Am radio from The Art Of Electronics Lab Course not working.

[mjn2024]

I recently picked up the book learning the art of electronics along with the components and equipment needed to complete the course. In lab 3l.4 you try to design and build a very simple am radio. The book does not give a lot in terms of help, so I was basically on my own to design it. Going off of the few hints the book gives I designed this schematic AM Radiomk1.pdf (14.92 kB - downloaded 63 times.), however, when I built it I get nothing but static even when I swap values I just get static. I am guessing I either am using the wrong values or I laid it out wrong(or both). I am using a decent length extension cord with an alligator connected to the exposed end and the other end I put out my window and through my backyard. for GND I am using the earth terminal on my PSU and for a speaker, I am using a generic pair of headphones.

[TimFox]

That circuit looks very wrong for a diode detector, and the resonant frequency of 2.53 nF and 10 mH is 31.6 kHz, far from the MW band.

[bob91343]

Do some googling for crystal radios and try a different circuit.

[Kerlin]

There are so many things wrong with that circuit that they are to numerous to list.
As a fourteen year old I built radios that didn't work. When I look back on them now I think the designers should be ashamed of the errors they made and the disappointment they caused. Lucky there were a few that did work, thank you Gilbert Davey, I became an RF and Embedded uP engineer.

This one is the best book ever written on building crystal radios just for fun.
https://worldradiohistory.com/BOOKSHELF-ARH/Technology/Boys-Book-Of-Crystal-Sets.pdf
Interesting and fun reading.

[iMo]

Mind for a crystal radio you want a good antenna and grounding, see the "Boys.." book above..
Also the headphones should be high impedance ones - at least 1-2kohm..

[Paul Ed]

Hi,
Although simple to make a working crystal radio set, they can actually be quite a challenge to make a really good version - indeed there are (or were?) various competitions held to compete with new variant builds with the aim (usually) either the most station received or the furthest stations heard - or a combination of both. Am not sure if such competitions are still held, but a basic crystal radio set is still easily do-able.

For a design that is more or less guaranteed to work, try this one:

https://www.petervis.com/Radios/making-a-crystal-radio/crystal-radio-circuit.html

(Peter's other pages/article are also worth a browse to help get you started)
This might be read in addition to the book(PDF) already posted, which is great but a tad dated now, so this post was just to help you actually have a working basic design which you can then tweak and change to see how it might be improved.

Oh, and don't be disappointed that your first set wasn't great, it's actually harder now to hear stations on such a set as there are fewer (and often much lower power) AM stations to listen to  (certainly that holds in Europe, not sure about the US).

Good luck!

[BrokenYugo]

Plenty of AM broadcasting still in the US, mostly religious or right wing pollical stuff but it's there, along with some more corporate news/sports/talk 50kW "clear channel" stations scattered around the country that are impossible to miss.

Crystal sets demand a lot to effectively give something for nothing, a long high antenna and proper ground (I guess a big loop antenna would work too), proper high sensitivity earphones (which can be made, e.g. http://makearadio.com/misc-stuff/piezophones.php), good diode selection, and a lot of attention to construction detail (maximizing Q, minimizing impedance mismatch, etc.), it's a hobby of it's own to get much performance from one.

EDIT: This guy does some good crystal radio content. https://www.youtube.com/user/MIKROWAVE1/videos

[floobydust]

OP that circuit is pure crap, it would never work. A tank circuit with a rectifier diode across it?
SPK1 cannot be a speaker, although historically it was always a "crystal earphone" - after the detector diode. R1 at 15k ohm ensures the output signal is too weak to drive anything but a microphone input on something and C3 at 20pF seems useless.

Just search for crystal radio circuits and you'll see what is popular. Don't follow the lab course example.
The hardest part is finding high-impedance earphone or headphones, they are uncommon nowadays.

[Terry Bites]

That aint gonna work.  Its a bad design from the get go. H&H is good but has lots of little errors.

[Neomys Sapiens]

Nearly any haphazard interconnection of a diode and a resonant circuit will exhibit some AM demodulating action, if blasted with some amplitude at the right frequency. But I would not call that an radio.

[floobydust]

Um, the tank is 10mH with 2,730pF... including the Schottky's huge ~200pF.
This means it's tuned to dumbness at 30.5kHz 

[rsjsouza]

This one is the best book ever written on building crystal radios just for fun.
https://worldradiohistory.com/BOOKSHELF-ARH/Technology/Boys-Book-Of-Crystal-Sets.pdf
Interesting and fun reading.

