Experiments with a detector unit

Back to the index

Introduction

On this page measurements are described on my detector unit 1 .
In this detector unit is as diode used the HSMS-282K (2 diodes parallel), and the load resistance by  transformer unit 1  is 1.6 MΩ.
For the choice of diode and load resistance I assumed that maximum sensitivity would occur if load resistance and diode resistance were about equal to the impedance of the unloaded LC circuit ( this assumption is however not correct).
Via some experiments I will now determine if the sensitivity and selectivity can be improved by the use of another detector diode and/or another load resistor.
Also frequency shift of the detector circuit is measured as a function of the voltage across the circuit. Test setup used for the measurements on this page.The oscilloscope measures via the measuring amplifier the voltage across the LC circuit. The voltmeter measures  the detected DC voltage. The load resistor is formed by resistor R parallel with the voltmeter with 10 MΩ input resistance. The value of capacitor C is: 100 nF. The RF signal is supplied by a DDS signal generator, the signal is via a coupling coil coupled to the LC circuit.

All measurements are done at a frequency of about 1500 kHz.
At 1500 kHz, the impedance of the unloaded detector unit is: 1.97 MΩ.
There are 4 (combinations of) diodes tested.
HSMS-282K Schottky diode ( 2 diodes in one case parallel)
HSMS-282K Schottky diode ( 1 diode)
5082-2835 Schottky diode ( 2 diodes parallel)
5082-2835 Schottky diode ( 1 diode)

Detected voltage at constant input voltage.

In this measurement, the detected DC voltage is measured, so the voltage at the output of the diode.
The diode is in all cases loaded with a load resistor of 1.52 MΩ, this is a combination of a 1.8 MΩ resistor parallel to the voltmeter of 10 MΩ.
The value of 1.52 MΩ is however a little bit lower then the impedance of transformer unit 1.
The input voltage (RF signal across the LC circuit) is constant (20 mV, 50 mV, 100 mV etc.) when measuring the different diodes.

 Voltage across LC circuit HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 20 mV 0.4 0.2 0.4 0.2 50 mV 2.2 1.2 2.6 0.9 100 mV 8.4 5.5 9.5 4.3 200 mV 29.1 23.5 32.3 19.2 500 mV 130 121 137 112 1 V 330 319 340 314 2 V 780 772 813 728 5 V 2170 2180 2210 2170
Table 1

Detected voltage (mV dc) as function of the voltage across the LC circuit (peak-peak value across the LC circuit).

Below 200 mVtt input voltage these diodes work in the square law detection region.
When doubling the input voltage, the output voltage increases about 4 times.

Above 2V input voltage, we more come into the linear detection region

When using 2 diodes 5082-2835 parallel, the highest DC voltage is measured.
The (single) diode 5082-2835 gives the least output voltage.
From this, one might conclude that the use of 2 diodes 5082-2835 parallel gives the most sensitive receiver.
And the use of a single 5082-2835 the least sensitive receiver.
This conclusion is however not correct, as we shall see later in this article.

Q factor at a constant input voltage.

In this measurement the Q factor is measured of the (by diode and load resistor) loaded circuit.
The diode is in all cases loaded with a load resistor of 1.52 MΩ.
De input voltage is constant (50 mV, 100 mV etc.) for the different diodes.

 Voltage across LC circuit HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 20 mV * * * * 50 mV 442 600 395 601 100 mV 429 536 376 578 200 mV 385 469 358 518 500 mV 306 333 289 341 1 V 273 300 259 284 2 V 264 283 247 278 5 V 247 250 236 251
Table 2

Q factor as function of the voltage across the LC circuit.
The diode is loaded with 1.52
MΩ.

* There is not measured at 20 mV input voltage, this voltage is too low for an accurate Q measurement.

The use of 2 diodes 5082-2835 gives the lowest Q in this measurement.
A low Q means the LC circuit is (relative) heavy loaded by the diode.
A low Q also means a lower selectivity of the receiver.

Frequency change of the LC circuit as function of the input voltage.

Every diode has a certain capacitance (capacitor value in pF).
This capacitance is depending on reverse voltage across the diode, the higher the reverse voltage, the lower the capacitance.
At zero volt the capacitance is the highest.
A diode connected to a LC circuit shall rectify the RF signal, the rectified voltage forms a reverse voltage across the diode.
The higher the voltage across the LC circuit, the higher the reverse voltage across the diode, the lower the capacitance, and the higher the resonance frequency.

 Voltage across LC circuit HSMS-282K 2 diodes parallel HSMS282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 20 mV 0.0 0.0 0.0 0.0 50 mV 0.0 0.0 + 0.2 + 0.1 100 mV 0.0 0.0 + 0.3 + 0.2 200 mV + 0.1 0.0 + 0.5 + 0.3 500 mV + 0.9 0.6 + 1.5 + 1.0 1 V + 2.4 1.0 + 3.1 + 1.4 2 V + 4.2 2.0 + 4.8 + 2.5 5 V + 6.2 2.8 + 6.8 + 2.9
Table 3.

