This is a recovered file. The images in this post may be out of order, and there may be duplicates.
Detector Sensitivity Tests - The RadioBoard Forums Post by gzimmer » Wed Jan 04, 2017 Hi, For a long time now we've been discussing (and arguing about) the efficiency of the Diode Detectors in Crystal Sets. The difficulty in actually measuring the efficiency lies in the large number of variables: The size of the antenna, the degree of Selectivity, the Tuned Circuit loading, the sensitivity of the Headphones, and many others. So I've been thinking about a test set which could reliably measure the relative Sensitivity of different detectors, while eliminating the variables. I'm looking for definitive answers for the following questions: - Which diode types are the most sensitive? - What circuit impedances do they prefer? - Can battery bias improve Sensitivity? - How do Shunt Detectors Compare? - Do Voltage-Doubler and Full-Wave Rectifier circuits work? - Do passive FET detectors give much improvement? - Do Zero Bias diodes offer any improvement? - Best 50 Ohm Detectors? - The effectiveness of un-tuned detector windings. - The Capacitor in series, "floating diode" circuit. - Compare the efficiency of the old "no battery" circuits. On weak signals our diodes operate in the Square Law mode. This means that the detection efficiency will vary greatly as the signal strength changes, and on weak signals the Efficiency will be very low. So for meaningful tests, we must measure the efficiency at some constant and weak signal strength. When using my RCA "Big Can" headphones, the weakest audio that I can easily hear is approx 2mV (into 2K impedance), so we will use 2mV as our reference. -------------------------------------------------- A Crystal Set can only produce as much power as is available from the Antenna, because the Radiation Resistance of the Antenna acts as a Constant Power source. Likewise in this set, the 50 Ohm termination resistor in the generator (along with the attenuator itself), strictly limits the power which can be drawn. The maximum power transfer will only occur when the load is also 50 Ohms. So in this test, each set of impedances is selected in turn and the tuning is peaked. Whichever of those impedances best matches the diode to the input (and output), will cause the power in the load to be at maximum. Plus the attenuator setting (to give 2mV out) will then give the relative sensitivity of that detector. The more attenuation which can be inserted, the greater the Sensitivity of that particular detector. And because the loading on the tuned-circuit is relatively constant, the Q will also be constant, thus removing yet another variable. ------------------------------------------------ My approach was as follows: Using an accurate AM signal generator, I feed the signal (1MHz, 100% modulated with 1KHz, at 500mV) via a switchable attenuator, into a simple Crystal Set with a 50 Ohm input link. The Toroid tuning coil has twelve taps selected via a rotary switch, and the impedance of each Tap was measured by shunting the tap with a decade resistance box to find the value which gave half-power. This gives the source impedance of each tap (and which by definition, is transformed to a 50 Ohm impedance at the input). Then the test diode feeds a multi-tapped Audio transformer, again via a rotary switch. The impedance of the taps on the audio transformer have also been calibrated. An output tap at 2K Ohms is terminated with a 2K resistor (via a LPF) and an AC millivolt meter is used to measure the output level. Here's my rough prototype: The method is simple: Connect the diode, peak the tuned circuit to resonance, then adjust the Attenuator to give 2mV out, while adjusting the input and output impedance switches to give a maximum (not forgetting of course to peak the tuning after each adjustment). The setting on the switched Attenuator then gives the relative Sensitivity of each Detector (the higher the attenuation, the better). So, bottom line: By adjusting the input level (via the attenuator) to give a constant output level, the test set can accurately measure the relative Sensitivity and also the preferred impedance (in and out) for each type of Detector Diode. It was very pleasing to see how stable and repeatable the measurements are. It's easy to see the difference between different diodes, and even differences between individual diodes of the same type. More shortly .............Zim