My crystal radio circuit is not working as intended. What don't I understand?

Hello, I am a self-taught electronics hobbyist and I've been on a mission to understand radio transmissions. I've started with some simple crystal radio circuits and have had some success. Recently I attempted to build the circuit below:



This is my fully assembled attempt:



My ground in this case is a connection to the ground prong of my home's wiring. I have also tried a piece of buried rebar and a cold water pipe. My aerial is two vertical insulated parallel wires each 15 ft long.

When everything is assembled and ground is connected where it is specified in my picture, I have little or no static or audio from the earpiece; However, when I connect ground to the screw labeled ' * ', or even touch it, using my body as a ground, the signal becomes loud and fully audible. If I understand correctly then my ground connection is then on the opposite intended side. Does this bypass my LC tuner circut? If not, why is it not where my schematic specifies?

All of my connections are sound, I've tested with a multimeter.
Can any one of you folks explain to me what is happening?

Thanks for your thoughts and time.

connction to wiring ground or pipe is not necessary.


There may be a mismatch in coil/capacitor tuning circuit .

Are you able to tune the reciever and get varying noise in the headphone ?

Since you said that touching the '*' point have a reception , it is the coil/cap side problem.

Despite being simple, there is an error in the schematic which has been copied into the construction and a further error in the construction alone.

1. the bottom end of the coil (as seen it the schematic) should NOT go to ground. If grounded, it shorts out the section of coil between the metal wiper and the bottom of the coil which has the effect of stopping the top of the coil working as well.

2. The coil wire is insulated so the metal wiper will not connect to it anyway! To make an adjustable coil like that, you must use bare (uninsulated) wire and also leave a space between each turn so they don't all short together.

For point 2, as the coil will have to be made again, you might find it easier to construct it in a different way. The idea is to be able to select different numbers of turns so instead of the complication of insulating and spacing the wire, every 10 turns or so, make a small (~10mm) sideways loop in the wire and twist it to prevent it collapsing. So you have essentially the same coil as at the moment with lots of twisted 'spikes' pointing out of it. Then clean the insulation from the spikes and instead of the wiper arm, use a short wire and alligator clip to connect to one of them. It doesn't give you continuous adjustment, you don't need that anyway, but it still allows you to select multiples of 10 turns which is more than adequate.

Brian.

admiralento said:

Thanks for your thoughts and time.

Click to expand...

Do not connect the electrical power ground to your radio; they run parallel to the power cables and pick up all kinds of noise. Buried rebar and cold water pipes are FAR better.

Your circuit is designed for AM reception (not FM) and you just need one long piece of wire (preferably outside) as antenna. That will be an untuned general purpose antenna. It should be horizontal (vertical for FM) - just stretch it like a clothes line outside.

Now coming to the circuit, avoid tap on L1, use C for tuning.

Just short the diode and see the result. Do not touch any part while doing the tests. Perhaps the diode is dropping too much- you may need to use a galena crystal for diode.

I too did this circuit 50 years ago; I too did not understand what was going on.

The diode doesn't look like a Germanium diode. That's a Germanium diode:



- - - Updated - - -

I see that mentioned 1N270 is a replacement for classical point contact diodes like AA118 in my photo. But your diode looks more like a Silicon diode.

FvM said:

...That's a Germanium diode:...

Click to expand...

I doubt these point contact Ge diodes are easily available these days; the common junction diodes do not work well at RF rectification and they drop too much voltage to be useful in self-powered radios.

But what was common those days was a simple PbS crystal touched with a sharp pin; you need to experiment to find a good spot.

It's even harder to find good Galena!
Go for a small signal Schottky diode like the BAT85, they are easy to get hold of, inexpensive and work better than the Germanium diode in the original design.

Brian.

Hey Brian and thanks for the thought out response.
I neglected to inform you that the top of the coil's turns have been sanded to remove the top layer of enamel. Also I and took care to make sure the exposed wire did not connect between turns. The shiny band on the coil in my picture is not simply glare! ;-)

In my multi-meter checks I did confirm that the wiper does have electrical contact; However the tap system sounds very effective and would entail much less hassle! I will definitively try your design in future experiments.

After reading your response section titled "1.", I am now under the impression that whatever adjustable metal contact system is used(wiper or taps) in the coil cannot be connected directly to ground. So could you please clarify where this proposed alligator clip should connect to in my circut?

Thanks for your efforts,
Admiral

connction to wiring ground or pipe is not necessary.

Click to expand...

Thanks srizbf,
I was under the impression that the modulated signal could not induce oscillation in my circuit's current unless it had a sufficient drain for the charge be absorbed by, thus ground.
Perhaps I did not understand your point, could you please clarify?

Thanks for your time,
Admiral

I see you guys have been discussing the diode I choose, the 1N270. I believe it is a functional germanium diode as evidenced by this video:
https://www.youtube.com/watch?v=n3IeDz5yCSk

I know there are better models out there but I need a little bit of help before I go shopping.
I know a silicon diode will not conduct till it has a forward voltage applied exceeding ~0.7V (in other words has a forward voltage drop of ~0.7V) and a good germanium diode is at ~0.3V.