This table gives the measured frequency shift (in kHz) of the LC circuit, as function of the voltage (peak-peak value) across the LC circuit.
The resonance frequency of the circuit is 1500 kHz.
The load resistor is 1.52 MΩ in this measurement.

Detected voltage and power at a constant magnetic field (large signal).

In the previous measurements, the voltage across the LC circuit was constant, when switching from one diode to another.
This corresponds however not with the situation in practice.
In a real crystal receiver, the voltage across the LC circuit shall depend on both used diode and load resistor.
If the LC circuit of a receiver is loaded more heavy by diode or load resistor, the voltage across the circuit shall reduce, and also circuit Q shall reduce.

In the next measurement (table 4) the LC circuit is placed in a constant magnetic field.
We get such a constant magnetic field by placing the coupling coil at a fixed distance from the LC circuit, and keep the amplitude of the signal generator constant.
The constant field simulates a received station of constant strength.
The load resistor is varied between 0.91 and 10 MΩ.
The voltage across the LC circuit is now depending on both load resistor and used diode, but the level of several volts across the circuit compares to reception of quite a strong local station.

Table 4
Voltage across the LC circuit (AC) and detected voltage (DC) as function of load resistor and used diode.

 Load resistor HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 0.91 MΩ DC= 0.546 V  AC= 1.48 V DC= 0.561 V  AC= 1.58 V DC= 0.562 V  AC= 1.54 V DC= 0.564 V  AC= 1.55 V 1.07 MΩ DC= 0.625 V  AC= 1.65 V DC= 0.644 V  AC= 1.72 V DC= 0.643 V  AC= 1.70 V DC= 0.647 V  AC= 1.70 V 1.30 MΩ DC= 0.721 V  AC= 1.87 V DC= 0.745 V  AC= 1.95 V DC= 0.742 V  AC= 1.90 V DC= 0.747 V  AC= 1.94 V 1.52 MΩ DC= 0.790 V  AC= 2.00 V DC= 0.819 V  AC= 2.19 V DC= 0.815 V  AC= 2.07 V DC= 0.821 V  AC= 2.08 V 1,80 MΩ DC= 0.867 V  AC= 2.18 V DC= 0.900 V  AC= 2.27 V DC= 0.981 V  AC= 2.20 V DC= 0.900 V  AC= 2.26 V 2.13 MΩ DC= 0.966 V  AC= 2.38 V DC= 1.003 V  AC= 2.48 V DC= 0.992 V  AC= 2.40 V DC= 1.005 V  AC= 2.49 V 2.48 MΩ DC= 1.055 V  AC= 2.58 V DC= 1.094 V  AC= 2.66 V DC= 1.079 V  AC= 2.60 V DC= 1.097 V  AC= 2.68 V 3.19 MΩ DC= 1.191 V  AC= 2.85 V DC= 1.243 V  AC= 2.96 V DC= 1.222 V  AC= 2.90 V DC= 1.243 V  AC= 2.97 V 3.97 MΩ DC= 1.318 V  AC= 3.10 V DC= 1.376 V  AC= 3.21 V DC= 1.350 V  AC= 3.17 V DC= 1.370 V  AC= 3.21 V 4.38 MΩ DC= 1,362 V  AC= 3.19 V DC= 1.428 V  AC= 3.33 V DC= 1.395 V  AC= 3.26 V DC= 1.424 V  AC= 3.34 V 5.00 MΩ DC= 1.424 V  AC= 3.30 V DC= 1.498 V  AC= 3.49 V DC= 1.467 V  AC= 3.40 V DC= 1.495 V  AC= 3.49 V 6.67 MΩ DC= 1.566 V  AC= 3.60 V DC= 1.657 V  AC= 3.80 V DC= 1.614 V  AC= 3.70 V DC= 1.657 V  AC= 3.80 V 10.0 MΩ DC= 1.743 V  AC= 3.98 V DC= 1.852 V  AC= 4.20 V DC= 1.775 V  AC= 4.10 V DC= 1.843 V  AC= 4.20 V

The given voltage across the LC circuit (AC) is the peak-peak value.
The voltage across the circuit is proportional to the loaded Q of the LC circuit.
So, the higher the AC voltage, the higher the Q, and the more selective the receiver is.

We see the voltage across the circuit and the detected voltage are not very depending on the used diode, all diodes give about the same result.
But AC voltage is very depending on the load resistor (at equal diode).
This is a typical symptom at high signal levels, where the diode is working in the linear detection region.

In the following table, the detected voltages from table 4 are calculated to power in the load resistor.

Table 5
Power in the load resistor (in nW = nano-Watt) as function of the load resistor and used diode.

