Monday, January 03, 2005

Regen Rcvr - 1

My hobby as well as my business is electronics. I'm in the final stages of a project I have been thinking about for 47 years. A regenerative radio. My original design (the design was a school project) was a one tube job with a tickler coil run on batteries.

The latest version (which is actually being built) consists of an RF stage (done - gain of 12 measured), followed by a Colpitts type regenerative stage (a short discussion of the Colpitts Oscilator), and an infinite impedance type detector. The audio is handled by an LM 386 and two minature oval speakers wired in series. The audio stage has been tested (a few minutes ago)and while not disco in volume will hurt your ears up close the response goes to about zero below 100 Hz and above 7 kHz. Good for a short wave reciever.

A schematic which is in part similar to my final design. I have changed some part values (the input cap fer instance is now 270 pF vs 47 pF). Plus I have a heftier audio out stage (the LM386). I have already added input jacks (3.5 mm and 1/4") so I can use it as an audio amp and later I will be adding output jacks for ext speakers or headphones.

The frequency band is going to be a few hundred kHz (or more) centered around WWV at 10 MHz. I have this nice old 75pF variable as the main tuning and a pair of back to back 1N4148's used as varactors (semiconductor capacitors that vary with voltage) for fine tuning. I'm using a 10.7 MHz IF transformer as the interstage coupling (L1 in the schematic at the above url) and tuning coil.

Back when I designed the original tube job adding more tubes for audio power or an RF stage was not very common. Tubes were hot (burn your fingers hot), used high voltages ( 67 1/2 volts and up - 90 volts was very common for portable equipment), and required a watt or two of heater (filament) power for each tube and another watt or so of plate power. Plus they were kind of bulky. The fewer the better. Vacuum tubes, or as my friend Brad Boyer liked to say: "space bottles".

Now my input RF stage (PN2222) uses about 55 milliwatts worst case. (I should rebias it because with a 12 V supply the collector sits at 1.56 volts - but it amplifies well enough in the voltage ranges I expect ( 1 micro volt to 100 millivolts ) that I'm not going to fool with it 'til later.

At full crank (maximum undistorted audio output) the whole radio ought to draw about two to two and a half watts. Average power should be in the range of 1/2 watt or less.


Here is a regen reciever of the old style with a tickler coil/Armstrong detector. it uses a LM386 for the audio power amp.

Here is a one tube regenerative reciever. and a report by some one who has had such a reciever on his mind for a few decades.

A nice picture of a 1920s commercial regen reciever.

Another nice picture of a 1920s commercial set. It is for sale $650. Way out of my price range. The set I'm building would cost about $30 - $40 for parts if everything was bought new.

Here is the front panel of a modern regen radio.

Another Armstrong type detector.

Here is a Hartley feedback regen schematic. The schematic is poor but you can get the general idea. Construction details here. Here is the original ARRL article (PDF) by Kitchin. The design shown covers around 7 MHz (40 Meters) and with a switch 3.5 MHz (80 Meters)

Another old time commercial regen set.

More neat pictures and a book on building a regen. I haven't read the book so I can't comment on the quality. The review looks good (don't they always?).

Here is a regenerative radio with a design (electronically) very similar to the one I designed 40+ years ago.

This is a two tube job (audio amplification) again with a regen stage similar to my 40+ year old design.

Here is a page that has a nice circuit for a ham band (about 7.0 to 7.5 MHz) regen. He also explains from an operational view point all the pluses and minuses of a regen. He also has an interesting one transistor design. It uses the tuning coil as a loop antenna. Fascinating.

This is a schematic of a sort of super hetrodyne reciever. It is a crystal controlled converter followed by a 3.5 to 4.0 MHz tuneable regenerative IF. Here is an explanation of how the circuit works along with construction details. It is a three tube job. It could easily be transistorized.


Reported 06:59z 04 Jan 2005.

Well I finally fired up the regen stage - nothing. So I'm thinking what could go wrong? Tis is supposed to be almost fool proof. Well sit back. Have a smoke - THINK. What could it be? The resistors are all 1/4 W jobs with color codes. They are correct. The Colpitts capacitors? Hmmm. 39 pF going to the JFET detector. Fine. 390 pF to the base of the Colpitts regen. Fine. 39,000 from the emitter to ground. BINGO. Replace it with a 390 pF. Oscillation. Outstanding.

Now there is a slight motor boat (low frequency oscillation). Hmmm. There are a few places where some more bypassing might do some good. I'll tack on some caps and report back later.

Updated 11:07z 04 Jan 2005

I tacked 3 or 4 .1 uF monolithic ceramics in likely locations and the motor boating stopped. I also cut the built in capacitor out of the IF can to increase the tuning rnge. I'm going to get the fine tuning pot attached to the board and then I'll solder the main tuning cap to the board and it will be done. A solid mechanical mount for the board will complete the process.

Updated 09:03 06 Jan 2005

I have connected the main tuning capacitor to the board. The reciever now tunes from about 8.7 MHz to 10.4 MHz. I could tweak this a little by adjusting the tuning coil but this is close enough to what I want - Coverage from 9.5 to 10.3 MHz which includes the short wave bands from 9.25 to 9.999 MHz. WWV at 10.000 MHz. And the amateur band from 10.100 to 10.150 MHz.

The fine tuning works but is a little too fine. I think I will mod it by adding diodes in parallel to the ones already installed.

I did change the collector to base resistor from 1 Meg ohm to 2.2 Meg. That raised the collector voltage to about 6 volts out of 9. That should allow signals about 3X those the stage could originally handle.

My first station recieved was a Spanish langusage station on 9.72 MHz (roughly). I did it with three jumper clips connected together. I'm going to add a good ground and some kind of indoor antenna (10 - 15 ft) and see what I can get.

Update: 05:17z 07 Jan 2005

I tacked on another pair of 1N4148 diodes to the pair already on the board and the fine tuning is still too fine. So I decided it was time for heavier ammunition. A 1N4002 rectifier diode ought to have more capacitance as it is rated for 10X the current that a 1N4148 can handle. More current = bigger junction area = more capacitance.

Did it work? Yep. Fine tuning now covers about 100 KHz. I could reduce it some by reducing the voltage to the tuning control but it works OK so I'm leaving it as is.

I have buttoned up the case. Time for a new project.

WWV only comes in well during the day. Even then it fades in and out. Which is to be expected in the short wave bands.

My 13 year daugghter played with it today. She thought it was fun. She likes the more modern jobs with digital dials better.

My first mate keeps asking if I can get Europe. I tried to explain that there has to be a European station transmitting on a frequency I can recieve. This did not satisfy her. She said that I needed to find a short wave listing on the net and get some Europe. She must be right. (Tact again).


Dave Halliday has a nice review of this project.


BTW. Any one interested in getting help on a radio project - leave a comment or send an e-mail.

I'll add to this as I progress. Probably in new posts once this moves down the page.

1 comment:

Anonymous said...

The best thing about building a tube radio for a school project - you are guaranteed to get a B+.

triticale - the wheat / rye guy