1
$\begingroup$

issue: trying to make a crystal radio, doesn't work right

components:

  1. diode pulled from a working crystal radio set
  2. variable air-gap capacitor gang (VC), 125/250/250 pF in parallel, 40-625 pF total
  3. variable inductor (VL), probably 13-55 uH
  4. variable resistor (VR), 0-117k ohms
  5. 50 foot long wire antenna
  6. high impedance earphone made for crystal radio sets
  7. various wires

method:

I have a diagram of a simple crystal radio.

enter image description here

  1. antenna connected to one end of the VL, ground connected to other end.

  2. connected in parallel to that, is the VC.

  3. connected in parallel to that, is the VR, but one wire has the diode connected in series.

  4. connected in parallel to that, is the earphone.

problem:

I only hear the 2 strongest AM stations in the area. one of them, around 560 kHz, varies in volume with adjustments to VL and VC, which indicates the circuit is being tuned. but I can't get any other radio stations and the other strong station is always heard in the background.

the diode works in either direction.

VR gets 0 volume at 0 ohms, optimal volume around 10k, more than that has no change, maybe 70-117k it sounds a little choppy.

it gets quite sufficient volume with VL ~ 13 uH and VC ~ 40 pF. I can change L and follow the tuning by adjusting C to compensate, but with a loss in volume. apparently that was optimal and anything else reduces volume, to nothing at extremes. but where VL= 13 uH and VC= 40 pF, f= 6.98 MHz, not 560 kHz.

if I put VL on 55 uH the volume goes to about 0 no matter what VC is set to.

why can't I receive any other stations? why are L and C not what they should be?

$\endgroup$

1 Answer 1

1
$\begingroup$

why can't I receive any other stations?
why are L and C not what they should be?

Strong stations likely drive the detecting diode beyond its square-law region: to the point where it acts as a rectifier (fully turning off on an RF peak). This is an efficient detector.
For weaker signals, the diode is conducting for the whole RF cycle: passing more current on one peak, while passing less current on opposite peak - this is square-law operation. Not so efficient.


OP's antenna is shorter than 1/4 wavelength, and adds a capacitive reactance to the tuning "LC", which is not accounted for.

55uH seems somewhat small for "L". To resonate with 625pf, at least 129 uH is needed to reach 560 kHz. It is difficult to account for the antenna capacitance, because it is tapped into the "LC" resonator. Coupling coefficient is unknown.

Many of these air-wound coils are wound tightly, so that each turn butts-up to the next turn. While compact, inter-winding capacitance might be significant. This capacitance adds to the 625 pf, and may limit the highest frequency of resonance.
Inductance of this coil might be measured at low frequency - perhaps 1kHz. At much higher frequency, inter-winding capacitance becomes significant, making the "apparent inductance" larger...at even higher frequency you reach self-resonant-frequency. You want to construct a coil whose self-resonant frequency is significantly higher than the upper limit of AM broadcast band.

Some coils are wound in a manner that separates one winding from the next, helping to distribute currents in each turn evenly around its perimeter. This also reduces inter-winding capacitance, but results in a bigger coil than more common close-wound coils.

$\endgroup$
3
  • $\begingroup$ the VL is not a wound coil but a plunger-type pulled from a working radio. it has what appears to be a small ferrite bar going into a metal cylinder, long range of travel adjusted by the tuner with a worm gear. measuring with LCR meter showed something like 13-55 uH I forgot exactly. won't be the full L because it's only a tuner component in a system. depending on how the radio works it might add L in other circuits, or the C it uses actually does require a range of only 13-55 uH. $\endgroup$
    – Fyodor
    Commented Oct 17, 2023 at 15:15
  • $\begingroup$ if those numbers are accurate, the circuit should be able to tune down to 860 kHz. but it's not working at any VC or VL, except minimums of both where 560 kHz comes in strong, which should be 6.98 MHz, not 560 kHz. $\endgroup$
    – Fyodor
    Commented Oct 17, 2023 at 15:22
  • $\begingroup$ If max signal peaks at the end-of-travel of L or end-of-travel of C, then LC resonance is at a frequency that differs from the signal you're listening to. Your original description seems to me describing true resonance: signal strength maximizes part-way along L's travel, and also signal strength maximizes part-way along C's travel. Plunging L's core further into the coil should require C's plate meshing to be less, so as to restore maximum signal. It is difficult for me to justify your LC measurements...many LC meters have difficulty with these small values - I would try using other methods. $\endgroup$
    – glen_geek
    Commented Oct 18, 2023 at 13:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .