Hello, and thank you MrChips.

First post, really, apart from this, which I link here to shorten the introduction to the present one. I apologize beforehand if this post is long, or is a well covered subject here, but I am still finding my way around. They won't be so long in the future!

So, crystal radios. In brief, so far;

Mk 1 did not work. As expected.

Mk 2A did not work, either. New coil, 2" diameter with 26 gauge wire, using a biased LED as a diode. Made an earphone from a small, salvaged piezo speaker out of a junked appliance, and used non-polarized ceramic capacitors to attempt to tune. I think I heard some static at one stage, but not sure.

Mk 2B worked wonderfully. I found some idea out on the internet about the Mystery Radio that uses two coils, so I reworked the Mk 2A coil to include a second, coupling coil on the same diameter (note, no taps). By this time my germanium diodes had arrived in the mail, also, as had a small variable capacitor and a "proper" piezo earphone. Reception and volume of my targeted AM station were good, showing that the math was working (I prefer doing the math manually rather than using the abundance of coil calculators I have found on the internet, otherwise I am not learning anything). However, there was some lack of selectivity, as I could hear some other stations simultaneously in the background. Whatever the case might have been, it convinced me the Mystery Radio was the way I wanted to go.

But I wanted to do it with a difference; concentric coils. I later found that this idea has been applied with success before. Here are the specs of the Mk 3A;

Coil 1;
71 winds of 22 gauge on a 3" diameter, with 4 intermediate taps. Originally, I wanted to make this around a 350 uH coil, but there was some imprecision in the wire gauge - it was salvaged from a defective starter motor solenoid and may be British Standard gauge (when I measured it later with a micrometer it turned out to be 28.7 thou)- and this resulted in the coil being a bit longer than desired, lowering the calculated inductance to 315 uH, exactly, funnily enough.

Coil 2;
50 winds of 26 gauge on a 2" diameter, with 2 intermediate taps. Wanted 140 uH, got 136 uH on calculation of the finished product, for similar reasons as above.

I assembled these two coils concentrically. I was thinking sort of "transformer", if that applies. See the diagram. Apologies for that schematic, I have no circuit drawing software and did it on Inkscape. The antenna, incidentally, is 55 feet of 18 gauge, insulated, solid copper wire. Straight, tapped from one end.

Okay, results and questions;

On first testing, I got a slightly weaker reception of the signal than the Mk 2B, but much better selectivity, especially when using the diode taps. Strength was increased to a par with the Mk 2B by adding the 47K resistor to earth before the earphone (I have no idea why 47K appears to be the magic number, but all references I have looked at that have a resistor seem to point to this value). Some signal "saturation", if that is the right term, was improved by using the antenna taps. I worked around the limited selectivity of the variable capacitor by placing different value ceramic capacitors in parallel with it, and can get the whole AM range this way.

Then, as an experiment, I hooked it up using only the outer, 3" diameter coil as the sole source, disregarding the inner coupling coil altogether (I will call it the 3B). The reception volume was phenomenal, but the strangest thing is that it works better (better selectivity) without the ground connected at all! Connecting the ground results in many stations being received simultaneously, with no apparent improvement in strength. Weird and unexpected.

Why can I do away with the ground in this configuration? It makes no sense to me.

Finally, though I do not want to fully spoil the joy of experimentation, I want to improve my Mk 3A. Would, perhaps, closing the gap between the two coils produce better reception strength? For example, if I wound a 140 uH coil on a 2.5" former and placed this concentrically within the 3" coil?

My thanks again, wonderful forum, some excellent reading material that I am endeavoring to read piecemeal, and not least, my thanks in advance for any responses to this post.

 

What form of antenna are you using, as a rule a long wire antenna is needed.
Max.

 

Hello,

For good functioning of the crystal radio a good ground is a must.
How long is your antenna?
You will need an antenna of several meters to recieve anything.

The following page of the EDUCYPEDIA has a lot of links on the crystal radio sets:
http://educypedia.karadimov.info/electronics/circuitsrfcrystal.htm

Bertus

 

Yes, it is a long post indeed, isn't it? I can understand you not wanting to read it! My apologies, again.

What form of antenna are you using, as a rule a long wire antenna is needed.
Max.

I fully agree; good, sound advice. However, the reception is good, as described in the post, and I am confident that the antenna, also described in the post, is adequate. The questions were about something else.

Hello,

For good functioning of the crystal radio a good ground is a must.
How long is your antenna?
You will need an antenna of several meters to recieve anything.

The following page of the EDUCYPEDIA has a lot of links on the crystal radio sets:
http://educypedia.karadimov.info/electronics/circuitsrfcrystal.htm

Bertus

Thanks, useful links.

One of the questions is exactly on the subject of grounding.

How come my radio variation, as described in setup 3B, works really well (better, in fact) without a ground.

The other question;

With concentric coils, is bringing the detector's coil closer in to the reception/tuner coil's field a good idea to improve signal strength?

