Transistorized Crystal Radio

Transforming a crystal radio with a few extra components

Features

• Single Darlington pair transistor stage for rf amplification, detection and audio amplification with optional detecting diode

• Standard headphones

• Signal taken from tuning coil via tapping point

Overview

Crystal radios are great - if you have a long aerial, good earth connection, strong transmitters nearby and a working crystal earpiece (modern ones are hit-and-miss) that is. That's quite a few 'ifs' and plenty of scope for disappointment, especially for the younger user.

The Maxitronix crystal radio works well, subject to the above conditions. It's of good design, with a separate coil for the aerial and a tapping point on the main coil for detection and output. It also has spring connectors, which allow for experimentation.

However, with just a few extra components, and taking advantage of those spring connectors, it can be turned into a simple transistor radio (see picture) that gives good reception using a short aerial. And it uses modern headphones to boot. This does lead me to wonder why 'educational' crystal radios don't (or perhaps more accurately, didn't, as they no longer seem to be made) come with a few extras like these so that they can be turned optionally into transistor radios - disappointment averted, even if the crystal radio fails to work.

A quick side-by-side comparison with the Ladybird transistor radio using the same aerial shows that this radio is able to pick up the same local stations at reasonable volume, which again aren't loud enough to warrant a volume control. Tuning is less precise, because regeneration isn't involved. 

The inspiration

The circuit behind this modified radio was taken from a Porthcurno Wireless Receiver radio kit that I bought some years' ago from ebay seller tgs1950 (Poldhu Radio Kits). It has a single Darlington Pair transistor (two transistors in a single package) which it uses for detection and amplification, making it an audion receiver, according to the seller. The radio is shown below. Originally it had two tiny tuning capacitors, C3 & C4, mounted in parallel on the board, but I replaced these with pins to connect to physically larger tuning capacitors off the board. The colours on the inductors are difficult to identify (red/brown, blue/black), but suggest values of 0.41 mH and 0.01 mH for the AM and HF inductors respectively, confirmed approximately by my cheap (and possibly unreliable) LCR meter/component tester.  

Some points:

• I imagine the diodes are there to protect the circuit from the battery being installed the wrong way around, but one should be sufficient. I experimented with this design, also using two diodes at first. I noticed that these provide substantial (if unintended?) detection in addition to that provided by the DP transistor - with just one, the sound was quieter and quieter still if I bypassed both diodes. Perhaps we shouldn't make too much of the audion feature? Anyway, I decided to retain the original germanium diode from the crystal radio.

• aerial, tuning capacitor and detection/amplification are all connected to the top of the coil.

• the capacitor marked C2, value 1uF, on the PCB, as supplied was actually marked 104, i.e. value 0.1uF.

• C1 seems unnecessarily high (10uF) simply for passing rf signal from the tuned circuit to the transistor base. Maybe now, I'd start with a non-electrolytic one, perhaps even as low as 0.01uF, as used in the Ladybird radio.

• An earth connection, while not essential, will improve volume.

• Here's another audion receiver. It's essentially the same circuit as this one, but, there are no diodes and there is a tapping point on the coil, 20% above ground, for detection/amplification. 

The mods

See picture at the top of the page.

Extra components:

• 3V battery and connectors

• Stereo headphone socket. The top tag in this case - ground -  is not connected. (Mono) sound should be heard in both sides of the headphones. Alternatively, a single phone from an unwanted headphone could be wired directly between the terminals without need for a socket.

• 100k ohm resistor

• 10 uF electrolytic capacitor

• 0.1 uF (marked 104) disk capacitor

• MPS A13 Darlington Pair transistor

• (optional) diode between the anode and headphone socket to protect the circuit. This seems to have the added benefit of providing additional detection.

The tuning capacitor and coil are wired under the board as per the instructions that came with the crystal radio and connections above the board are as shown. I'm using the part of the coil between terminals 8 and 10 only. A separate winding is provided for the aerial, with terminals 6 (to aerial) and 7 (to ground). However, with my short aerial I couldn't get this to work and connected the aerial to terminal 12 (& 10) instead. 

I'm making use of the tapping point for detection/amplification.

With the diode placed between the electrolytic capacitor and transistor, the transistor amplifies (mostly?) the audio signal; if placed after the transistor, then the transistor amplifies (mostly?) radio frequency. Either way, volume seems to be about the same, i.e. the transistor seems to work well with both types of signal.

