Message	Date	From	Subject
7977	2012-12-24 08:28:44	davidpnewkirk	A regenerative detector at IF behind a diode-ring doubly balanced mi
Fun experiment with a regenerative detector at IF behind a diode-ring doubly balanced mixer: 1. Rebuilt Kenwood R-599/T-599/TS-520 VFO tunes from 4.9 to 5.5 MHz (J310 oscillator driving two-2N3904 buffer designed by Roy Lewellen, W7EL, followed by sufficient padding to bring its output down to 7 dBm to drive the mixer LO port) to convert the amateur 3.5- and 7-MHz bands to an IF in the vicinity of 1.7 MHz. 2. Homemade diode doubly balanced mixer (two trifilar transformers wound on FT37-43 ferrite cores and a ring of four Radio Shack silicon switching diodes [1N4148/1N914-like) matched for forward resistance with DMM, all ports "50 ohms." Its IF port is terminated with a 6-dB pad followed by a 47-ohm resistor. 3. Switchable 80- and 40-m (3.5 to 4.0 MHz and 7.0 to 7.3 MHz) top-coupled double-tuned circuit filters (50 ohms in and out). 4. High-C 1.7-MHz regenerative detector using a 12SJ7 or 12SK7 pentode, with a "50-ohm" signal input port added by inserting a 0.1-uF capacitor between the ground end of the detector tuned circuit inductor and chassis. The ungrounded end of the 47-ohm load resistor in Item 2 is connected directly to the junction of the inductor and 0.1-uF capacitor. 5. Vacuum-tube audio stages, including switchable passive low-pass filters. 6. Low 50-foot doublet antenna, fed with ladder line via a homemade balanced tuner. 7. At least 50 feet of RG-58 coax between the tuner at Item 6 and the input of the switchable filters at Item 3. Even though the signal path from the antenna to the detector includes ladder line tuner (up to perhaps 0.5 dB of loss) 50+ feet of RG-58 coax (will add a dB or two of loss, more at 40 m than 80 m) top-coupled filter (loss, perhaps 1 dB) homemade diode DBM (loss, perhaps 7 dB) 6-dB pad (loss, 6 dB) the regenerative detector at 1.7 MHz handily hears down to the band noise on both bands (3.5 and 7 MHz); that is, disconnecting the antenna makes the receiver quieter. The noise floor of the antenna system clearly overrides that of the receiver. I included a 6-dB pad (and its 47-ohm load) between the mixer and detector mainly because I wanted to forestall disturbance of the mixer by the signal from the oscillating detector. (I have no means of measuring the intermodulation distortion dynamic range of the system, but theory and experience tells us that sufficient incident signal at the IF port of a diode mixer will affect its IMD performance. [In superhet practice, one would add a diplexer and/or low-noise, strong-signal-capable amplifier between mixer IF port and subsequent stages to keep it termination-happy.]) Best regards, Dave amateur radio W9VES
7978	2012-12-24 12:02:17	kb1gmx	Re: A regenerative detector at IF behind a diode-ring doubly balance

7979	2012-12-24 14:56:28	KK7B	Re: A regenerative detector at IF behind a diode-ring doubly balance
Hi Dave, Interesting new-old receiver topology. Thanks for all the descriptive numbers. A conversion stage in front of the regen solves many of the known regen issues, and I'm not at all surprised at the ability to hear the band noise floor. Will you have it on with a transmitter for SKN? An experiment it might be interesting to try is to use the 1.7 MHz regen stage as a very high-Q tuned circuit and follow that with a second diode ring to downconvert the 1.7 MHz IF to audio--perhaps driven by a 6C4 Hartley oscillator with a link on the tank circuit to drive the diodes. You might need some additional isolation between the regen tuned circuit and the diode ring detector though...perhaps a 12AT7 with one triode as an oscillator and the second as a common grid isolating IF stage with little gain. That would give you independent selectivity and bfo offset for single-signal CW reception, passband tuning, etc. A 6BE6 could also be used in several places--or a 1L6 if you want to keep the current drain down to battery portable levels. I've sketched a battery portable all tube superhet receiver that draws less current than the solid state receiver Wes used on Field Day the last time I hiked into the hills with him. A backpack portable all hollow state Field Day rig is entirely possible. It's really quite easy to design an all-tube portable receiver with respectable performance with lower total battery drain than an FT-817, and I think those have been used successfully for Field Day. A useful number to bear in mind is the battery life of a 1950 Zenith Transoceanic--200 hours with carbon zinc cells. Field Day is only a quarter of that if you leave the receiver
7985	2012-12-25 14:25:45	kb1gmx	Re: A regenerative detector at IF behind a diode-ring doubly balance

7991	2012-12-26 17:52:56	davidpnewkirk	Re: A regenerative detector at IF behind a diode-ring doubly balance

Fun experiment with a regenerative detector at IF behind a diode-ring doubly balanced mixer:

1. Rebuilt Kenwood R-599/T-599/TS-520 VFO tunes from 4.9 to 5.5 MHz (J310 oscillator driving two-2N3904 buffer designed by Roy Lewellen, W7EL, followed by sufficient padding to bring its output down to 7 dBm to drive the mixer LO port) to convert the amateur 3.5- and 7-MHz bands to an IF in the vicinity of 1.7 MHz.

2. Homemade diode doubly balanced mixer (two trifilar transformers wound on FT37-43 ferrite cores and a ring of four Radio Shack silicon switching diodes [1N4148/1N914-like) matched for forward resistance with DMM, all ports "50 ohms." Its IF port is terminated with a 6-dB pad followed by a 47-ohm resistor.

3. Switchable 80- and 40-m (3.5 to 4.0 MHz and 7.0 to 7.3 MHz) top-coupled double-tuned circuit filters (50 ohms in and out).

4. High-C 1.7-MHz regenerative detector using a 12SJ7 or 12SK7 pentode, with a "50-ohm" signal input port added by inserting a 0.1-uF capacitor between the ground end of the detector tuned circuit inductor and chassis. The ungrounded end of the 47-ohm load resistor in Item 2 is connected directly to the junction of the inductor and 0.1-uF capacitor.

5. Vacuum-tube audio stages, including switchable passive low-pass filters.

6. Low 50-foot doublet antenna, fed with ladder line via a homemade balanced tuner.

7. At least 50 feet of RG-58 coax between the tuner at Item 6 and the input of the switchable filters at Item 3.

Even though the signal path from the antenna to the detector includes

ladder line
tuner (up to perhaps 0.5 dB of loss)
50+ feet of RG-58 coax (will add a dB or two of loss, more at 40 m than 80 m)
top-coupled filter (loss, perhaps 1 dB)
homemade diode DBM (loss, perhaps 7 dB)
6-dB pad (loss, 6 dB)

the regenerative detector at 1.7 MHz handily hears down to the band noise on both bands (3.5 and 7 MHz); that is, disconnecting the antenna makes the receiver quieter. The noise floor of the antenna system clearly overrides that of the receiver.

I included a 6-dB pad (and its 47-ohm load) between the mixer and detector mainly because I wanted to forestall disturbance of the mixer by the signal from the oscillating detector. (I have no means of measuring the intermodulation distortion dynamic range of the system, but theory and experience tells us that sufficient incident signal at the IF port of a diode mixer will affect its IMD performance. [In superhet practice, one would add a diplexer and/or low-noise,
strong-signal-capable amplifier between mixer IF port and subsequent stages to keep it termination-happy.])

Best regards,

Dave
amateur radio W9VES