 

Hi, Will, welcome to the forum! I am not the one to answer your questions, I'm just starting out again myself. I learned some electronics while getting a bachelor' degree in electrical engineering. I designed and built a few radios back in the 90s then lost interest and now I'm getting back into it.

I don't know why grounding your radio detunes it except that it seems to somehow be adding a large amount of capacitance which swamps your tuning capacitor.

When I was in the dorm at WSU and I had just learned the principle of Am demodulation, I set up a breadboard with a detector like yours and fed it into an op amp to drive a small 8 ohm speaker. For an antenna I used a straight wire. Reception was poor until I hooked that antenna wire to the heat radiator, then it was loud. I couldn't tune it but we had a dominant station coming in from Spokane that mostly drowned out the others. I told my professor about it and he was as confused as me. He said the wire must have been acting as a ground not an antenna.

 

Hi Will, PRS

Experimenting with crystal radio circuits is a fun and entertaining way to get back into electronics.

@Will - when you refer to MK1, MK2A etc. it would be nice to see the circuit schematics so that we can follow your successes and failures.

I looked at the Mystery Radio circuit and with the bit I know about crystal radios the circuit makes no sense to me.

Hence I cannot comment on the functioning of the circuit without a ground connection on the primary coil. As mentioned somewhere, there must be a capacitance coupling through the coil to the ground.

The crystal set I am currently working on is of the simplest design:

My antenna is a simple horizontal long wire about 30m long. The ground is to a cold water copper plumbing.

It is a good idea to experiment with your component values. The 100kΩ resistor from the diode to ground seems to work best for me.

In a second version, I have the antenna and ground connected to the tuning capacitor and the coil as a primary winding. Then I have a secondary winding that slides inside the primary winding. I can adjust the coupling between the two coils by sliding one tube over the other.

With this primary/secondary setup, the selectivity is phenomenal, much better than I have ever achieved in my younger days as far as I can recall.
Sorry, I don't have a current circuit diagram. It looks like this one shown below except that the primary coil is also grounded. (Interesting that this circuit shows the primary not grounded as well. Hmm, I will have to test that idea.)

Here is the link to the site of that circuit shown above:

http://www.bristolwatch.com/radio/crystal_radio.htm

Experiment and have fun.

Later I will show you how I amplify the RF signal with a simple MPF102 FET.

 

The diode radios we could afford as kids, before we had transistor radios, worked best connected to the finger stop on dial telephones.

Don't know why.

 

Maybe you weren't listening to radio? You were just tapping the phone lines?

 

Yes, it is a long post indeed, isn't it? I can understand you not wanting to read it! My apologies, again.

Do not worry for that at all.

Doing otherwise, sure would get you a request for being more clear on whatever you post. Such is life!
Crystal radio; a pending matter for me.

 

The diode radios we could afford as kids, before we had transistor radios, worked best connected to the finger stop on dial telephones.

Don't know why.

Really? Diid it work with any rotary disk?

Were those stops connected to something like ground or...?

 

Really? Diid it work with any rotary disk?

Were those stops connected to something like ground or...?

Mom wouldn't let me take the phone apart.

Heck, in those days, there were no phone privileges til high school.

 

Were those stops connected to something like ground or...?

On mine it isn't, ground that is.
Max.

 

Thank you, PRS and MrChips.

Before anything else, you both mentioned amplification methods. That is precisely the subject that rekindled my interest in electronics a few months ago, as I eventually have a desire to construct my own amplifier for a musical instrument. Some tentative experiments - which I won't detail here - have been done already (TEA2025 for my computer's audio output), but it is all in its infancy with me, at the moment. I will certainly be all eyes for reading and learning anything about the subject. Thank you, again.

Crystal radio seemed a fun and appropriate way to start, however.

Details on my other circuits, as per request.

Mk 1:
70 turns of 30 gauge on a 1" former for 100 uH, 2 x 220 pf ceramic capacitors in parallel to tune 760 Khz, a close local station for testing. I would say it is a "standard" crystal radio circuit with no intermediate taps (much like the one on MrChips post, with fixed value capacitors instead of a variable one), so no diagram is needed, really. I initially made a trial "diode" out of a segment of rusted box cutter blade and a shaped carbon brush from a motor. I attempted to listen to it through a computer speaker, but I am pretty sure it fell apart there, if not at the "diode", as the input (I imagine) is calibrated for a different impedance. I judged it a failure, but wait!

I revived it this morning out of curiosity, rigging it with the 1N34A diode and the piezo earphone and - what do you know - it works! It pulls in a variety of other stations, too, so the selectivity is not so good. However, that is now the Mk 1B, for my reference notes. I will try some antenna tapping to the coil on it, next.

The Mk 2A was not much different, only on a larger diameter coil, as described in the first post. 88 turns of 26 gauge on a 2" former for (approximately) 325 uH, 2 x 68 pf ceramic capacitors in parallel (for receiving the same frequency as above). The novelty was the attempt to use a green LED salvaged from an old CD-ROM drive as a diode. I am not at all sure I biased the LED correctly to function as a detector diode, so there is a diagram attached of this circuit for scrutiny.

