Trouble with getting AM receiver to work

[CGandC]

Hello,
I have assembled the following AM radio reciever circuit:

Purpose of the circuit: To receive AM radio signals and provide the output audio of these signals ( demodulated ).

Circuit elements: L ~ 1.5mH , C ~ 300pF , Amplifier IC: LM324N , Diode: 1N4004 ,
The speaker is small and it's resistance is ~ 5 ohm
There are 2 variable resistors, to adjust the amplification.
The antenna is a long wire
The voltage provided to the amplifier is 9V DC from battery ( not drawn )

My expectations from this circuit:
- the antenna would receive a signal
- the bandpass filter would filter the signal so the output signal's ( to the amplifier ) frequency spectrum is centered at ' Fc '
- the filtered signal would get amplified by the amplifier which is connected in negative feedback configuration
- the signal's negative voltage would get truncated by the diode
- the signal would pass through the speaker to the ground and thus I would hear AM radio station broadcast.

Experimental results
- There is no voltage drop on the speaker at all and no current flows through it, thus it is mute.
- Most voltage drops on resistor R2

Notes:
- the central frequency of the bandpass filter is Fc=1/sqrt(LC) ~ 1.49MHz
- the ground in the circuit is the cathode of the 9V battery

My question is:
Is this circuit wrong? , if not, then what do you propose as being the fault to the failure of the system?
If it is wrong, then why?

Thanks in advance for help.

 

[berkeman]

Some quick thoughts...

• The LM324 is pretty slow to get much gain out of it at 1MHz (AM band)
• Even though the input and output range of the LM324 includes ground, you will get better gain if you bias up the input (AC couple the signal) and AC couple the connection to the speaker
• It seems your rectifying diode should be before the opamp, not after. In that case the LM324 would be fast enough to amplify the audio frequency AC envelope signal...

 

[tech99]

It might be useful if you could draw the power supply connections.

 

[berkeman]

Circuit elements: L ~ 1.5mH , C ~ 300pF

I'm used to an AM band of 0.5 to 1.7MHz. Why did you choose those LC values? What are the frequencies of the AM band you are wanting to tune to?

Also, you will get a lot better performance if you make the LC tank circuit variable, so you can tune to the desired AM radio frequency selectively.

https://en.wikipedia.org/wiki/Broadcast_band

 

[berkeman]

Also, you will get a lot better performance if you make the LC tank circuit variable, so you can tune to the desired AM radio frequency selectively.

Quiz Question for @CGandC -- Are you familiar with how you would implement each of these methods for a tuned LC circuit in your AM receiver application?

http://www.learnabout-electronics.org/ac_theory/images/Tuned-cct.gif

 

[the_emi_guy]

Issues I see:

OP's formula for frequency of tank circuit is wrong.

Circuit will require 700mV of RF signal to overcome the diode drop and deliver a signal to the speaker. No way one amp will get us up to that level. The diode needs to be biased to its knee. One way to do this is using an active half wave rectifier (op amp with 2 diodes inside feedback).

5 ohms is a pretty low impedance to try to drive.

I suggest OP download and learn LTSpice, simulate first, then build. It can be a real pain to try to get something like this working by just hacking at it. It's a lot more satisfying getting it to work in simulation then having it work reasonably well once built. Start with a generous 1mV antenna signal (like you are close to a strong AM station and you your L is wound around a big oatmeal box).

 

[berkeman]

the central frequency of the bandpass filter is Fc=1/sqrt(LC) ~ 1.49MHz

Agreed -- OP, what is missing from this equation?

 

[CGandC]

Quiz Question for @CGandC -- Are you familiar with how you would implement each of these methods for a tuned LC circuit in your AM receiver application?

http://www.learnabout-electronics.org/ac_theory/images/Tuned-cct.gif
View attachment 217636

Yes, then i'll have to order such elements as to make ' Fc ' variable.

Issues I see:

OP's formula for frequency of tank circuit is wrong.

Circuit will require 700mV of RF signal to overcome the diode drop and deliver a signal to the speaker. No way one amp will get us up to that level. The diode needs to be biased to its knee. One way to do this is using an active half wave rectifier (op amp with 2 diodes inside feedback).

5 ohms is a pretty low impedance to try to drive.

I suggest OP download and learn LTSpice, simulate first, then build. It can be a real pain to try to get something like this working by just hacking at it. It's a lot more satisfying getting it to work in simulation then having it work reasonably well once built. Start with a generous 1mV antenna signal (like you are close to a strong AM station and you your L is wound around a big oatmeal box).

