Radio LC RC help.

[perfect_disturbance]

Ok so I decided to build a basic AM radio and attached is my schematic so far.

But I've gotten stumped selecting values C3 then for L1 and C5 (upper and lower values).

So here is what I've tried for C3 I know I need to pick a value that in coordiation with R3 at 100k will filter out the 540khz to 1700khz carrier but leave the < 20khz audio wave intact.
to calculate this I turned to

fc=1/(2*3.1415*R*C)

where
fc = 50000 Hz
R = 100000 Ohms

I get C = 3.18319 F or C= 0.031pF

Which seems wrong as I'm intuitively expecting something closer to 31uF

Next problem is picking L1 and C5 (lets just look at the top  of the range).

For this I turned to fr = 1/(2*3.1415*sqrt(L*C))
where
fr = 1700000 Hz (I think 1700kHz is the top of the am band right?)
L = 0.000150 H (I read a random page on the internet that 150uH would give me good selectivity)

and I get ... C = 0.00000000005843 F or .05843 pF which again seems way wrong from what I've read this should be more like 365pF.

I know I'm and idiot so please nail me just as long as you tell me what the freak I'm screwing up.
Also feel free to mock and insult (productive criticisms please) the rest of the circuit.

Thanks

[ElectroIrradiator]

Well, firstly then you don't need nor want audio all the way up to 20 KHz, as the AM stations limits their audio bandwidth to about 5 KHz (depending on where you live). Not that a single RC filter will provide much attenuation, but worth keeping in mind.

Secondly, the 100K load resistor for the diode is a bit on the large side. I'd suggest 10K instead. When I plug 5 KHz and 10K into the first formula, I get C3 = 3.2 nF, which sounds about right.

You are right about the top of the AM band being up around 1.7 Mhz, and 150 uH should work for L1. Your result is correct, you will get 58E-12 F, which is 58 pF, not the low value you mention.

The audio amp doesn't have much gain, and it won't be able to drive a speaker. You will need some very sensitive headphones to hear much of anything. A simple receiver like this won't have much selectivity, but that may actually be an advantage, if the sensitivity isn't all that hot.

Depending on the antenna you use, you probably want to connect it to the top of L1 through a small capacitor, like 100pF or even smaller still. This may help a bit with the low selectivity.

[Paul Price]

Completely remove the regulator. Use a 9V battery + to VDD pin - to VEE pin of op-amp.

Any small-signal transistors with >=.6amp Max IC will work, the higher the Beta the better.

A germanium diode will give you the best sensitivity, 1N82, 1N34, 1N60 or the base-collector junction of a PNP germanium transistor will do.  The capacitor values shown can be shifted upwards quite a bit. A 1000uF10V cap across the battery might also do good.

The small capacitor value connecting to the antenna and to post-diode filter help selectivity and sensitivity. See attached schematic.

[perfect_disturbance]

Thanks for the help I have a few more questions.

OK so the  variable cap should have a range of 50pF to 680pF?

I was planning on just 6 feet of 30 awg wire as the antenna. I was also thinking of using the opamp(maybe a ts922?") to  power a set of headphones.  I read somewhere that the output of the antanne would be about 1vpp  I thought If I boost it to double that would be 2vpp and 1.4 rms appx? my headphones have a resistance of 32 Ohms and I figured that would be 61mW RMS or 30ish mW per channel which according to here
http://electronics.stackexchange.com/questions/28404/what-is-the-voltage-range-of-a-standard-headphone-jack-from-a-phone and here
http://www.head-fi.org/t/97419/what-is-voltage-output-of-ipod-headphone-jack
seemed about right.
If not how much input signal/gain should I figure for?

Next if I only need to drive the output signal to under 2 volts peak to peak do I need to drop the regulator I thought it would help me get more even output over the lifetime of the battery.

And I had considered a germanium diode but I couldn't find any at mouser.  I saw somewhere else where thy had used a silicon or Shottky diode with a low forward voltage. Such as a 1N5711, or a 1N4150TAP if they have a low forward voltage is there any other advantage to a germanium diode?

Thanks for your help.

[Paul Price]

The voltage regulator circuit is not needed and will decrease the voltage output to the speaker.
The LM358 works without a negative voltage voltage supply.

The regulator circuit you've drawn has the ground pin connecting to the - of the battery, this would mean you are always wasting >7mA of quiescent current to power the regulator. That's more than 60mW of wasted power. So if the regulator wastes power and is not needed for the circuit to work best, it should be eliminated.

With a standard loop antenna coil and the AM band, that tuning capacitance is about right.

