N0WVA’s One-FET Regen Optimized For SSB/CW Sounds Great!

Since I began posting about regens, it has been gratifying to hear from fellow regen builders, and quite interesting and inspiring to communicate with other hams who are also taken by their charms. The Sproutie has proved to be a great little receiver and it’s a pretty straightforward circuit, but it’s not exactly a minimalist design. Some hams are so intrigued by the regen’s ability to deliver an impressive amount of gain and selectivity in just one stage, that they focus on building a station with an absolute minimum number of parts. Doug N0WVA has the goal of building a complete (and usable) station with just 2 active devices, and he’s halfway there, having come up with a regen receiver that works well on SSB and CW, contains just one active device, and has a total of just 9 parts including the coil. Better still – it works well. Doug writes, “I’ve been building regens since the 90’s and always wondered why the tube versions always ran circles around the solid state rigs that I had been building. Seems like it always took two transistors to equal what one triode could do as far as recovered audio. Well, I think I might have figured it out. You really can get all that and more from just one FET. Actually, I’d say now that a two tube Doerle will not be able to touch this one transistor rig for serviceability. My intentions are to have only two active devices for an entire CW station. It’s not difficult to get a half watt or so from one device, and even be able to slide around the band a bit via VXO. So the main problem was a decent receiver. Well, I now have that problem solved. All the SS (solid state) regens I had been building had a source resistor and cap. This is to supply self negative bias to the gate. It works, but audio suffers because of the use of a small bypass capacitance at audio frequencies. Larger electrolytics here cause howl when oscillating. By taking the source to ground you can then inject the negative bias through the 1 meg gate leak and set your optimum operating point with a 10k pot. However, this adds to the total parts count. I found that by using a green LED in the source worked almost as well, and bypassed all the audio to ground as well.  Also, since the receiver was to be optimized for CW, I wanted minimal pulling, even on strong stations. To do this we must keep the tank coupling very light. That way the capacitive changes in the junction of the FET has minimal effect on the tank. Finally, another plus for SS regens is the need for much less feedback to achieve regeneration. A properly built SS regen should need no more than 1 turn for a tickler. Much more than this and you get wild instability and a heavier loaded tank. One experiment I did was to try different coil forms and see which ones took less tickler feedback. I found that pill bottles were about the best form for low loss, as I could use just one turn spaced well away from the tank. Even better was a pill bottle with slots cut out of it for less loss, but the wire would try to crush the form. I am using a 1k to 8ohm transformer from Radio Shack to couple into Radio Shack phones, which are a bit more sensitive than other cheap headphones I have used. This results in audio that is loud enough to be highly usable, probably where you would normally run the volume on a regular receiver. The dial drive is a ball bearing vernier integrated into a gear reduction capacitor. The NPO trimmer is a band set. All capacitance is kept to a minimum. I find the less plates on a variable capacitor the better for keeping drift down. The regen control is actually a homemade tickler variometer inside the tank form.  This eliminates the need for a choke and throttle capacitor. I will attach some photos and schematic. Hopefully you will find them interesting. I still need to get the transmitter together and get this on the air. Having just two transistors for a complete CW station will be a blast I think.” Doug made the shaft of his variometer from the plastic barrel of a syringe. I have a number of syringe/pipettes here that I use for giving medicines to my cats and am thinking one of them would be useful for this purpose. Of course, anything cylindrical could be used for a shaft. Doug’s schematic was a fairly low-res scan, so I redrew it. Here’s my re-draw of the schematic of Doug’s one-FET regen –

Schematic of N0WVA’s single-FET regen. It is a monobander on 75/80M. Note: Barry W6YE looked closely at the bandset trimcap in the photos below (the one marked in the schematic above as being 10pF) and discovered, using the part number visible in the picture, that it is a 7-45pF part. Thank you for this information Barry!

I re-drew Doug’s schematic, as the one he sent me was a little small. I figured my re-draw might be a bit easier to read. However, for the sake of completeness, here is what Doug sent me –

I don’t think Doug mentioned the type of FET he was using, so I assume the usual suspects (J310, MPF102) would be fine here. He also doesn’t include coil winding details on the schematic but you should be able to get a rough idea from the photos and besides – if you’re hardy enough to give this receiver a go, you’ll want to figure out the exact values of inductance and capacitance for yourself. It’s good for the soul  🙂 These two online calculators should help – Resonant Frequency Calculator Coil Inductance Calculator (NOTE – Doug provides more details in the comments section at the end of this post) It occurred to me that a good pair of balanced armature headphones (the type referred to as sound-powered headphones and liked by crystal set enthusiasts) could work well. He replied that he had tried a pair of 2000 ohm headphones and they were rather loud, so he would probably need to incorporate a volume control if he used them. Just imagine that – a pair of headphones being too loud when the only thing separating them from the antenna is a single transistor – and with ham signals too, as opposed to big broadcast signals! Here are some pictures of what Doug’s single-FET regen looks like –

N0WVA’s single-FET regen. View from above. Photo courtesy of N0WVA.