Can you please provide a schematic? it would help me very much.

Agreed -- OP, what is missing from this equation?

You are right, I was calculating the Angular frequency ω not 'Fc' turns out I had to divide my frequency by 2*pi.

So I chose L=40µH and kept the same capacitor value ( Fc ~ 1.45MHz) , yet , no audio is heard.
just to elaborate, I'm using an inductor that looks like a resistor... but I saw everyone use an exposed copperwire coiled around some material , is it wrong to use such an inductor ( which I use ) when attempting to pick up signals?

 

[berkeman]

So I chose L=40µH and kept the same capacitor value ( Fc ~ 1.45MHz) , yet , no audio is heard.
just to elaborate, I'm using an inductor that looks like a resistor... but I saw everyone use an exposed copperwire coiled around some material , is it wrong to use such an inductor ( which I use ) when attempting to pick up signals?

That inductor looks pretty lossy. Can you measure its resistance with your DVM?

Also, have you moved the diode and added bias to your opamp circuit yet? If not, I'm not surprised that you are not getting an audio signal yet.

The inductor and capacitor used in such a tuned circuit often look like these:

https://upload.wikimedia.org/wikipedia/commons/6/6c/Variable_Capacitor.jpg

https://sc01.alicdn.com/kf/HTB19xLt...stable-IFT-inductor-coils-slug-tuned-coil.jpg

 

[TurtleMeister]

I would move the diode to the front end for detection, as berkeman suggested, and use the op amp for audio amplification. Also use a germanium diode (such as the 1N34A) instead of silicon.

 

[berkeman]

Can you please provide a schematic? it would help me very much.

Where did you get the schematic that you drew? If you just drew it yourself as something you wanted to try, that is fine of course.

You can use Google Images to find circuits for simple AM receivers. Have you tried that yet?

 

[TurtleMeister]

Here's a close match to your circuit (if you move the diode to the front):

http://www2.ensc.sfu.ca/~ljilja/ENSC220/Labs/lab_5.html

 

[berkeman]

Here's a close match to your circuit (if you move the diode to the front):

http://www2.ensc.sfu.ca/~ljilja/ENSC220/Labs/lab_5.html

Nice! And for @CGandC -- note how that opamp uses split supplies, If you only have the single 9V supply, you should still make a center-bias circuit for the opamp. Or else add a second 9V supply (battery?) to give you split supplies to make things easier.

 

[tech99]

Nice! And for @CGandC -- note how that opamp uses split supplies, If you only have the single 9V supply, you should still make a center-bias circuit for the opamp. Or else add a second 9V supply (battery?) to give you split supplies to make things easier.

I have built similar circuits for measuring field strengths, and these are my personal experiences.
If the diode is placed before the amplifier, R1 is not helping anything. The diode already has very high resistance to small signals, and can be connected direct to the inverting input. R2 can be raised to a high value to obtain max gain. Typically 1M.
We do not want DC across the speaker so a balanced battery supply, as mentioned by berkeman, is a good idea. An op amp will work quite well with 32 Ohm earphones or can operate a voltmeter to indicate output.
I would also mention that, depending on distance from the transmitter etc, the antenna might need to be several metres long, and the circuit will require a ground connection, maybe to water pipes etc.
For MF reception, the tradition inductor was about 50 turns of wire 1 1/2 inches diameter and the variable capacitor was 500pF max. It is also possible to wind the inductor on a ferrite rod or to wind it as a multi-turn loop. A few turns of say 12 inches diameter.

This is a note about the operation of the circuit I have described:-
In operation, the capacitive reactance of the wire antenna is resonated by an inductance created by the parallel combination of L and C. This steps up the resistive part Ra of the antenna impedance, which is mainly loss resistance and might be 100 Ohms or so, to a high value equal to Ra x Q^2, where Q is essentially the ratio of antenna reactance to resistance.This creates maximum voltage, required to drive maximum current through the high resistance of the diode. In other words, the LC circuit tends to match the antenna impedance to the very high diode resistance. The current in the diode, Id, flows into the virtual ground at the amplifier inverting terminal, and the amplifier creates an output voltage equal to Id x R2. So that if 1 uA of diode current flows, and R2 is 1M, this gives 1 volt out.

 

[AlexCaledin]

How about trying a serial LC at the input?

