When we build an electronic project in 2016, the chances are that the active components will be integrated circuits containing an extremely large amount of functionality in a small space. Where once we might have used an op-amp or two, a 555 timer, or a logic gate, it’s ever more common to use a microcontroller or even an IC that though it presents an analog face to the world does all its internal work in the digital domain.
There was a time when active components such as tubes or transistors were likely to be significantly expensive, and integrated circuits, if they even existed, were out of the reach of most constructors. In those days people still used electronics to do a lot of the same jobs we do today, but they relied on extremely clever circuitry rather than the brute force of a do-anything super-component. It was not uncommon to see circuits with only a few transistors or tubes that exploited all the capabilities of the devices to deliver something well beyond that which you might expect.
One of the first electronic projects I worked on was just such a circuit. It came courtesy of a children’s book, one of the Ladybird series that will be familiar to British people of a Certain Age: [George Dobbs, G3RJV]’s Making A Transistor Radio. This book built the reader up through a series of steps to a fully-functional 3-transistor Medium Wave (AM) radio with a small loudspeaker.
Two of the transistors formed the project’s audio amplifier, leaving the radio part to just one device. How on earth could a single transistor form the heart of a radio receiver with enough sensitivity and selectivity to be useful, you ask? The answer lies in an extremely clever circuit: the regenerative detector. A small amount of positive feedback is applied to an amplifier that has a tuned circuit in its path, and the effect is to both increase its gain and narrow its bandwidth. It’s still not the highest performance receiver in the world, but it’s astoundingly simple and in the early years of the 20th century it offered a huge improvement over the much simpler tuned radio frequency (TRF) receivers that were the order of the day.
![Armstrong's regenerative receiver circuit. Chetvorno [CC0], via Wikimedia Commons.](https://hackaday.com/wp-content/uploads/2017/01/armstrong_regenerative_receiver_circuit-svg.png?w=400)
The simplicity of a regenerative receiver did not come without problems though. The coupling adjustment became a small variable capacitor in later designs, and this could be found as a regeneration control on the front panel of a typical receiver. At every retune to a different station this would require readjustment for best performance, resulting in tuning a regenerative radio becoming something of a black art. In addition, if poorly adjusted they could sometimes oscillate and become transmitters in their own right. When the more complex but superior superhetrodyne receivers (another Armstrong invention) arrived around a decade later the popularity of regenerative receivers went into decline, and they had almost entirely disappeared by the end of the 1930s. Today they survive in niches such as amateur radio, toy walkie-talkies, toy electronics kits, and unexpectedly in very cheap UHF remote control modules.
It is this last application that points to one of the regenerative detector’s useful features. While most regenerative receivers are designed for AM broadcasts, the principle works at almost any frequency. It is possible to simply construct receivers using the principle that extend well into the UHF spectrum, and though they aren’t the best receivers on the block they can surprise you with their performance. [Roger Lapthorn, G3XBM] for example has published simple designs for a range of transceivers for the VHF bands with regenerative receivers, including the rather minimalist 2 metre (144MHz) “Fredbox”.
The regenerative receiver may not be the most advanced receiver ever conceived, and it certainly isn’t the most sensitive. But it’s one of those circuits that everyone should consider trying once, for its simplicity and ingenuity, and because it delivers results for relatively little effort. Go on, have one on your bench!
[Header image 1920s regenerative receiver, Charles William Taussig [Public domain], via Wikimedia Commons]
Oh I don’t know. Some of those “super-components” are doing some pretty clever things behind the scenes.
Made a VLF receiver with a pentode triode combo tube in 1963. Tuned frequency from 11 to 33 KHz. Tied in sensitive relay to receive code from military transmitters.
The link to the internet archive up there under the Ladybird cover has the full text of the book, along with the equally wonderful “Magnets and Electricity” and “Magnets, Bulbs and Batteries”.
https://archive.org/search.php?query=Ladybird
Anyone building a regenerative receiver should consider putting an RF amplifier stage ahead of the detector, so it doesn’t transmit via the antenna if the regen is turned up too high. There was a time when the AM broadcast band was so full of signals from poorly adjusted regen receivers that it made finding the actual stations hard!
This was a big issue back in the day, when receiver tube would dissipate few Watts at anode. Today with milliWatt power levels in RF transistors, it is much less of an issue.
