If you are going to build a one-transistor radio, make it this one!
This reflex radio project was inspired by Robert Bazian 's design. His reflex radio is the "darndest" thing I have seen and his spectacular results inspired me to come up with my own version! These designs are similar to two-transistor circuits used in some ancient Japanese "Boys Radios" except that the inter-stage transformer directly drives the speaker.
My version is restricted to components that are fairly easy to get or make and the results are amazing. Instead of a loopstick antenna, I use an air-wound coil that is easy to make and I use an ordinary "telecom" transformer for the feedback signal. (I have a plentiful supply of these, by the way.) The one part that might be a little hard to get is the MPS-A18 and I don't recommend substitutions unless you can find a very high gain transistor. A Darlington can be made to work but there may be noticeable distortion. Here is the schematic:
The antenna coil is just 20 turns wrapped around four insulated posts as seen in the photos and shown in the diagram below. I used nylon posts but ordinary bolts covered with tubing would work fine. The dimensions are not particularly critical; those shown happen to fit inside the case I chose. If your antenna tunes too low, remove a turn or two and if it only picks up stations at the top of the band, add a couple of turns. The signal is coupled to the receiver by two turns wrapped right on top of the 20 turns. This winding technique is simple but it has the drawback of having more than desirable capacitance and that limits the tuning range with a standard 365 pF variable capacitor. To get around this limitation, a switch and 100uH inductor has been added as a band switch to allow the reception of the higher frequency stations. This antenna system also tends to give more regeneration at the high end of the band and the volume control (5 k) must be turned down a bit.
The detector diode is a Schottky type intended for RF applications but a germanium detector diode will also work. There is a bias current flowing so other types of diodes might be worth a try (see detector diodes). I may have to try an LED just because it is so odd.
The output transformer may be a 1k to 8 ohm audio type or even a 12 volt filament power transformer. Some power transformers sound muffled due to poor frequency response so if your radio sounds like the treble control is all the way down, try changing to a different style of transformer. But don't hesitate to try a power transformer; the best performance I got was with a power transformer, outperforming my smaller 1k to 8 ohm audio type seen in the photos and they are easy to find. (The 12 volt winding goes to the speaker and the 120 volt winding goes to the transistor.) Although I didn't try it, a 6 volt power transformer might work well with a 4 ohm speaker.
A second transistor audio amplifier driven by a 10k to 2k transformer would give incredible performance!
A line-powered version using two high voltage transistors might be interesting, easily driving a larger speaker.
A varactor tuned receiver might allow the regeneration to be adjusted automatically as the radio is tuned. The voltage that tunes the varactor could also adjust the gain in some way.
Robert Bazian suggests using his circuit as a 455kHz IF amplifier in conjunction with another transistor as a traditional autodyne front-end. He claims that his prototype works about as well as a typical transistor radio!
Make sure to visit reader Mike's page, http://oldradiobuilder.com/One%20Transistor%20Radio.html (another website) to see a true work of art version of this radio! He also made some interesting improvements.
Also see Tom's work of art with interesting improvements:http://www.tompolk.com/radios/macrohenrydyne/macrohenrydyne.html
Make sure to see Don's page detailing his version.
Marcos from Brazil built this version:
Karen from the UK built this one:
The case is a CD storage box, bought from a UK supermarket. It was difficult to work and not really a good choice - the plastic is brittle and cracks easily.
My radio uses a BC549C transistor. This has a lower typical beta than the MPS-A18 so if anyone is thinking of using this part they might like to buy a half dozen and try them to find the best. I suppose I got lucky - my only BC549C works about as well as an MPS-A18.
I used a pulse transformer for the RF transformer. This has three windings, two of which I joined together to form the non-collector winding. I discovered the windings were all wound in the same direction, so to join two in series it was necessary to have a link that passed under the device (had I joined two adjacent pins the windings would have cancelled).