The only diode test I have available is a measure of resistance (same space on my multi-meter as continuity, Craftsman 82141) and I haven't been able to decipher it as of yet. When is look online at data sheets they report forward voltage drop as such:

Diode Vf (V) If (mA)
1N270 1 200
1N277 1 100
1N191 1 5
etc.

This doesn't look a value like a ~0.3V like I was expecting. Could one of you kindly help me to understand what these values mean and how I can convert it to a forward voltage drop value?

Thanks for all your time thus far,
Admiral

If it is still not working, please try the following suggestions:

Put the tuning coil 1-2 cm above the board;

Replace the wooden board with some plastic (perf board is good);

Replace the tuning coil with a new one as suggested in post #3; do not connect the other end of the tuning coil to ground;

Check the impedance of your headphones.

What does the diode test show with your actual diode?

Diode test is running at low current, e.g. 1 mA which gives you around 0.3 V with germanium or schottky and 0.6 V with silicon diodes.

After reading your response section titled "1.", I am now under the impression that whatever adjustable metal contact system is used(wiper or taps) in the coil cannot be connected directly to ground. So could you please clarify where this proposed alligator clip should connect to in my circut?

Click to expand...

Ok, I'll explain. The coil of wire is an inductor, the tuning control is a capacitor. When a signal is received, even an incredibly tiny one, it starts the circuit resonating, the signal voltage starts a kind of 'see-saw' where the inductor builds up a magnetic field and the capacitor stores an electrostatic charge alternately. There is a natural speed it want to do that for every combination of capacitance vale and inductance value. Mathematically the formula is "frequency = 1 / (2 * pi * sqrt( L * C ))" where L is the coil and C is the tuning capacitance. The resonant frequency is the one you want to receive the radio station on, other frequencies are lost, rather like they push or pull the ends of the see-saw out of step with each other.

To select a different resonant frequency you therefore have to change the tuning capacitance or change the inductance. The capacitor is already variable by the way it is made and it would normally be your main tuning adjustment. Making the inductor adjustable as well isn't too important but it does allow you to extend the range the circuit can resonate over, in other words a wider tuning range.

The inductance (L) depends on several things, mainly the diameter of the coil, the number of turns of wire and what type of core is inside it. You appear to be using a ferrite core which is fine. When you move the wiper, you select how many turns are being used and therefore the inductance, the remaining turns are still there but essentially disconnected and not playing any part in the inductance. Where you have a problem is that instead of selecting how many turns are being used, you a have it wired so it shorts out part of the coil. The schematic is wrong - you followed its mistake when you built it! If you short out the coil, that tiny magnetic field is also shorted out and it ceases to work properly. To fix it, disconnect the wire at the bottom of the coil completely and leave it unconnected. The wiper will then select the number of turns being used instead of shorted out.

This is a photograph of part of a crystal radio I used for teaching a long time ago, it isn't the best of photographs but you should be able to see the 'taps' in the coil. The big metal thing is an ancient variable capacitor. Your wiper would become a flying wire to clip to the tap points.

(That got me thinking of who the last pupil was. It was the mother of someone who is now an Edaboard contributor, I'm getting old!)

The diode type is fairly important but there seems to be a lot of misconception about how it works. You want one designed for use at high frequencies, the kind used in power adapters are not suitable and you want one with a low turn-on voltage. A low Vf is better than a high one but in a crystal receiver the 'mA' in the Vf rating is only going to be around 0.01mA for a strong radio station and much less for a weak one so the published figures aren't much use. Don't confuse the 0.3V and 0.7V figures as being good or bad, the voltage levels off at those figures as the current is increased but that effect doesn't start until MUCH higher currents are involved. What matters is how the diode works at tiny currents and in general Germanium and Schottky types perform better. Galena works, I have a crystal set made in 1920 here which still works but it can take hours to find a 'good' spot on the Galena to probe it - much easier to buy a diode you know will work out of the box!

Brian.

This is a photograph of part of a crystal radio I used for teaching a long time ago

Click to expand...

Ha! I vaguely recall the coil specs; we used 100 turns of enamelled wire on a 1 inch dia paper tube with taps every 10 turns. We could tune in MW stations -at least three stations- using the capacitor tuning. I could not get any SW (they are weak) but the excitement was great. The headphone must be very high impedance - they are also difficult to get now- most common headphones will not work with this circuit. The tuning cap seen in the photo is an excellent one- still available at some places- but the one I used was a very flimsy and poor quality. But it worked!

A moving ferrite might be better than a wiper on the
coil.

Did the radio finally work?

it seems the coil is insulated copper coil. if it insulated copper the metal plate touches on the center will not work as variable inductance as per the diagram. please check the inductance of coil with meter. correct me if i am wrong

What is the value of the variable cap(Gang) that is connected in the circuit ?

Is it an AM or AM/FM Gang ? If it is second type did you connect it to AM tags ?

Give the coil details (No of turns /dia . wire ).