# Help troubleshoot my crystal set

My attempt at building a crystal set has resulted in failure. The circuit I used is the simplest one I could find:

I tried without a ground and with the ground from the power outlet (I heard that this was dangerous, but it's the next best thing I have to a rod in the ground). The result was the same for both. I transmitted from my radio at around 145 MHz. The moment I keyed, I could hear a small click on the earpiece, but nothing more. I couldn't hear myself when I spoke into the microphone.

Here's a picture of the actual circuit:

I believe the diode is a 1N34A.

My guesses at the problem were impedance mismatch and the lack of a good ground, but I also remember that the only issue that would cause is a weaker signal, and I'm right next to the transmitting station, so I'm not sure if that's the reason I can't hear anything. It's really discouraging to spend money on something that doesn't work, so before I go out to get a grounding rod and a sledgehammer or a huge length of wire for an antenna, are those likely to be the reasons I'm not getting any sound? What else could the problem be?

-
The power outlet ground is safe! In Belgium the ground pin is actually on the outlet (in lieu of the plug), and I think it makes more sense: if you touch it you're touching ground, touching the ground pin on a plug is not touching ground. – stevenvh Jul 7 '11 at 6:14
145MHz? That sounds like you're using FM to me. The crystal set will only demodulate AM. – MikeJ-UK Jul 7 '11 at 8:10
What is the resistance of earpiece. In old radios it was about 1000-2000 ohm. Contemporary ones usually have very low resistance, about 4-50 ohm – Rocket Surgeon Jul 7 '11 at 20:38

That's a very very very very minimalist circuit and very 'laid back' construction. It will work with a suitably high impedance ear-piece and proper construction and enough transmitter power and close proximity and ... ie It MAY work but it's not surprising if it doesn't as there aretoo many potntial pitfalls.

• The earpiece looks like it is low impedance - either 8 ohms or perhaps a few hundred ohms.

Without a tuned circuit you need all the sensitivity and lack of loading that you can get. Use of a very very high impedance "crystal earpiece" will make a significant difference.

• Your general construction is excessively rough for what you are trying to do. You MAY make it work but one bad join may doom it and stability is unlikely.

• Your wiring lengths and stray capacitance and inductance are long for use at 145 Mhz.

Clip leads of the sort you are using usually use crimp construction. They can be high resistance or open circuit or have a thin skim of oxidation at the wire to clip contact so that you need a certain level of current flow before they will conduct.

Soldering the core of the circuit with short lengths of wire is far preferable.

Clip leads of the sort shown should ALWAYS be soldered at the wire to clip join. Failure to do this will give you many hours of harmless and unproductive fun. When I buy leads like this I put a loose knot in the wire and remove it once the lead has been soldered.

The diode MAY be a 1N34a. That's a Germanium small signal diode. Rare and not cheap. More usual would be a 1N4148 or 1N916. Mot small glass diodes are probably OK. Do NOT use a low Voltage power diode.(1N400x family etc)

The "rabbits ears" antenna leg is OKish but a length of wire is just as good

Power outlet ground MAY be safe as safe and MAY kill you when you put the earpice in your ear.

-
 Germanium diodes used to be preferred over silicon types due to their lower forward voltage drop - typically 0.45V @ 10mA for an OA47 (though those would be rather hard to find these days!). But I was surprised to see the data sheet for the 1N34A quote Vf max as 1V @ 5mA! But you're right about the crystal earpiece - that's essential. – MikeJ-UK Jul 7 '11 at 8:15 Cmon guys, we are in XXI, use Schottky diode with 0.15V drop :-) One only need to make sure it's hi-frequency one, not a power one. – BarsMonster Jul 7 '11 at 8:34 @BarsMonster - Yeah OK. I won't mention the CV1092 then :o} – MikeJ-UK Jul 7 '11 at 9:19 @Russell: How do I tell if it has high impedance? I thought these looked identical to the earpieces used with some of the circuits I've seen online. @MikeJ-UK: Does that 1V drop mean that it's unsuitable for usage in a crystal set? – howardh Jul 7 '11 at 21:56

That circuit can't work, because the headphones are across the diode and will effectively short it to ground.

This is the simplest possible crystal set:

I built one like it when I was about 12 years old.

Use a Ge or Shottky diode. It needs high-impedance headphones; if you want to use the usual low-impedance headphones that are more common these days, you need to use a suitable transformer on the output, such as the output transformer from an old radio.

With an inductor consisting of about 50 turns of enameled copper wire on a suitable cardboard tube and a 300 pF variable capacitor you should receive several local MW stations with a few feet of wire as the antenna and a good ground. The mains earth might introduce noise, so it's best avoided. A rising water main was recommended when I was young, but probably won't work these days because of the use of plastic pipes. A bedspring was often used as an antenna, but they are thin on the ground these days, as well.

-
If the mains ground is not touch-safe, then why is it a pin in Belgian wall outlets? Everybody, toddlers included, can touch them. – stevenvh Jul 7 '11 at 12:58
The wiring regulations differ from country to country. It's frowned upon in the UK. Anyway, it will pick up a lot of electrical noise. – Leon Heller Jul 7 '11 at 15:11
I was only replying to your statement that it would be dangerous. – stevenvh Jul 7 '11 at 15:56
@Leon - It's encouraged in the US. In fact, I plug the banana jack on my anti-static wrist strap directly into the mains ground on my surge protector. Perhaps you're referring to the neutral line? – Kevin Vermeer Jul 7 '11 at 16:48
@Kevin - actually it's not connected "directly", but through a series resistor in the order of 1M$\Omega$. – stevenvh Jul 7 '11 at 18:18

There are actually a number of different simple circuits for crystal radios. One web site is here.

-