Let’s not forget that most MW superhets have local oscillator as a first stage. They radiate as well.
I still agree with you, just nitpicking.
Military receivers designed for and before WWII had more than one RF amplifier stage ahead of the local oscillator, with an RF gain control… This was to minimize the radiation of the oscillator… As many found out over the years – you can send a nasty present back down a radio, RADAR, or SONAR beam…
Not advisable, a pre-amplifier necessarily will alter the entire design. Each stage including the regenerative one must then be fully shielded and separated. Also, This is not an issue in transistorized designs.
No, it won’t alter the design. It actually enhances it. And you don’t need to shield everything (you DO need to keep the coil at least one inch away from large metal objects). There are three good reasons for adding a pre-amp.
1. It prevents the antenna from radiating a signal when the radio is operating PAST the critical point and is oscillating which is necessary to receive CW and SSB.
2. When properly designed, the pre-amp really isn’t a pre-amp. It’s more like an attenuator. Strong nearby signals can block the receiver (or rather LOCK on the the strong signal and thus block all other signals from being received).
3. Without a pre-amp the antenna is a part of the tuning circuit. As a dipole sways in the wind it often de-tunes the signal. This might not be a problem with AM signals but it’s definitely noticeable on CW and SSB, enough to make the signal un-intelligible.
Thank you for reminding me that I need to sell off my Kennedy Type 110 and Grebe CR-12 regen radios.
Regenerative receivers are good but the real magic is in the super-regenerative receiver (which really should have an RF buffer amplifier in front of it).
While regenerative receiver might oscillate if you are doing it wrong, super-regenerative must oscillate if you are doing it right!
Superregen is more sensitive, but it has wider receive bandwidth (at least: quench frequency * 2). Furthermore, it is not good for CW reception. Therefore it is mostly used at VHF and UHF.
For technical reasons superregens are practical pretty much only for VHF (broadcast FM and aircraft-to-ground AM) and similar UHF transmissions. For mediumwave and shortwave reception (“AM radio,” international broadcasters, shipboard and ham radio operators using Morse code and single-sideband voice), a _regenerative_ receiver is the thing. I routinely use regenerative-detector-based receivers of modern design for two-way Morse code ham-radio communication, and they work just fine.
No mention of the ZN414? Or am I too young…
Howard Armstrong didn’t invent the ZN414.
Besides, it’s a TRF receiver, Tuned Radio Frequency, which was a step up from a “crystal radio”, and was the trend except for Armstrong’s receivers.
Back then, you’d string a bunch of amplifier stages in sequence, and each would have a tuned circuit. Lots of knobs, until someone thought up the notion of ganging the variable capacitors together so only one knob needed. If things were done right, you could end up with a regenerative receiver due to feedback.
The ZN414 was a modern equivalent. A single tuned circuit, but lots if gain, and since there was AGC, automatic gin control, the illusion of better selectivity because strong signals caused the IC to reduce gain, making it harder for wek signals to get through.
Michael
I could do with an automatic gin control.
Everywhere I used a ZN414, a BC108 worked just as well.
What should we use for OC71 and OC45 nowadays? None of the listed transistors seem easily available.
2N3904, put the PSU the other way up ‘cos it’s NPN. That’s where I’d start.
Speaking of this issue… does anyone know of any good online transistor substituting guides?
I have a couple of decades old books that I use for this. It seems like there ought to be something better than that on the net somewhere by now. I tried Googling it a few times but didn’t find anything. I even tried to OCR my books but it never came out usable.
All modern small signal silicon bipolar transistors (BJTs) are very similar. Typically they all have gain in the 50-300 range, maximum usable frequency above 100 MHz, etc. Therefore, for 95% of circuits, you only need to make sure that you have the correct polarity (NPN or PNP), and that you identify the leads correctly. For example, 2N4401 (NPN Si type) is usable where 2N3904, BC548, BC108 are indicated. If the circuit works, but with only one of those, it is very critical (5%) and/or badly designed.
Power transistors require a bit more thought. Maximum current, voltage, and gain are often critical parameters, so they should be the same or better in a replacement transistor.
Back in the day Germanium transistors often had issues with large leakage currents (collector current that flows even with no base current applied). Some inexperienced designers relied on this fact, often omitting the bias resistor(s). Now those circuits are often critical of certain leakage levels to work at all.