I decided to use a small breadboard, rather than place all of the extra components on the spring connectors on the radio, as these are fiddly to use and I wanted to experiment readily. For this reason, I also put the diode there. The downside of using a breadboard is that connections between components are less robust, so it's not a good option for a permanent solution.

A schematic of this circuit, but for now without the diode and earth connection, is shown below.

Performance

With the configuration shown, and the now sadly very limited number of MW stations available to me locally, there is daytime reception of BBC Radio 5L (reasonable volume) at  ~6 on the dial and Smooth Radio (higher frequency, quieter) at ~10. 

However, I was astonished to find that it also picks up shortwave (SW) stations Radio Romania International Channel 1 & CGTN from early evening onwards, interfering with R5L and swamping anything of much higher frequency. At times it is even possible to hear these without an aerial! This also means that I am unable to listen to higher frequency overseas MW stations at these times, so it's no good for DX-ing on MW.

A partial explanation perhaps: higher frequency stations are tuned in with lower capacitance (resonant frequency of a tuned circuit, capacitance C, inductance L, varies as (1/LC)**0.5) and as this approaches zero, we might expect to pick up some SW, non-selectively, towards 10 on the scale. However, this doesn't really explain why MW stations should be drowned out over such a wide tuning range.

 For now it's possible to mitigate against this somewhat by reducing the section of coil used, e.g. by connecting terminal 12 to terminal 9 instead of 10. This requires higher capacitance to tune into stations and shifts the problem higher up the MW band.

On further investigation, I found the same issue with SW reception when I reinstated the original crystal radio, so it's not the mods that are making this happen. It's been a while since I last played with crystal radios and I always regarded hearing anything as a bonus, so if I had heard these stations at the time, I would likely have assumed they were on MW, although the lack of selectivity with these stations should have indicated an issue. 

Nor is it inherent to the structure of the radio (I was thinking that perhaps those spring connectors might be having some unforeseen effect), as I it also happens when I make similar mods to a crystal radio of different design.

This does not happen with the Ladybird Radio, btw.

I wonder - could a suitable rf filter could be used to suppress these signals, is this caused by the environment I'm in - same room in a house with steel piles (terminology?) inside the walls?

May 24: also noticed with a 6 transistor radio project, the cause is likely the proximity of my 'workbench', a window ledge, to an aerial, a metal-framed window, which concentrates these signals. Away from the window, SW reception disappears...

Alternative circuits

Reflex receiver

I tried a different option, based on the first stage only of this reflex receiver. It was always going to be on the quiet side without the additional amplification stage, but as a result needed fewer components and I was interested to see how this type of radio would perform. I again used a small breadboard with the radio. A major difference with this receiver is the use of the coil: it has a separate aerial coil (provided in the crystal radio by using terminals 6 and 7) and the tuning coil and variable capacitor are electrically isolated from the rest of the circuit. I didn't use the 0.0047 uF capacitor that is attached to ground here (intended to remove any unwanted signals?*) nor the 330 uF electrolytic capacitor across the battery. I adapted the wiring used for a crystal earpiece with transistors in the Ladybird radio book. One wire of the crystal earpiece was connected to the end of the inductor (I used a ~ 4.7 mH, not 10 mH, btw, as that is all I had to hand) and the other to the +ve terminal of the battery, with a ~3k ohm resistor in parallel**. At first, the radio was silent, but after checking the connections and slowly turning the tuning dial, I was able - without an aerial wire - to hear BBC Radio 5L and to just about make out the presence of a  couple of other stations - possibly Talk Sport and Absolute Radio; reception seemed better at night. Tuning was very precise indeed and stations could be easily missed as a result - slow-motion tuning would be so much better. 

Compared with the audion circuit:

• pros: can function without an aerial

• cons: needs more components, needs higher voltage battery, very quiet (fair enough - it's using just the one transistor to the audion's two), requires careful tuning, needs high impedance crystal earpiece or headphones. 

* If so, then I wonder why this is so much smaller in value than the one used for this purpose in the Ladybird radio (100 uF)?

** It seems to work OK without this resistor!

Apr 24: I'm curious as to whether this will also pick up unwanted SW stations at night and will build it again to check.

Simple regenerative receiver

This comprises just the first stage of the Ladybird radio with a crystal earpiece, as described at the bottom of this page. 

Compared with the audion circuit:

• pros: more selective tuning through regeneration

• cons: needs 11 extra components, compared with 4 for the audion (5 if we count the Darlington pair as two transistors); uses a crystal earpiece; regeneration needs to be adjusted each time you tune in to a new station.