In this side-view, you can see Doug’s home-made variometer tickler, with the syringe body that he used for the control shaft –

Side-view of N0WVA’s one-FET regen, showing the home-made variometer tickler, and control shaft. Photo courtesy of N0WVA.
N0WVA’s one-FET regen. Photo courtesy of N0WVA.
N0WVA’s one-FET regen. Photo courtesy of N0WVA.
View from the other side of N0WVA’s one-FET regen. Photo courtesy of N0WVA.
The front of N0WVA’s one-FET regen, along with the RS headphones that give good sensitivity. Photo courtesy of N0WVA.

Doug made a video of his receiver in action. He says it was a little difficult, as he had to use one hand to hold the headphone to the microphone for recording, while tuning with the other hand. If you can, while watching this, have another window open so you can see the schematic and remind yourself that the stable and well-resolved SSB signals you are hearing are coming from that simple circuit. Great stuff –

I can’t wait to see the 1-transistor transmitter he pairs this great little receiver up with. Thank you Doug, for allowing me to share this info on your regen receiver. This is what ham radio is all about.

25 thoughts on “N0WVA’s One-FET Regen Optimized For SSB/CW Sounds Great!

  1. Wow, cool having my receiver on your blog! Thanks Dave!
    A quick description on the tank circuit. I am using 31 turns of #26 enamel on a 1.5 inch pill bottle. The trimmer capacitor is I believe 7-45pf and Im not sure about the actual tuning capacitor. Maybe around 30pf max?
    And yes the JFET is a J310. I did not try an MPF102.

    1. Thanks for the details Doug! My friend Joel KB6QVI watched your video and was really impressed. It’s amazing that so few parts can produce that level of performance,

      Dave
      AA7EE

    2. I saw an update to this circuit. It it is true someone corresponded with you Doug to improve on the design. Here is the thread:

      https://soldersmoke.blogspot.com/2019/11/notes-and-video-on-doug-n0wvas-amazing.html

      And the new schematic:

      https://forums.qrz.com/index.php?attachments/scan006-jpg.121895/

      I did not have good luck with the regen control (popped in and out with hysteresis). However I was using a supposedly equivalent NTE312. I saw in this updated schematic that you switched to a throttle capacitor instead of the vari-coupler, and that you had dropped the LED for a “C” battery on the potentiometer. I was thinking of measuring said gate voltage once a nice bias was found and use a diode (or a couple) that had the same drop. Not sure where it is biased if it is near cutoff, or in the linear region of the device. I had a ready made vari-coupler in the junk box, but it had 4 turns instead of one. That would explain the sharp transition to oscillation, but not the hysteresis of where oscillation starts and when it drops back out after backing off the controls.

  2. Dave,

    Great circuit! I will try it later.

    It looks like you left out of the schematic a capacitor that by-passes RF from the junction of the variometer coil and the audio transformer to ground. I would use a 1 nF to 10 nF ceramic cap.

    Thang Le
    AA6SV

    1. You are absolutely right – thank you very much for taking the trouble to point that out. I corrected the schematic.

      Dave
      AA7EE

  3. Dave, Isn’t there a ground connection symbol missing at the LED cathode end on the schematic? Love your blog – it’s always fascinating to see what you’re up to with regens. It seems that one type is better at AM and another at CW or SSB. I understand that has to do with tank loading generally, but, have you found a topology that is reasonably good at all modes?

    1. Ooops, you’re right Paul. I will add that later today.

      I have not experimented very much at all with different regen topologies though once I realized that a good way to operate a regen was to adopt the practice of leaving the AF gain at, or near, full gain, and using the RF gain control to adjust volume, SSB reception improved greatly. Previously, I had been doing it the other way around i.e. leaving the RF gain set fairly high and using the AF gain to adjust the volume. This is the way most of us operate our modern superhet receivers and for these modern rigs, equipped with product detectors, it’s a fine way to do things. For anyone who has used an older receiver with a diode detector that was primarily designed for AM reception, they will remember that when they needed to receive SSB or CW, they switched the BFO on, left the AF gain turned up high, and used the RF gain to control volume. This was standard practice with older receivers, and it’s a good operating practice with a regen. It’s why I have come to appreciate audio stages that are relatively low-noise.

      If you look at the first comment on this post, qrp-gaiijin neatly summarizes how to design a regen to minimize frequency pulling –

      https://aa7ee.wordpress.com/2015/01/07/does-the-sproutie-regen-work-on-ssb-and-cw/#comments

      He also provides links to discussions of a design that automatically reduces the gain on strong signals (which would help with pulling).

      I used to think that pulling didn’t happen when receiving AM. It does, but much of the time you don’t hear the effects as readily as you do with SSB. However, if you adjust the regen control very, very close to the critical point of oscillation, and have the RF gain up fairly high, then if your received signal has fading, you’ll hear the signal move slightly out of, and back into, the relatively sharp “nose” of the passband as the signal fades in and out. Once I realized that this effect was noticeable with AM reception too, I always keep the AF gain up fairly high and ride the RF gain control, regardless of what mode I am listening to. It’s just a good operating practice to adopt with a regen.

