Experimenting with a crystal radio

[WaveyDipole]

A few years ago as a lad I experimented with building a crystal radio. One coil of wire., one variable capacitor in parrallel and one germanium diode in series with a crystal earpiece was enohg to pick up a couple of local stations, in particular Radio Leicester. Back then I lived in the city.

Well a while ago, I came across a crystal radio kit in an electircal shop. It was being sold off at 1.99 to clear to I thought I'd give it a go. My daughter was also interested to see whether it would actuall work, only unfortunately it didn't. The kit comprised of the basic components above and a board to clip them on to except it came supplied with what looks like a small silicon signal diode (1N4148?). I used the radiator as a ground and about 10 foot of wire for the antenna but nothing. I tried different diodes including a BAT80 and a 1N34 but still nothing. We were both disappointed and the kit got put into a drawer and forgotten. I should point out that I now live in the County, about 10 miles from the city so I figured that my distance from a transmitter might be a factor.

I came across it again tonight and pulled it to experiment with. The base board has gone, so I just used a connector (domino) block to connect everything up and a random length of wire for the antenna. At first I used just the coil - probably 60 or so turns on a plastic former about an inch and a quarter in diameter - and again I got nothing. I tried adding a resistor (68k) across the earpiece. I also added a few feet of wire for an antenna. Still nothing. I then conected the 20m longwire outdoor antenna that I use for the radio. This finally worked and I picked up just a station! Just the local Asian one, but it at least it worked - sort of.

Firstly, to make it work, the 68k resistor was required. Without it, the set would not work. Secondly, I can only connect the antenna. If I connect the earth (as per most diagrams), the set goes dead. Finally, I discovered that I need to keep my hand near the coil, otherwise the signal fades away and then tuning becomes difficult.

Why does the earth make things worse? I should point out this is a connection to an earth rod, not mains earth. And why does my hand need to be near the coil for the signal to be sufficiently strong enough to hear?

[ThePillenwerfer]

I'm no expert on these things but have always believed that you need a germanium diode in the OAxx series. I have read that a low current LED will work, though my experiments with one failed.

There are some more complicated circuits available here: http://rexresearch.com/xtlradio/boysbook.pdf, which may be of benefit if you're in a poor reception area.

Good luck with your experiments.

— Joe

['LIVEWIRE?']

I would have thought that a Germanium point contact diode, such as the OA81, OA90/91 or AA119 was essential as the detector in a Crystal radio. These are the nearest 'modern' components to the od 'Cat's whisker' radios, which used a crystakl of Galena being touched by a fine wire - hence the term 'Cat's Whisker'

[Herald1360]

Gold bonded types like OA47 are best, though a bit of judicious dc bias can help a lot with Si or Schottky types.

If you get your dc supply from another crystal set tuned to a strong local signal, I guess it's not "cheating".

Large low loss hi Q coils, multiple tuned circuits and careful impedance matching seem to be the way to go. Some of the professional early '20s designs from Marconi et al for ship's sets were pretty effective.

Have a look at what the "Crystal set society" gets up to.......

[ms660]

The average xtal set with a tuned tank with the antenna connected directly to the top of the tank and the detector input from the top of the tank isn't the best of performers, as far as I understand it the average random bit of wire for MW will be capacitive, not only that it will have resistance too, a bit of a Q killer if connected to a high impedance resonant tank, the capacitance can be tuned out by connecting an inductance with the same reactance value as that capacitance in the antenna circuit but that still leaves the resistance, the resistance is likely to be low, so low impedance connected directly to high impedance is not good, in other words with a decent earth connection your capacitive antenna will detune the tank and at the same time reduce Q (selectivity)

A common trick is to connect a low value capacitor between the top of the tank and the antenna, that will reduce antenna loading (selectivity increased) but can reduce the available signal strength.

There are other ways to connect the antenna to the tank to try and improve things, the antenna can be coupled via a link winding with a small number of turns or via a tap towards the bottom of the tank coil, in effect impedance transformation. With either of those methods you can also connect an inductance in series to cancel out the capacitive reactance of the antenna.

The detector and it's load can also be a Q killer, one solution to improve things is to feed the detector from a tap lower down on the tank coil, a matching transformer can also be used, there is one now being manufactured and available, the bloke that does them is a xtal set enthusiast as far as I can gather, It taps from 200K down to 4 ohms with 17 taps, transformer link here:

http://www.ebay.com/itm/KPB-02-Cryst...e=STRK:MESE:IT

Lot's of experimenting, good luck.

Lawrence.

[G8HQP Dave]

Originally Posted by WaveyDipole

Firstly, to make it work, the 68k resistor was required. Without it, the set would not work.

If you use a crystal earpiece then you may need a resistor across it to provide a DC path.

Originally Posted by pwdrive

The average xtal set with a tuned tank with the antenna connected directly to the top of the tank and the detector input from the top of the tank isn't the best of performers, as far as I understand it the average random bit of wire for MW will be capacitive, not only that it will have resistance too, a bit of a Q killer if connected to a high impedance resonant tank, the capacitance can be tuned out by connecting an inductance with the same reactance value as that capacitance in the antenna circuit but that still leaves the resistance, the resistance is likely to be low, so low impedance connected directly to high impedance is not good, in other words with a decent earth connection your capacitive antenna will detune the tank and at the same time reduce Q (selectivity)

A short MW antenna (often incorrectly called a 'long wire') will look like a capacitor in series with a small resistance. However, when you connect this in parallel with a tuned circuit it is the parallel equivalant of the antenna which matters, and this is a capacitor in parallel with a high resistance. Hence if the antenna is not too long it will not load the tuned circuit by much. The longer the antenna the smaller will be the parallel resistance so longer antennas may have to be tapped down the coil, or matched in some other way.