@PRS: That is just it. This crystal radio hobby appears to tread a border between art and science, far as I can tell. There must be some descriptive formulae for these phenomena of apparent synergy or interactions of capacitance and inductance floating around, but yet to be put on paper (?).

@MrChips: I am completely defeated in trying to understand how that Mystery Radio primary coil sets up a current with no ground, unless of course as you say, it is getting some form of "ethereal" grounding from the coupling of the secondary coil. Weird. I don't get it. I have tried measuring the current in these coils with a multimeter, but nothing registers, not even on the grounded coils. Must be less than milliamps.

That air variable capacitor is the sort of thing I need, but cannot locate them anymore (a neighbor is offering me a broken, old B/W TV that has two, but I am loathe to accept it solely to cannibalize. More likely, it will be another project!). I am going to make one of those capacitors, soon. Any suggestions on the best materials to use? I understand that copper sheet, and even aluminum, is good. I do have some galvanized mild steel sheet, which I was thinking of employing for the purpose. I am assuming (unfortunately) at this stage that a more "robust" capacitor like that might better resist the swamping PRS mentions.

@atferrari: Thanks! And I was going to ask inwo the same sort of question. It made me go and look up analogue telephone system operation!

 

Mom wouldn't let me take the phone apart.

Heck, in those days, there were no phone privileges til high school.

Funny that. I actually did take a phone apart at age 8, about the same time I made my radio kit. My father, who was an engineer himself, was quietly amused. My mother was furious.

 

Will, I have never seen a radio like your Mk2A. What is the purpose of the 3 volts across the diode? Does it work without it? Does it work only as long as it's forward biased?

 

As a kid I actually owned a 'real' cats-whisker radio, the chunk of Galena had to be tickled with the whisker.
Max.

 

Will, I have never seen a radio like your Mk2A. What is the purpose of the 3 volts across the diode? Does it work without it? Does it work only as long as it's forward biased?

Good questions; it did not work at all, initially. It was before I had the 1N34A.

But back to the subject. Does the LED works as a detector diode? Yes, it does! Without the 3 VDC, you actually do get a very faint signal off the Mk 2A. Adding the 3 VDC in parallel across the LED, with the resistor as per diagram, seems to "over-activate" the diode, and you get nothing. Putting another resistor (47 KΩ) in series brings the whole set to life in a very acceptable fashion.

The parallel bias circuit that works (let me clarify this. The LED circuit itself is series. It hooks into the radio circuit parallel);

3 VDC --> 47 KΩ --> 1 KΩ --> LED --> |

The 1 KΩ resistor stays in not because I want 48 KΩ specifically, but because I soldered it to the LED anode to protect it.

The results seem to point to a requirement to have the minimum current possible flowing through the LED for it to act as a detector. I will try this with a 100 KΩ potentiometer; might be able to regulate to get the best reception strength.

And yes, it only works forward biased.

EDIT: Hang on a moment. I was not fully understanding what you meant by forward bias. I thought you meant reversing the battery polarity on the LED itself. Sorry. I now understand that you meant turning the whole LED circuit around. Just tried it. Yes, it does work. Thanks!

As a kid I actually owned a 'real' cats-whisker radio, the chunk of Galena had to be tickled with the whisker.
Max.

I understand that iron pyrite works, too. I have a geologist friend who might be able to procure a sample for me, though I am not sure how eagerly the neighbor's cat is going to contribute to this experiment!

 

This one that I was just looking at, on the other hand, appears not to ground the detector circuit, as far as I can see, which is the complete opposite of the mystery radio.

The article does answer my question about proximity of the primary and secondary coils, however. Post it in case someone else finds it interesting, too.

 

The tuning coil will receive some RF, not as much as a long wire combined with a good ground, but if you have strong radio signals at your location then the tuning coil alone will run a radio.

The phase of the received RF between the tuning coil as an antenna and the long wire/ground antenna system may be destructively interfering with each other, which could explain why connecting a ground weakens the received sound.

Tuning coils as an antenna have some directionality - try rotating it, both horizontally and vertically, and see how your signal changes. You may find a sweet spot where the grounded tuning coil gives the greatest reception for a particular station. Or by rotating the tuning coil you may be able to reduce the strength of an interfering station, another way to improve the sound of a selected station.

Something I've learned about radio reception is that almost nothing is as simple as it seems like it should be. That wire that you connect to ground and call ground is itself a length of RF receiving wire, and can supply some amount of RF energy to your crystal set. Even a wire connected to Earth itself doesn't automatically show zero RF energy at the surface of the earth, depending on the salt content and moisture in the soil, some length of wire below ground can still have RF energy on it.

 

tapping the detector and antenna down on the coil helps the "Q" of the tuned circuit, puting the detector on the end of the coil loads down the circuit, making it even more broadbanded. to try how much to tap, bring out taps on the coil and try them in different combinations.