Thank you for the link. My dad always talked about a book he had whan a kid about galena radios - it seems very similar to this one.

Oh, and to the OP: indeed the circuit is terribly bad. 100 turns of enameled wire around a 1 cm^2 of a ferrite core; a variable capacitor for AM (30~400pF), a germanium diode (OA95), a small ceramic capacitor and a crystal earphone (or a dynamic earphone plugged on a small audio output transformer) and the radio is done. That obviously tied to a very long antenna and a solid ground connection.

[Kerlin]

The Boys Book of Crystal Sets is a book I have known as the best for decades.
Its not just a link I just googled and posted, like so many do on forums, whats the point of doing that ?
The first one is the easiest one, just use any broadcast band coil.
Number eight worked best for me, as a kid hand wound the Pi-wound coils my self.

It lead me to a career in radio design and onto Senior Avionics RF Engineer, so I really owe that author, W.J. May.
Currently R.F. engineers that can come up with practical results are hard to find and well paid, very easy to get a job.

A good thing to do for a hobby is the get a full Ham radio license, great fun collecting old radios too.
I obviously found RF very addictive.

[bob91343]

I'm with you, Kerlin.  My own resumè parallels yours in many ways.  Ham license, BSEE, some rf work.  Ended up owning a retail biz repairing stereo and such.  Now a demented hobbyist with more high end test gear than any sane person could afford (at original prices).

At the age of 19 I designed and built a kit for a one tube regen AM radio for my gf to assemble in HS science class; she loved the look on the teacher's face when it actually worked.  I still have the radio and 1N5GT tube.  That was 70 years ago.

[Kerlin]

Bob very glad you still have the radio.
Please post a pic of it.
But do you still have the girl ?
If not I am sure you have fond memories of her.
Maybe you got the radio and the girl?

[vu2nan]

Hi @mjn2024,

Here's a conventional schematic.



A 60' long, 30' high horizontal wire would suffice for the antenna.

The coil could be wound on a 2" former (on-line calculators would be of help to determine the number of turns).

For tuning, an air-spaced variable capacitor would be desirable. Should it not be readily available, a 'PVC gang capacitor' could be tried.

A pair of high impedance headphones would be preferable. Sensitive 'balanced armature' headphones would be the best.

Good luck!

Nandu.

[Ian.M]

Traditionally, high impedance headphones were 'moving iron', with a ferromagnetic diaphragm in the field of a permanent magnet with a very high turn count fine wire coil wound round it.  The coil resistance was typically several K, and if wired in series, the total impedance of the headphones could exceed 10K.   Back in the day, (which is now close to a century ago) they were typically the most expensive part of a crystal set, and are now vanishingly rare antiques and priced accordingly.  Also you need expert knowledge to distinguish them from the far more common WW2 era military headphones without doing a DC resistance check so the odds of getting lucky and finding a set other than in the estate sale of a deceased ham are minimal.
 
Modern 'high impedance' headphones are an order of magnitude or more lower impedance and DC resistance, and are optimized for quality not sensitivity and are generally totally unsuitable for use with a crystal set.

Crystal (piezoelectric) earpieces are essentially capacitive i.e. open circuit at DC so will not function in a classic crystal set circuit without a resistor across them (try 10K to 100K) to provide a discharge path for the signal current from the detector diode, so they can follow the modulation envelope rather than just hold at the peak voltage from the detector.

[timenutgoblin]

That circuit looks very wrong for a diode detector, and the resonant frequency of 2.53 nF and 10 mH is 31.6 kHz, far from the MW band.

Is there any possibility that it's just a font formatting error during printing? Perhaps L1 should read 10µH (10 micro henries) instead of 10mH (10 milli henries). The resonant frequency would then be closer to 1MHz (1000.6 kHz) instead of only 31.6kHz 

[VooDust]

man I would love to try that out.

in this golden age of consumerism surely there's an amazon link to a complete set....?

[TimFox]

That circuit looks very wrong for a diode detector, and the resonant frequency of 2.53 nF and 10 mH is 31.6 kHz, far from the MW band.

Is there any possibility that it's just a font formatting error during printing? Perhaps L1 should read 10µH (10 micro henries) instead of 10mH (10 milli henries). The resonant frequency would then be closer to 1MHz (1000.6 kHz) instead of only 31.6kHz

Traditional crystal sets and other MW AM receivers used variable tuning capacitors with max 365 pF capacitance (and roughly 35 pF minimum).  His schematic probably mis-located the 2.2 nF capacitor and diode.  Note the drawing in reply 15 above.