Originally Posted by WaveyDipole

Why does the earth make things worse?

With the earth you have an antenna, with perhaps significant capacitance, connected straight across the tuned circuit - this will detune it by quite a large amount. Without the earth you simply have a bit of stray capacitance; too little to detune it, but also too high an impedance to feed in much signal.

And why does my hand need to be near the coil for the signal to be sufficiently strong enough to hear?

Putting you hand near will significantly increase the stray capacitance, thus reducing the impedance of the signal source and so increasing the signal seen by the tuned circuit.

Crystal sets often seem to deny the rule that "there is no such thing as a voltage; voltage is always between two points" but when you add stray capacitance you find that mysteries disappear. Crystal sets are so sensitive to stray capacitance because they usually operate at quite high impedances.

[ms660]

Also worth remembering that MW transmissions are vertically polarized, so the longer the vertical component of the receiving antenna the better.

Folks often knock the humble xtal set but a well designed well constructed one with a decent antenna and earth is quite capable of good performance, as has been said impedance matching is very important for good results.

There's always the loop antenna as well.

Lawrence.

[G6Tanuki]

Originally Posted by pwdrive

Folks often knock the humble xtal set but a well designed well constructed one with a decent antenna and earth is quite capable of good performance, as has been said impedance matching is very important for good results.

The biggest issue these days is - once the initial attraction wears off - finding content that's worth listening-to: broadcasters are abandoning MW/LW in droves.

[ms660]

Yeah, some might say what's the point of any MW/LW receiver.

Lawrence.

[Restoration73]

Loads of MW still in the capital - I had success with Year 6s building these;
https://www.quasarelectronics.co.uk/...elenco-mx-901c
If you connect a ready made 10uH inductor across the MW coil, reception of SW is
possible, some stations e.g. CRI (China) are almost guaranteed.

[WaveyDipole]

I agree that for best results a Germanium diode is required, which is why I was initially surprised when the original kit contained what looked like a standard silicon small signal diode. Then, of course, I realised that they are much cheaper than Germanium diodes so perhaps the objective was to keep the lit as cheap as possible. I was told that a Schotky such as a BAT80 should work euqally well although I haven't put this to the test as yet. The diode I had to hand is a 1N34, which is a germanium diode, but I had considered getting an OA series one as well and I had heard that the gold bonded ones have good sensitivity. I have noted the earlier mention of the OA47 and I was also thinking about the OA95. Does anyone know whether it makes much difference. Are some diodes in the OA series better suited than others?

Thanks for the link to that Boys book. I have seen it before but had forgotten about it.

I also realised that I hadn't considered the antenna properly. The outdoor antenna is fed via a 9:1 balun into a coax cable. Obviously with the earth connected, this means that the crystal set is being presented with a low impedance and then there is the capacitance of the cable. By disconnecting the earth and only using the core the impedance goes high again so it works.

I AM stations are the traditional focus for crystal radios and there are still a few stations around. My main reason is curiosity and the desire to experiment, but if I get fed up or the stations get closed down then I guess I could also build one of these:
http://solomonsmusic.net/FM_CrystalRadio.html

[FrankB]

Germanium diodes work the best. But if you can go to a rock & mineral show, likely you can pick up some galena, silver, platinum, copper, pyrite, tin, etc. raw crystals and make your own cat whisker detector to use them with. A nice asst. here runs > $5.00 USD.
(Of course if you have a friend who owns a mine, likely you can get them for free, like I did. ).
Its fun to try different types and see the results. (Also carborundum detectors are cool, but you will need a battery to make it work properly.)

[G4XWDJim]

I've often thought about building a VHF crystal set to see if I can receive air band to listen to the pilot when I'm on a plane. Should be possible.

Jim

[Herald1360]

Won't the body of the plane screen the inside fairly effectively from anything that's being transmitted by an external aerial?

[G4XWDJim]

Dunno but people use mobile phones on planes seemingly well enough.

Jim

[AC/HL]

Googling passive airband receiver returns a wealth of information.

[julie_m]

I'm not too sure I'd want to be carrying anything onto a plane that looked home-made and could be used for listening to the pilot .....

[Radio Wrangler]

Originally Posted by G4XWDJim

I've often thought about building a VHF crystal set to see if I can receive air band to listen to the pilot when I'm on a plane. Should be possible.

Possible, but perhaps a bit disappointing in use.

How will a will a security person (hired on minimum wage) interpret home-made electronics in the crazy age we live in?

You may find it a bit less than enthralling listening to only one end of the conversation, also transmissions are very curt and there are long long gaps between contacts.

The transmitters are a bit over 16W to allow for coax losses and environmental effects, but not a lot over. Antennae are on usually top and out of line of the windows. There will be a bit gets in.

You'll need some other sort of entertainment for most of the time and you'll just realise something is being said by the time the transmitted phrase ends.

David

[Radio Wrangler]

oops! What timing!

David

[turretslug]

A long time ago, I'd toyed with the idea of making a loop of standard many-ways 0.05" pitch ribbon cable around the inside of the semi-rigid suitcase I use for overseas breaks/work so I'd have an instant-use MW/LW/NDB loop in conjunction with the little multi-band portable that also accompanies me to such places. Also a long time ago, it occurred to me that passing such a thing through a scanner might result in some distinctly hostile questioning and delays.

Things have only got worse since.