[it’s not ‘superhetrodyne’, it’s superheterodyne. ]
‘The regenerative receiver may not be the most advanced receiver ever conceived, and it certainly isn’t the most sensitive.’
Actually, a carefully-made regenerative receiver has equal or better sensitivity than many inexpensive receivers; you have to get into a better ham or communications to get improved sensitivity.
Lots of good info on regenerative designs in google groups and on http://theradioboard.com/rb/index.php
Last week I was listening to shortwave broadcasts from 2000 miles away, on a regen receiver built with 2 transistors and an LM386 audio amp, jammed together on a breadboard, with 8′ of hookup wire as an antenna. It’s fun.
Indeed, one of the classic hacks back in the day was to put in a regenerative circuit in the IF stage of an inexpensive SW general coverage superhet to boost performance
Some commercial low end receivers did that, a tiny capacitor (usually two pieces of insulated wire) from plate to control gird, and the cathode resistor made variable. So it increases selectivity, but also when put into oscillation, act as a BFO for CW.
Maybe more common was the Q-multiplier, which was connected in parallel with the mixer’s output, so it only needed one point. But it boosted selectivity, or with a but more circuitry, could notch out an unwanted signal. That was a common accessory for low end receivers, but some high end receivers included them. One early trick was to use the Q-multiplier (which had a narrow peak, but the skirt was wide) to peak the carrier on an AM signal so the carrier was strong enough for good reception. A precursor of sort to the synchronous detector.
Michael
Oh yes, there were all sorts of kit you could put on those cheap receivers. Along with Q-multipliers, and notch filters, there were preselectors as well if you remember, but the easiest, and least costly was a bit of regen.
Ken: “on a regen receiver built with 2 transistors and an LM386 audio amp” this sounds like a great project too – can you share a link to the schematic?
It’s this one:
http://theradioboard.com/rb/viewtopic.php?f=4&t=7446#p71919
… but that whole thread (and the site) contain many more.
For a first regen circuit, if I wanted to try building one, what would Hackaday suggest? Go for the G3XBM Foxbox or something else?
One year when I was forced to spend the Christmas vacation in the town where I was attending school. I used some my beer money to purchase the Radio Shack regen. shortwave receiver kit for something to do. Respectfully having built one I can’t understand the expectation others do so as well. As I expect others pointed out that a regen. recover can be quite sensitive. More than enough given it’s lack of selectivity. your mention of that 2M AM transceiver remind me that on of my older ARRL radio handbooks has a portable 2M AM transceiver. I recall it uses a superhet receiver.
A regen is simple, and you can get a lot of gain out of it compared to a single stage of amplification without positive feedback. But remember, it came along just over a hundred years ago, it was way better than a “crystal” radio. But all radio was happening on low frequencies. A single tuned circuit down there has fair selectivity, and gets better when boosted by regeneration. “Shortwave” wasn’t even seen as useful, until hams spanned the Atlantic with some tests in December of 1921 (and Howard Armstrong was involved in one of the transmitters for that). So as frequencies moved further up, the regen still boosted the selectivity of the tuned circuit, but that could still be wide.
Regens were simple to build, and that factors in. A beginner could get a receiver going at low cost and effort, and no alignment needed. They remained popular for that reason, until “direct conversion” took the role in the late sixties.
As I mentioned on January 19th, when regen receivers came up here last, the “supergainer” was a hybrid, a superheterodyne, but just a regen receiver after the mixer. Still relatively simple, but a step up from the basic regen.
Note that a superegen is a variant. It’s a regen with an oscillator “modulating” it. This second oscillator runs just above audio frequency. It is wide, but no finicky adjustments. Most of the time, the regen tube is used to oscillate at the ultrasonic frequency, so operation is less obvious. Indeed, like other things, the operation of the superegen became kind of invisible with time, barely an explanation, but then who cared, few were using it? People would try endless things to improve selectivity of the superegen, but forgetting that the quench oscillator was doing the damage.
But it was simple, didn’t need adjustment like the regen and worked well if you didn’t need selectivity. So it was a boon as frequencies went higher up. A single tube would give you a receiver, selectivity or stability didn’t matter when the bands weren’t yet populated. A single tube could be the transmitter. Real simple equipment, and sometimes the same tube was used as a superegen on receive, and a modulated oscillator on transmit. That could populate a ham band, while fancier equipment came later and in smaller numbers. As things changed, the simple equipment would be banished, but the same sort of simple equipment moved to the next higher band. 28MHz, 56MHz, 112MHz, 420MHz, the simple equipment came first.