The circuit is hanging on the loopstick, which is conveniently supported by two slots on the inside of the case! A rubber foot on the back of the speaker provides all the extra support needed. I didn't fit a back to my radio - how can you prove its a one transistor radio if you have a back on it? :)
I had to lop about two centimetres off my loopstick to make it fit. This is always a fraught affair because ferrite is so very hard and brittle. There are diamond-tipped tools that will do the job but if, like me, all you have is a junior hacksaw then it can be accomplished by carefully cutting a groove all around the point you want to cut it. With a but of luck you can then cleanly break off the excess with a little force. But there is a lot of luck involved so don't attempt to cut down your loopstick unless you absolutely have to.
My loopstick is held to the board - or more correctly, my board is held to the loopstick - by cable ties and cork stand-offs. DON'T do what I did many years ago and use wire loops to secure the loopstick. They represent shorted turns and will completely prevent your radio from working!
I can get all three of our local MW stations on this radio, one at excessive volume. I should point out that I am less than thirty miles from our Droitwich transmitter, which doubtless helps. I found the circuit to be very well behaved and not at all critical in construction.
UK hobbyists may be interested to know that most of the parts I used are obtainable from our Maplin chain of high street stores:
|Part Maplin part number|
|BAT43 schottky diode VR19V|
|BC549C transistor QQ15R|
|Pulse transformer PT6 N79CC|
|Audio transformer LT700 LB14Q |
Maplin also do a loopstick that you could almost certainly use, though I haven't tried this - Maplin part number LB12N. The only part you can't get from Maplin is the tuning capacitor so you'll have to look to an alternative supplier for that.
What an amazing circuit! It breaks all the rules. Is anyone up for the challenge of an FM version? :))
|Here is a two transistor radio that really performs! The first transistor acts as a reflex stage similar to the one-transistor reflex radio above except that no RF transformer is used. Instead, a 10 millihenry choke sends the RF to the diode detector and the audio to the output stage. This simpler approach is not as sensitive as the transformer feedback version but the addition of a simple one-transistor audio amplifier makes this receiver really scream. You will not be using this radio with the volume control turned up full! And you can really pull in stations with a simple loopstick antenna.|
Here is a tidy, attractive version using brown phenolic standoffs built by Marcos Kusnick from Brazil.
Marcos says: Hello Charles. That’s the boy. As you said in your page it really performs! It is the best MW radio that I've made.
|Here is a version built into a surplus computer speaker using an LM386 audio amplifier IC. The circuit is built onto a little piece of phenolic perfboard that is secured by wedging the edges of the perfboard into a couple of bifurcated standoffs. The on-off switch is part of the volume control which is mounted next to the tuning capacitor. The loopstick antenna is glued to a couple of plastic posts that are part of the enclosure as can be seen in the photo. The 9 volt battery is located in the C-cell battery compartment with plenty of room to spare. The performance of this simple receiver is quite good, giving a blasting full volume on a couple of stations with no antenna. Play with the number of turns on the secondary to get the best results; more turns will give more sensitivity but the circuit might become unstable, especially at the low end of the band. Fewer turns will give better sound but at some point the sensitivity will drop off. Three turns makes for a great receiver with my version. I had to remove a significant number of turns from the loopstick to make it tune properly with the 365 pF variable capacitor I chose because the loopstick was designed for a significantly lower value capacitor.|
The photos below show two versions of a three-transistor version; the third transistor really makes the receiver come alive. In fact, the breadboard version is entertaining me with the "Clark Howard" show from across the room as I type. It works quite well without an antenna on the stronger stations. In order to avoid an audio transformer I connected two 16 ohm speakers in series and mounted them side-by-side in a wooden box. Keep the "phasing" of the two speakers the same by connecting the left terminal of one to the right terminal of the other (assuming they are identical speakers). The pocket radio version uses a tiny 39 ohm speaker I happened to have. Don't be misled; the "pocket radio" version really doesn't work all that well due to the small ferrite antenna coil, small speaker and low battery voltage. But it does work! It would serve better as an "earphone only" design.