It seems interesting because the input resistance of the amplifier can easily be made negative by adding some part of the opamp output voltage to its " + " input, making the receiver regenerative.

 

[tech99]

How about trying a serial LC at the input?

View attachment 217860

It seems interesting because the input resistance of the amplifier can easily be made negative by adding some part of the opamp output voltage to its " + " input, making the receiver regenerative.

Where is the detector in your circuit?

 

[AlexCaledin]

Where is the detector in your circuit?

why, it's the transistor, just notice the big capacitor to the emitter. The emitter junction works like any diode.

 

[berkeman]

why, it's the transistor

Very subtle if true.

Can you share a SPICE simulation that shows it? I tried to visualize it when you first posted the circuit, but didn't see it right off the bat...

 

[berkeman]

Also, what's the GBW product of that opamp? (I'm too lazy to look it up, sorry)

 

[AlexCaledin]

Also, what's the GBW product of that opamp? (I'm too lazy to look it up, sorry)

800 MHz

 

[berkeman]

800 MHz

Holy smokes! I really need to see your simulation of this circuit. 800MHz into a few-MHz transistor into a speaker...

At least send me your SPICE deck for this circuit so I can play with it please...

 

[AlexCaledin]

few-MHz transistor

2N2369 is more than 500MHz, chosen for low input capacitance.

 

[berkeman]

Look, you're not a newbie here. Post the plots please. How does the envelope end up at the speaker?

 

[TurtleMeister]

How about trying a serial LC at the input?

View attachment 217860

It seems interesting because the input resistance of the amplifier can easily be made negative by adding some part of the opamp output voltage to its " + " input, making the receiver regenerative.

That's an interesting circuit. Is it your design, or where did you get it? This takes me back to my younger days when I would spend many hours experimenting with simple am receiver designs. However, I didn't have IC chips or SPICE. Just a junk box.

 

[Averagesupernova]

@AlexCaledin
I don't see that the tank circuit is a series resonant circuit. It may appear so but the right side of the inductor is at virtual ground. Circulating currents will behave as if a parallel resonant circuit. Interesting setup. An antenna hanging out in the wide open liable to pick up who knows what concerning low frequency noise is tied right to an input of an opamp. I have a hard time believing the transistor will detect anything.

 

[tech99]

why, it's the transistor, just notice the big capacitor to the emitter. The emitter junction works like any diode.

It looks to me as if the transistor is biased for linear operation, so that the B-E junction will not rectify as it is forward biased.
The base is at half supply voltage and the emitter is connected to -ve via 1k. This looks like normal common emitter operation for a linear amplifier stage.

 

[AlexCaledin]

It looks to me as if the transistor is biased for linear operation, so that the B-E junction will not rectify as it is forward biased.
The base is at half supply voltage and the emitter is connected to -ve via 1k. This looks like normal common emitter operation for a linear amplifier stage.

- again, think of the 100 (or better 1000) microF capacitor at the emitter. It makes the emitter potential practically constant during a typical audio wave period. Therefore, the collector current is an exponent of the radio wave, the average is the audio signal. The only disadvantage here is that the detector is not linear at strong signal or strong noise ( - but the thing is, where I live, the MW signals are all very small, so anyway my detector is doomed to be just as quadratic as any simple detector at small signal).

(The other thing is the emitter resistor, it must of course be much more than 1kOhm if the load is an old high-resistance headphone.)

 

[AlexCaledin]

@AlexCaledin
I don't see that the tank circuit is a series resonant circuit... antenna hanging out...

- sorry, I was actually thinking of using a big coil with ferrite working as magnetic antenna (with regeneration), in that case the "tank" capacitor should be grounded and tuned, the coil to the opamp.

In case of a big electric antenna, the constant capacitor ought to be between antenna and the opamp input, the coil grounded and tuned.

<- a classic tuning inductor (two coils in series)

 

[CGandC]

Thanks for all the help everyone :)
Due to shortage of time to try out all of these ( likely ) solutions to my problem, I think i'll settle for now.
If I'm unfortunate again after I have had made up some time to attempt these propositions to my mishap, I will open another thread to readdress the issue.
So any mentor/admin reading this... you can close the thread.

 

[berkeman]

I will open another thread to readdress the issue.
So any mentor/admin reading this... you can close the thread.

No need to close the thread. If you want to come back later with more questions about this subject, feel free to keep posting in this thread.