When I had trouble getting the regen receiver in W2UW's ET-1 circuit to work, I turned to the internet and -- through AA7EE's site -- found the circuit of Doug N0WVA. This circuit has completely changed my attitude toward regen receivers. I have been exchanging e-mails with Doug -- below is a compilation of the info and regen-wisdom that he has shared. More to follow... Thanks Doug.
I came up with the diode after exploring ways to ditch the source r/c combo. The thinking was the closer I could get the source to ground the less voltage/capacitance fluctuations the gate would see. Also I hated seeing everyone using .01 bypass to avoid audio oscillations and also losing audio gain.
The green LED works good but even better is directly grounding the source. Then feed a small negative bias through the gate leak resistor , adjustable via a potentiometer.
On video, the audio is taken straight from the radio shack headphones that are connected to the audio transformer. The headphones are held directly to the phones case ( no hole for the microphone seen on the phone)
The variometer is made with I think a 1.25 inch pill bottle and the tickler inside is around an inch in diameter pill bottle. I used a pharmaceutical syringe's outside tube for a shaft. The tickler form has a couple holes cut for the shaft to pass through, it is a friction fit, more like slits cut and the rod pushed through. I used the soldering iron to melt round holes on the actual outside coil form for the shaft to turn on. On the back of the shaft is a small screw that goes through the outside coil form and screws into the syringe center hole that holds it in place. The tickler is one turn, I think, and routed through the inside of the shaft via small holes melted with the soldering iron.
A couple tweeks to mention is instead of a resistor in the gate, use a choke for less noise, makes a big difference, especially if you listen to AM. Also I have been using a gimmick for the gate cap. Just maybe a #36 enamel wire wrapped around the hot tank lead 5 or 6 times and then I remove turns till the thing stops oscillating, then add a turn. This helps cut down even more on strong signal pulling.
I have always been on a quest for more performance out of the least parts. This design was about as far as it could go, I think....
I have never done any real sensitivity tests on the regen, so you have gone farther than me already. One thing was noticed though is the gate resistor does add a lot of noise, especially noticeable just under oscillation in AM detection mode. So I took a one meg 1/4 watt resistor and wound as many turns of #38 wire on as I could, probably around 80 turns, then subbed it out for the gate leak. This dramatically improved the noise level just under oscillation. This was with a simple antenna band noise test. I think it also improved the noise under oscillating conditions.
Adding extra antenna coupling will probably help a lot, but, there is a point where we start getting too much strong signal pulling. The strong external bias battery trick will also improve this, although at the cost of extra parts.
Posts
Interesting, but I think given that you posted additional mods from Doug, you should also include a link to or posting of the schematic he is referring to
ReplyDeletehttps://forums.qrz.com/index.php?attachments/scan006-jpg.121895/
This schematic shows how he grounded the source, added the bias to the grid leak and used the wire in place of the small cap feeding the gate.
maybe a nice low noise op amp could replace the 2n2222a and drive headphones w/o needing an output transformer. Of course that would have lots of transistors and be considered blasphemy...lol
ReplyDeleteElectronics Illustrated in the 1960s had a 2-transistor transmitter using 2 high voltage transistors in parallel for 17 watts input at 175 volts. You could use one such transistor and lower the voltage slightly for safety and still have useful output power. The output circuit was the usual tube type for high impedance, not the usual transistor LP filter. If harmonics were too high, you could still use high impedance output circuit, just use 2 tuned circuits.
ReplyDeleteI have the "QST Novice Pair" built from their Novice License book and i am thinking i will modify it for transistors using these circuits.
Why is the spec asking for a Green LED? Is it based on it's forward voltage? will another color work? or what is the required forward voltage so I can make sure it is to spec?
ReplyDeleteThanks