The circuit I recommended, the two transistor complementary pair will work just using the LM358 to drive a headphone or a speaker at low volume.

Input signal will vary with the transmitter power and your distance from the transmitting antenna, but the voltage P-P will always also depend on the Q of the resonant circuit and selectivity and it is important for it not being heavily loaded by a too long antenna or  by the detector circuit, so adding a 18pf cap in series with the antenna and will actually increase Q of the resonant circuit and the p-p voltage of a received signal. You can always play around with larger cap values, maybe up to 100pf.

A Schottky diode 1N5711 would be an excellent detector diode for this circuit, better than germanium diodes.

[vk3yedotcom]

Others have discussed the audio section so I'll concentrate on the RF bit. 

The project is effectively an amplified crystal set so anything written about those will apply to your radio.

I was planning on just 6 feet of 30 awg wire as the antenna.

Bear in mind that AM radio operates at low frequencies (compared to almost all other radio and TV) and the wavelengths are very long.  That means the antenna wire needs to be very long to pick up sufficient signal.  6 feed is way too short.  Instead make it 60 feet and preferably install it outside as high as possible on a chimney, pole or tree.

An earth is desirable and will greatly improve volume.  This can be metal strip stuck into the ground or possibly a cold water tap (unless your plumbing is plastic).

The wiring of the antenna and diode to the coil needs to be different to that in the circuit, especially if you have more than one station in the area.  As it stands the diode will load down the coil and you won't be able to select between stations - all will appear as a jumbled mess. 

The same is also true of the antenna connection.

The way to fix this is to have taps on the coil made evenly along it every few turns.  The diode goes to one alligator clip while the antenna goes to another.  This way you will be able to vary the tapping point to best seperate stations. 

If you have the antenna and diode near the ground end of the coil you will be able to select stations but volume may be low. But if you have it near the top volume may be high but stations could be jumbled.  So it's a compromise and may vary with each station you tune.

This video explains a bit about crystal sets

 

[JackOfVA]

I would start by reading Ben Tongue's web pages on the design of crystal receivers. http://www.bentongue.com/xtalset/xtalset.html

With respect to the input LC and diode detector, it may be advantageous to use a  tapped inductor and connect the detector diode at a lower impedance point, so as to increase selectivity. Although this will reduce the voltage due to the transformer action, it will improve selectivity and increase overall Q which offsets some of the transformer voltage division.

The post detector filter need not be all that elaborate, usually a simple RC filter is enough since the modulating  frequency is relatively low compared with the RF carrier frequency. 

[perfect_disturbance]

I've made some changes how does this look and attached a board layout.

[vk3yedotcom]

I've made some changes how does this look

Still not a fan of the front end.

The 100pF in the antenna line may improve selectivity but will reduce volume. Especially at the 530kHz end of the band.

You really do need taps on the coil and for the settings to be determined by trial and error.

This is both for the antenna and diode connections. 

Alternatively a primary winding for the antenna might work better.

There are so many variables (antenna length, position, earth, nearby stations etc) and with such a weak front end you cannot afford to lose a microvolt of signal or suffer poor selectivity though a badly configured circuit.

[Paul Price]

Again, the voltage regulator circuit is not needed and will decrease the voltage output to the speaker.
The LM358 works without a negative voltage voltage supply.

Using a 7805 would mean you are always wasting >7mA of quiescent current to power the regulator. That's more than 63mW of wasted power. So if the regulator wastes power and is not needed for the circuit to work best, it should be eliminated.

Forget about tapped coils..too much trouble and not needed. If you couple your antenna to the coil with a small capacitor(10-47 pF), you can connect either a long or even a short antenna to the tuned circuit with good selectivity.  If you are reasonably close to any AM radio station, this should work fine.


If you want to increase selectivity, buy a ganged-dielectric capacitor used for superhet receivers and add a second coil. Couple the fist coil to the antenna and to the second coil with a 10-47 pf capacitor and  you will have twice the  good selectivity between stations.

The dual-ganged capacitors are made for superhet receivers and they have different in max capacitance because it is designed to work with a 455KHz freq. difference of the oscillator from the received signal that makes a superhet work. Most come with small trimmer capacitors built in to adjust the tracking of the the two tuning circuits of a superhet.

You mayl need to have at least one  loopstick antenna that can be tuned with its adjustable core or else a add a second fixed capacitor across the smaller of the ganged capacitor so that the two capacitors/coils track(resonate) with each other at the same frequency of the station you want to listen to. Adjusting the trimmer capacitors built into the variable capacitors should make fine tuning a snap.