      PS – I hope to be posting something about my next regen in a few months. The front end will be almost the same as the one in The Sproutie. There is a small mod designed to provide a greater audio level to the AF pre-amp, and there will most likely be slightly lighter coupling between the tank and the Q-multiplier. I’m not sure if that will help any, as the detector will still be coupled into the top of the tank. I’m pretty happy with the performance of this circuit, so decided to stick with it for the time being. The new regen will have a different, and slightly more complex, audio stage though. I’m planning to use a series of switched active filters on separate boards using the trusty (and low-noise) NE5532 for different AF bandwidths. I just built a 4-stage filter with a 3dB cut-off of 2.4KHz, which is 80dB down at 5KHz, and am keen to see how SSB sounds through it. One of the issues I have with comfortably listening to SSB on a regen for long periods is the rushing/hissing sound, due to the fact that the circuit is oscillating. I’m thinking that a filter with a nice sharp cut-off can help reduce that, and am keen to see how it sounds. A nice narrow filter for CW would be really cool too.

      Details to follow but as I work extremely slowly, with lots of long breaks, don’t hold your breath!

      73 for now,

      Dave
      AA7EE

    1. Thanks Doug – I just made the change. I’m glad your video is still available to view. People need to see how well a regen can handle SSB and CW when optimized for those modes,

      73,

      Dave
      AA7EE

  4. I built a two-band version using an inductor/variometer from an old TU-9-B antenna tuning unit and a 25-250 pF. varicap. It covers the range 2.5 to 8 mHz. I changed the gate cap to 100 pF., as I intend to use it only for AM and SSB signals. For a minimalist set, it works well.

    73, Bruce, VE3EAR

    1. That’s great Bruce. I think this might be the only schematic I have published on my blog that I haven’t actually built. Still mean to do so some day. Good to know that yours works well.

      73,

      Dave
      AA7EE

  5. Electronics Illustrated magazine had a 17-watt transmitter that looked kind of like the Ameco, but used two HV transistors in parallel at 200 VDC. One transistor would give you an 8 watt transmitter in a single device, and with the more conventional old-time output circuit rather than the low-Z tbandpass filter that power FETs all use. -Hue K7HUE

    1. W8DIZ has NPO’s with values as low as 10pF. You could put a few of those in series. Alternatively, Mouser has 3 and 3.3pF MLCC (Multi-Layer Ceramic Capacitors) that are C0G (zero temp co-efficient). Just go to Mouser, and look for C0G (NPO) MLCC capacitors. They have lots of them in all sorts of values. Here’s one. There are plenty more though, so do perform a search –

      https://www.mouser.co.za/ProductDetail/Murata-Electronics/RDE5C1H3R0C0P1H03B?qs=IAJIiSVA5G%252BsYS9CjtSRDA%3D%3D

      Here are Diz’ NPO caps –

      https://www.kitsandparts.com/caps_npo.php

  6. I made a SKCC contact with KI4POV. He successfully built the N0WVA One-Regen and he has pictures on his QRZ webpage. This lead me to surf the net and find this blog.

    1. Thanks for the tip Bob. I went to his QRZ page to look. It looks great! I haven’t built anything in a few years, but one of these days, I’m going to build this little receiver. It’s on the list of things to do. I have a spare HRO dial and drive, and am thinking it would be great to build it with really high quality vintage and sturdy components. My idea is that it would be mechanically quite solid and large, yet electrically simple. One of these days, maybe…………..!

  7. Thanks for the response Dave. I am gathering the parts for this project. I found a green LED from a garage door sensor I replaced a few years back and will give it a try. BTW, your home brew projects are most professional!

    73’s Bob K5MBA

  8. Hello Dave
    A delightfully simple little reciever, er, but I havn’t been able to get mine to work. I used a J310 fet, made a variometer from a pill jar and a piece of plumbing pipe with a one turn tickler. LT 700 transformer, walkman type ear piece 10pf coupling capacitot between tuned cct and the gate of the fet. 22pf air space variable for tuning. Connnected long wire ariel via three turn coupling coil. All conections on a 2 amp chocolate block.
    Somewhat quiet, er silent. With variometer in the right setting, the green led glows a little bit brighter and a good oscilation can be heard on 7mcs. There is a very slight shushing as the variometer takes the cct in and out of oscilation. Touching a finger on the gate of the fet produces a good audio buzz apart from that the set is silent.
    Any ideas. Anyone ?
    Regards Chris

    1. I have not built this set and have no suggestions, I’m afraid Chris. Out of interest, I just included Doug’s original schematic in this post, underneath my re-drawing of the circuit diagram he sent me. I was wondering if your feedback winding is connected with the correct polarity, but you mentioned that you are getting oscillation, so that part of the circuit must be in order.

      I hope you get your problems with this little receiver resolved. Can you let me know when you do please?

      Dave
      AA7EE

Leave a comment