Would this work (and some other questions)?

[Resistanceisfutile]

I'm interested in building a crystal radio with an amplifier and was wondering if this circuit would work. If so, where do I connect -9V (do I connect it at the ground markings?)

 

[ted]

It might work, at least "sort of". However, it is hardly too optimum way of building an amplified crystal receiver.

I say "sort of", because there are some questionable aspects in the circuitry. Without some modifications, not even "sort of"....

Regarding the power supply connections: I am afraid this is not a good configuration at all, as the OPAMP output can never swing below the negative rail. Therefore you can't connect the minus pole of 9V to ground, if you want output to be able to reproduce the negative half-part of the AC waveform. The easy fix would be to use two batteries, one +8V and other -9V towards the ground, to feed the amplifier.

Another is lack of RF filtering after the diode. I would at least connect a small capacitance from cathode of diode to ground.
The second is the configuration of amplifier. It is essentially a transimpedance amplifier (converting current to voltage). That might work well. However, it is a bit unusual configuration in this context, and therefore I would analyze and simulate it before "buying" it.


Btw. for analog simulations, a program called LTSPICE is free and pretty intuitive to use for this kind of circuitry. You can download it from Linear Technology (https://www.linear.com/designtools/software/#LTspice)

 

[Audioguru]

The "opradio" from Arron Cake's site has a few errors:
1) A lousy old 741 opamp DOES NOT WORK when its inputs are within a few voltage from ground. One input is at ground and the other input is almost at ground. The inputs should be biased at half the supply voltage.
2) The detector diode needs a DC load (a resistor to ground at its output).
3) The detector diode needs an RF filter capacitor to ground at its output.
4) The opamp is inverting with an extremely low input impedance that shorts the tuned circuit. The opamp should be non-inverting with a high input impedance.
5) The supply voltage needs a filter capacitor to ground.

I fixed the horrible circuit about 10 years ago:

 

[ted]

I absolutely agree, that 741 would not work at all with one input to a voltage rail. However, not even an op-amp with rail-to-rail inputs would work in configurations, if the biasing doesn't allow some headroom both at inputs and outputs. Therefore, the bias must be arranged in-between the rails, as you say. This circuit would try to force its output below the minus rail (here drawn to be ground). And that would not work. Ever.

You make a good point in most, if not quite all, of the weaknesses of the design. However, at least in theory, your point 2) is not quite correct, as the circuit has a DC load through the feedback resistor. Also your point 4) is not obvious, as the circuit as described would work as a current-to-voltage converter for the detected signal (but has to have at least the operating point / the bias fixed, of course).

The "opradio" from Arron Cake's site has a few errors:
1) A lousy old 741 opamp DOES NOT WORK when its inputs are within a few voltage from ground. One input is at ground and the other input is almost at ground. The inputs should be biased at half the supply voltage.
2) The detector diode needs a DC load (a resistor to ground at its output).
3) The detector diode needs an RF filter capacitor to ground at its output.
4) The opamp is inverting with an extremely low input impedance that shorts the tuned circuit. The opamp should be non-inverting with a high input impedance.
5) The supply voltage needs a filter capacitor to ground.

I fixed the horrible circuit about 10 years ago:

 

[Audioguru]

The OP does not say his country.
Crystal radios were used in North America and Europe about 80 years ago when there were very few AM radio stations. Now there are so many stations in our countries that the single tuned circuit in a crystal radio will have a few or more stations playing at the same time (poor selectivity), or if the Q of the tuned circuit is very high then a single very muffled station will be heard with no high audio frequencies (much worse than the "normal" muffled sounds from a REAL AM radio).

A real AM radio has AGC so that a weak distant station sounds as loud as a strong local station. A crystal radio has no AGC.