[Ian.M]

That circuit looks very wrong for a diode detector, and the resonant frequency of 2.53 nF and 10 mH is 31.6 kHz, far from the MW band.

Is there any possibility that it's just a font formatting error during printing? Perhaps L1 should read 10µH (10 micro henries) instead of 10mH (10 milli henries). The resonant frequency would then be closer to 1MHz (1000.6 kHz) instead of only 31.6kHz

The lab in question does NOT give a schematic - its up to the student to figure it out.  It refers to the 3L.1.1 tank circuit of a 10mH inductor in parallel to a 10nF capacitor, fed by a 100K resistor from a signal generator, but says:

• to detect the carrier, use an LC circuit like the one you built at the start of this lab, but showing the following differences:
  • the resonant frequency should be around 1 MHz;
  • you need no upper resistor in the "divider:" the antenna can drive the LC directly.
  
  
• The value of the resistor to ground is not critical; try 10k;
   
• the low-pass filter's job is to kill the carrier, keep the audio. Fortunately, these two frequencies are very far apart; so, you have a lot of freedom in placing f_3dB. The form of the low-pass you design may strike you as odd (though this depends on the way you choose to do the task: the "odd" configuration, described in §3N.8.4, Fig. 3N.36 uses the rectifier's resistor to ground as the R in the RC low-pass). Just make sure to put this in time-domain terms: that RC is very long relative to the period of the 1 MHz "carrier," but short relative to the "signal" or "audio" period.

Fig. 3N.36 suggests a leaky peak detector, i.e. a diode feeding a parallel RC.

The student is thus expected to choose parts from the parts kit for a 1MHz resonant frequency.  Unfortunately there are only two inductors in the kit to choose from: 10mH and 100uH both radial unshielded ferrite bobbin construction, so the student  is expected to realize that the 2.53pF required to resonate an ideal 10mH inductor at approx. 1MHz is unrealistic, or at least note that the smallest capacitor in the kit is 10pF, and thus choose the 100uH inductor  which will need a much more achievable 253pF, given the parts available (5 each 10pF, 68pF, 100pF, 470pF).

Unfortunately as the circuit is only tunable in discrete steps (unless you get cute and try two of the 100uH inductors in series and vary their coupling by varying the angle and distance between them), the book misses the key step of actually checking there is a strong local AM radio station transmitting near 1MHz!  Also it doesn't mention that direct coupling the antenna will detune the tank circuit due to the antenna capacitance, so less capacitance will be required than expected.

The addition of a ferrite rod and a spool of magnet wire, and even a variable capacitor  to the kit would have been essential if the course went into RF design in any detail but as this was the only radio experiment, it would be hard to justify the several bucks extra cost.

[timenutgoblin]

That circuit looks very wrong for a diode detector, and the resonant frequency of 2.53 nF and 10 mH is 31.6 kHz, far from the MW band.

Is there any possibility that it's just a font formatting error during printing? Perhaps L1 should read 10µH (10 micro henries) instead of 10mH (10 milli henries). The resonant frequency would then be closer to 1MHz (1000.6 kHz) instead of only 31.6kHz

Traditional crystal sets and other MW AM receivers used variable tuning capacitors with max 365 pF capacitance (and roughly 35 pF minimum).  His schematic probably mis-located the 2.2 nF capacitor and diode.  Note the drawing in reply 15 above.

My browser appears to be blocking some images and not others. I noticed when using a different browser the schematic you refer to appears without issue.

[BrokenYugo]

man I would love to try that out.

in this golden age of consumerism surely there's an amazon link to a complete set....?

Only modern mass market kit I've seen is a really cheap/crude one (tuned with a sliding tap on the inductor) marketed towards kids.

The fun here is in the tinkering and building anyway, build your own kit. Go get a magnet wire assortment, a couple 365 or 500 pf air variable tuning caps (can be salvaged from beat up tube era radios), build your headphones from the plans at the link I gave earlier, and have fun. I've had good luck getting 1N34A ish germanium diodes (look to be surplus) very cheap from Aliexpress, just buy the ones where the picture is of glass package point contact diodes, not the smaller glass packages that look suspiciously like small signal Schottkys, because they are and schottkys do behave differently. The rest of the parts are junk (board offcut for the base, polypropylene bottles make for good high Q coil forms, etc.) or easily sourced modern components. From there you can also build other early radio designs for AM broadcast band and HF, regen, reflex, etc.