with one of two date-codes: "248" (November 1952) or "308" (February 1953). They arrived in the original Raytheon cardboard "egg carton," on the side of which was printed: "Job 102," "R.T. 6822," "Part# QF721," "Type 43," "Coded 308." I also noticed what appeared to be an inspector's stamp in which a letter "H" was adjoined to the letter "R". What was this "HR"?  

Playing the technological "history detective," I uncovered a single reference to the QF721. In this patent (filed November, 1952) the inventor chose a QF721 pentode for the second audio amplifier/modulator stage of a miniaturized UHF walkie-talkie. The patent drawing shows a crystal-controlled AM transmitter and superegenerative receiver, all built from Raytheon sub-miniature vacuum tubes.

Ah...but look who the patent is assigned to: Hoffman Electronics Corp., Los Angeles, California; the sister of Hoffman Radio Corp., also of Los Angeles. That would explain the "HR" inspector's stamp. What's more, the fellow that sold me these tubes lives on the outskirts of Los Angeles.  

Hoffman Radio began making tube-type consumer radio equipment in 1941. Following the war they went on to produce televisions, tape recorders and even Geiger counters. You may recall they were an early adopter of photovoltaic cell technology. For example, they produced the "solar cells" used in our Vanguard Satellite program. In fact, they produced a solar powered transistor radio as early as 1957; the Hoffman P411.

I suspect that Raytheon made these tubes as a special order for Hoffman Radio. It may be that Hoffman used the QF721's in one or more of their military contracts, and the engineer working on the UHF walkie-talkie design pulled them from the stockroom in order to get on with his project.

Although I haven't located a data-sheet for my QF721's, at least I now have an example of it used by a (presumably) competent designer. For example, the patent drawing shows the QF721 filament in parallel with a Raytheon CK527AX (for which the specifications are available). Thus, I know the QF721 was designed for use with a 1.25v filament supply voltage. At this potential the QF721 filament current measures ~100mA. Comparing these numbers with other T2x3 bulb sub-miniature types in Raytheon's catalog (along with its established usage as an output AF amplifier stage) leads me to believe the maximum cathode current rating would be in the range of 3 to 4mA at a B+ potential of 45 to 60Vdc.

My first project using the QF721 was a 0V0 regenerative receiver for 40m CW using tickler feedback. A potentiometer in the screen supply provided the regeneration level adjustment. I filled a sheet of paper with DX calls in two pleasant evenings listening with this receiver: ZS1JX, 4X4FC, PY2VJ, YO3AAJ, US3IZ, CP4BT and many others. 

I next built a 20m superheterodyne receiver. It begins with a frequency converter made from a single QF721 having two L-C parallel resonators in series with the grid (at 14MHz and 3MHz). Positive feedback via a coupling link on the 3MHz resonator drives it into sustained oscillation in order to provide a tunable VFO. Due to the wide frequency separation, the interaction between the 14MHz input tank circuit and the VFO frequency is insignificant.

An 11MHz L-C resonator in series with the anode couples the IF output to a one-stage crystal filter; a configuration known as the "Telefunken Filter." It was famously used in the Köln E-52 receiver (you may listen to the motorized pre-set frequency gears whir here!). The filter appears to have been popularized in amateur radio circles in the early 60's (at least in Europe) by DJ2KY; thus, it's occasionally referred to as a "KY Filter." My filter uses a single 11.0592MHz "uP" type crystal taken from my junkbox. The opposite sideband rejection is ~25 to 30dB.

I originally connected a crystal-controlled product detector/converter made from a second QF721 to the output of my crystal filter. Again, this is a self-excited mixer. However in this circuit a second 11.0592MHz crystal is configured as a Miller oscillator using the tube screen (G2) as the oscillator input grid. An inductor in series with the crystal allows the frequency to be "rubbered" down to within 800Hz of the crystal filter bandpass frequency. The input IF signal arrives at the the grid (G1).  

Unfortunately, an annoying interaction between the BFO oscillator and the crystal filter made it necessary to include an IF amplifier/buffer stage between the crystal filter and the product detector. Although I'd had in mind a two-tube superhet, these three QF721's sounded so nice together that I decided to relax the minimalism constraint, at least until I had built a matching transmitter.

My transmitter begins with a Raytheon 1V6 (pentode-triode) that uses the same T2x3 glass bulb as the QF721. The triode section is configured as a Pierce, 11.2MHz crystal-controlled oscillator. The pentode section works simultaneously as a 2.8MHz VFO and mixer. This heterodyne exciter tunes from 14.0 to 14.7MHz. The exciter drives a QF721 PA, producing a maximum output of 30mW. Keying is accomplished with a PNP transistor in series with the PA anode and screen 50Vdc supply line. The exciter crystal-controlled oscillator is switched off whilst I'm receiving. It would be better to instead shift the transmit VFO frequency upwards while receiving in order to avoid several seconds of frequency shift until the 1V6 reaches its operating thermal equilibrium.    

Listening to my receiver while working on the transmitter, the RMS Titanic commemorative station, GR100MGY, came to my attention. I learned this station would remain on the air until April 15th; exactly one-hundred years to the day since the great disaster. My desire to work GR100MGY became an incentive to press on with the transmitter.

I judged the setup to be air-worthy on 12 April. In the afternoon I came across GR100MGY working through a pileup. I joined in, and spent at least the next hour calling. At last, there was a lull in the activity and I must have finally turned up alone. The GR100MGY operator replied, "EA1TJ 599." He picked up my correct call on the next and it was over in an instant. Got 'em with 30mW!

Sliding back down the band I heard G3HGE just signing with a state-side station. I quickly sent my call. Silence. I called again, this time with the /QRPP suffix. He came right back with a 429 report. He sent, "UR CHANCING IT WID QRP CNDX NOT SO GUD." He also commented (correctly) on the start-up drift in my transmitter. Alas, another station started in with a CQ and he heard nothing more from me after that.

In my haste I hadn't re-peaked my transmitter before calling him. Without changing any of the settings I measured my output power at 16mW.

I found a nice note from Tom in my inbox the next morning. He wrote

"I was an RAF Wireless operator with Bomber Command and we were trained to listen for weak and watery signals so finding you was about normal for me, I spend a fair bit of time grubbing around in the noise floor of my rcvr looking for signals like yours."

On April 14th I worked: OK6DJ with 29mW (599/559), F8DGY with 28mW (599/559) and SP3GXH with 24mW (599/459). When F8DGY heard about my output power he asked for the details of my setup. He copied everything with no repeats required! In a follow-up email, F8DGY wrote

Hello Mike,
Tnx for this nice QSO with ur QRPp station, Vy interesting but it must take patience! So congrats agn cu next time. Here 800 w 3ele spiderbeam. IC7400. Ur signal was good not much difficulty to hear you!
73 Chris

 The complete station is shown in the upper photo. The lower photo is a close-up of the transmitter. I hope to post the schematic diagrams in the near future.


  1. As always Michael, absolutely fascinating stuff! Thanks for sharing. 73, Casey, TI2/NA7U

  2. Thank you Casey! It's very kind of you to say so.

    Mike, AA1TJ