120x90 spot

 
 

Home Projects RF circuits The "UnFETtered Crystal Radio!

 

 

The "UnFETtered Crystal Radio!

author: Austin Hellier © April 2007

Introduction

Several weeks ago I became aware, via the Internet, of an experimental AM 'crystal set' style of receiver, that uses what is known as a ZVB (Zero Voltage Bias) device. This is in the form of an IC, which contains either a 2 pack or 4 pack of specialised FETs (Field Effect Transistors) that require no bias voltage on their gates to conduct current from source to drain. They simply rely on signal voltages to turn them partially or fully on. This new IC, dubbed the ALD110900A, is made by Advanced Linear Devices in the US and is basically an array of zero voltage MOSFETs, configured in the receiver as a 'synchronous detector'. The original article, which appeared in the January 2007 edition of QST magazine, authored by Bob Cutler (callsign N7FKI,) uses such a device, and both medium wave and short wave receiver versions are illustrated for constructors to build. Readers can access the QST website (just Google a few keywords) and obtain a copy of the article, free of charge.
 

 

Many years ago, crystal sets and such were basically abandoned by manufacturers and constructors, as a very poor cousin to modern superhet receivers, and even the novelty effect of 'toy' radios for the kids had all but lost its appeal. The discovery of the germanium diode during WW2 and the subsequent discovery of the point contact transistor by Bell Labs in 1947, rekindled the joys of simple receivers, as those components came onto the market via disposals shops and hobby stores. During the 1970's and 80's, the Ferranti ZN414 and its later incarnation, the MK 484 (AM radio chips) enjoyed tremendous success in the hobby market. Now, they have all but had their day, as manufacturers and retailers once again, move away from the hobby end of the market. If it were not for a band of very dedicated enthusiasts, the hobby part of radio may well have died a long time ago. I'm so glad they didn't give up, as we now have a whole new generation of up and coming technicians, engineers and operators, who can once again "cut their teeth" on 'crystal set' style radios, albeit with the context of a 21st Century update.
 

What is presented in this article, is my version of the QST arrangement, and I'm using the humble 2N5484 JFET, purchased from JAYCAR Electronics for around $2.00. Yes, a JFET, not a MOSFET! Why? Well, my basic knowledge of FET devices at the time of acquiring the QST design was a little rusty, to say the least, not having experimented with them for some time, so after a bit of scratching around on the net, and down at the local library, I reclued myself as to their peculiarities. Devices that are static sensitive have protection diodes on the inputs, and the ALD device certainly has those. Basic JFETs also have a protective diode between the gate and source connections, presumably for the same reason, but neither the QST design, nor my adaptation of it, use any internal diodes as a rectifier. You can experiment with that as a start - simply replace your germanium or Schottky diode with the gate and source of a JFET device, and you'll get reception all right - but it does seem a bit mushy or scratchy, somewhat like a poor Schottky diode that distorts on low signal levels (BAT46's come to mind…) You can still obtain reasonable results by tapping the gate further up the tuning coil, if you want to.
 

The basic premise behind the QST article, is that you drive the gate of the FET with an RF voltage derived from the top of the tank circuit via C1, and the gate switches on and off very rapidly, at the resonant frequency that you are tuned to. The source connection is tapped into the tuning coil low down, as a means of impedance matching with the headphones and acts as the anode, while the drain connects with your headphones as the cathode, to complete the detector part of the circuit. According to the experts on the net, much success has been enjoyed by one and all, and there's a lot of chatter about this most recent innovation in the realm of simple AM receivers (292 posts so far, to the 'Rap'n'tap' chatroom of the American Crystal Set Society alone!- www.midnightscience.com). One of the most curious aspects of this little beauty is that the antenna/ground system I employ basically entails a 'short' antenna, and a water pipe ground. Ideally, short antenna wires work best near the top of the tuning coil, but in this case, the best position seems to be right at the bottom tap! Normally, this would send all my weak locals into a spin, and shift the whole band up towards the top end of the tuning cap's range. The JFET device seems to act like a FET, (not a diode) as connecting it to the tank circuit via the gate lead does not appear to cause any loading, and flatten out the 'Q' of the LC tank, in any discernable way.
 

Another aspect of using a JFET in this manner, is the actual sound quality that you get in the headphones. The audio is very clear ans is in no way muffled, or distorted. Diode detectors often produce poor results, and while this can sometimes be attributed to bad antenna/ground systems & poor layout and construction of the receiver, at the end of the day, a FET device has it over a simple diode on a number of fronts. Diodes tend to introduce various distortions, and they also exhibit very high output impedance. With this design using the 2N5484 device, it is possible to get rid of most distortions and at the same time, use just about any kind of audio transducer that you may have on hand. I have successfully used a crystal earphone, my pair of 'Scientific' 2KR headphones, and a couple of low impedance telecom style inserts, and all work reasonably well, without the need for matching transformers, or extra passive components, beyond the usual 0.001uF cap, or 47KR ballast resistor. Diode detectors can also cause loading on the tank circuit if they are tapped too high up the coil windings. The JFET in this circuit is tapped way down, at around 10 turns, with the antenna lead sitting just under that on 5 turns from the grounded end of the coil. Sensitivity is OK, and selectivity is quite good, for a simple AM receiver
 

 

A Word about Coils, Caps, and Connections


This version of the receiver will work quite OK with air cored inductors, providing you use PVC insulated wire. I don't recommend enamelled copper wire, as it can be down a bit on sensitivity and selectivity. Get some good quality, thinnish and untinned (copper colour, not silvered colour) multi stranded hookup wire for winding the tuning coil and a tube made of plastic, rather than cardboard. If you want to use a ferrite rod antenna coil instead, then go right ahead, Ones that come from el-cheapo pocket radios will work but you will need to remove some of the turns from the primary coil for them to work with a combined 160+60pF PVC tuning cap, and then create one low Z tap, by soldering the end of the primary coil to the start of the smaller secondary coil. If you can get your hands on some high Q 'litz' wire, then use that on a bare ferrite rod. Simply fasten the wire onto one end of the rod using some tape, then wind on around 50 to 60 turns, with two taps - one at five turns from the grounded end (for the antenna lead in) and the other one at ten turns for the Source (anode) connection. You may need to experiment with the number of turns, and the tapping points on the coil, depending on the value of VC1.

 


To the left you can see a basic circuit diagram of my radio, based on the QST design, and to the right, a physical wiring diagram of how it all goes together. Note that both sections of VC1's fixed plates (A and O) are wired together, and that your headphone connections go between 'd' of the FET and the common ground 'G' point.

 

 

Two alternative methods for connecting either a crystal earphone (left) or low impedance headphones (right).

Parts List
C1 470pF styroseal capacitor
C2 0.001uF MKT capacitor
VC1 240pF (160+60 dual gang) Polyvaricon tuning capacitor
L1a 90 turns of pvc insulated wire wound on 3" diameter pvc tube or BCB ferrite rod with tapped coil
L1b 50 to 60 turns of litz wire wound around a good quality ferrite rod 100mm long by 10mm Ø
Q1 2N5484 JFET device
HP high impedance (2K-4KR) magnetic, dynamic headphones
EP crystal earphone and 47K ballast resistor
LHP low impedance headphones 8R up to 64R
TX1 impedance matching transformer - JAYCAR cat.# M-1109 or similar
 

Download this schematic in doc

 

 
 

 
120x90 spot

 

 

Search Site | Advertising | Contact Us
Elektrotekno.com | Free Schematics Search Engine | Electronic Kits | Electronics Projects