A lot of VHF ham equipment used suoeregens. Not perfect, but cheap.
Michael
At amateur radio W9BRD I routinely use regenerative-detector-based receivers for two-way Morse code communication, and with care given to design and input-signal level, _and the use of passive audio filtering as close to immediately downstream of the detector as is practical_, they can give basic-radio performance on par with a superhet. Signal-input attenuation such that the antenna-system noise floor is _just above that of the detector_ is where useful design divergence from merely and slavishly rebuilding last century’s regen circuits begins…
David,
Could you provide a valid link to the schematic of the RX you are using?
Thank You !
Jim WB4ILP
A friend of mine has an early Japanese AM radio that only uses TWO transistors…. But it uses one of them TWICE! Uses it first as an RF amplifier and then again as an audio output amplifier. Dang sneaky!
That’s called a Reflex circuit, another popular old-time design. It can be realized with a single vacuum tube.
I think such design is called ‘Reflex’. Same transistor acts as RF and Audio amplifier.
At the age of 10 I read the Ladybird book and I instantly knew what my job would be. At the time I didn’t read well (dyslexia) but this introduction to electronics ignited a thirst for books on the subject and my reading difficulties were left behind.
41 years on and I still have a copy of the book and a few Germanium OC71s in my junk box.
Thanks Dobbs and Robinson.
Amazing what the right stimulus can do for a kid. For me it was a 50-in-1 electronics kit from a prescient uncle.
I have designed many regenerative receiver circuits. My best design is a receiver that requires only two electronic
components. The rest of the circuit is built from things anyone would have in a house. The receiver sensitivity is
very good and selectivity is also good. The power required is 3 volts and a crystal earphone can be used to hear
the signals.
I’m curious to see your circuit. Was it using transistors (JFETs/BJTs), or using tubes? What frequency range does it operate in? How do you control regeneration? Some people use an adjustable “tickler coil”, some use an adjustable capacitor/resistor.
I have designed over 100 regenerative receivers. Both AM and short wave, I have enjoyed everyone. I even design
the poor man’s regenerative receiver that be built form material found in most people homes. The parts required(need to buy is one MPF 102 or equiv. and one 47K resistor. All other material is found in the home including building battery and acquiring an earphone.
“Been there, done that”. And so my first super-regen RX proven to be … “dual-head” arrangement, similar to multivibrator in spirit, self-quenched, using “barrier mode” to simplify circuit, with 2 x PBMS3904 sot-23 BJTs (like small version of 2N2222, very common on dead PC motherboards and so on), targeting 27MHz ISM band. It has been meant as data transmission and remote control thing, so after super regenerative stage there is just mere LM358 opamp, first half just amplifies signal (output filtered of quench), second half is a comparator & data slicer, giving me my data bits back.
Somehow this suprisingly simple arrangement (whole thing is below 2×1 inches in size, about 5×2 cm board, similar to ready factory modules, despite being 1st Rx ever made by me) proven to perform very well. Sensitivity of “dual head” barrier-mode super-regenerator proven to be awesome, at least it readily picks up crappy low power RC transmitter through whole house and even outside of that, despite of laughable output TX power. Funny enough this version also does not needs much tuning. The only tunables are barrier-mode multivibrator current and frequency. The former does not changes things much so one can go with fixed value resistor. So one just have to tune in to desired frequency – and it works. And it works awesome for such a simple, cheap, obvious circuit.
As funny note, sensitivity of circuit and relatively wide band allowed to conduct funny RF measurements, spotting couple of power supplies emissions managing to reach 20+ MHz and so on. Up to being able to detect if laptop in another room is sleeping or active, based on its DCDC noise patterns.
You may or may not be interested to know that sadly the Reverend George Dobbs G3RJV passed away (silent key) on 11th March, 2019. His contribution to the world of amateur radio is immeasurable and he will be sorely missed. His Ladybird book inspired so many of us (of a certain age) into radio and electronics as a hobby.
Looking for a schematic for a practical Regen for 40M and/or 80M to pair with a QRP TX.
Thanks !
Jim WB4ILP