 Load resistor HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 0.91 MΩ 328 346 347 349 1.07 MΩ 365 388 386 391 1.30 MΩ 400 427 424 429 1.52 MΩ 411 441 437 443 1.80 MΩ 418 450 441 450 2.13 MΩ 438 472 462 474 2.48 MΩ 448 483 469 485 3.19 MΩ 445 484 468 484 3.97 MΩ 438 477 459 473 4.38 MΩ 423 466 444 463 5.00 MΩ 406 449 430 447 6.67 MΩ 368 412 391 412 10.0 MΩ 304 343 315 340

Graph 1
Power in the load resistor (in nW) as function of the load resistor and used diode.
The results from table 5 are shown in this graph. Conclusion:

The highest detected power occurs at a load resistor of about 2.5 to 3 MΩ.
As diode, a single HSMS-282K and a single 5082-2835 give the best results.
Connecting 2 diodes parallel gives a little bit less output power, but the differences between are small.
Once again: this is the situation when receiving strong signals.

Detected voltage and power at a constant magnetic field (small signal).

The same measurement from table 4 is done once again, but now with a much smaller signal level, so with a lower amplitude of the signal generator.
This signal level compares to reception of a very weak station.

Table 6
Voltage across the LC circuit (AC) and detected voltage (DC) as function of the load resistor and used diode.

 Load resistor HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 0.91 MΩ DC= 6.2 mV  AC= 98 mV DC= 6.5 mV  AC= 119 mV DC= 7.2 mV  AC= 111 mV DC= 6.6 mV  AC= 138 mV 1.07 MΩ DC= 7.1 mV  AC= 99 mV DC= 7.5 mV  AC= 120 mV DC= 8.2 mV  AC= 112 mV DC= 7.8 mV  AC= 139 mV 1.30 MΩ DC= 8.1 mV  AC= 99 mV DC= 8.8 mV  AC= 121 mV DC= 9.4 mV  AC= 113 mV DC= 8.9 mV  AC= 140 mV 1.52 MΩ DC= 9.0 mV  AC= 100 mV DC= 9.7 mV  AC= 122 mV DC= 10.3 mV  AC= 115 mV DC= 9.9 mV  AC= 141 mV 1.80 MΩ DC= 9.8 mV  AC= 102 mV DC= 10.9 mV  AC= 124 mV DC= 11.3 mV  AC= 116 mV DC= 11.0 mV  AC= 142 mV 2.13 MΩ DC= 10.9 mV  AC= 104 mV DC= 12.3 mV  AC= 126 mV DC= 12.6 mV  AC= 117 mV DC= 12.3 mV  AC= 143 mV 2.48 MΩ DC= 11.9 mV  AC= 105 mV DC= 13.5 mV  AC= 128 mV DC= 13.8 mV  AC= 118 mV DC= 13.7 mV  AC= 144 mV 3.19 MΩ DC= 13.4 mV  AC= 106 mV DC= 15.6 mV  AC= 130 mV DC= 15.6 mV  AC= 120 mV DC= 15.8 mV  AC= 146 mV 3.97 MΩ DC= 14.8 mV  AC= 108 mV DC= 17.5 mV  AC= 132 mV DC= 17.4 mV  AC= 122 mV DC= 17.9 mV  AC= 147 mV 4.38 MΩ DC= 15.3 mV  AC= 109 mV DC= 18.2 mV  AC= 134 mV DC= 17.9 mV  AC= 124 mV DC= 18.8 mV  AC= 148 mV 5.00 MΩ DC= 16.0 mV  AC= 109 mV DC= 19.3 mV  AC= 136 mV DC= 18.9 mV  AC= 126 mV DC= 20.1 mV  AC= 150 mV 6.67 MΩ DC= 17.7 mV  AC= 110 mV DC= 21.9 mV  AC= 138 mV DC= 21.2 mV  AC= 128 mV DC= 22.9 mV  AC= 152 mV 10.0 MΩ DC= 19.8 mV  AC= 110 mV DC= 25.3 mV  AC= 140 mV DC= 24.0 mV  AC= 130 mV DC= 27.2 mV  AC= 156 mV

We see the voltage across the circuit and the detected voltage are rather depending on the used diode.
But the AC voltage as function of the load resistor (at equal diode) is just relative constant.
In other words: the circuit Q (which is proportional to the AC voltage) is mostly depending on used diode and not so much on load resistor.
This is a typical symptom at low signal levels, where the diode is working in the square law detection region.

Also now, the detected voltages are calculated to power in the load resistor.

Table 7
Power in the load resistor (in pW = pico-Watt) as function of the load resistor and used diode.

 Load resistor HSMS-282K 2 diodes parallel HSMS-282K 1 diode 5082-2835 2 diodes parallel 5082-2835 1 diode 0.91 MΩ 42.2 46.4 56.9 47.9 1.07 MΩ 47.1 52.6 62.8 56.9 1.30 MΩ 50.5 59.6 68.0 60.9 1.52 MΩ 53.3 61.9 69.8 64.5 1.80 MΩ 53.4 66.0 70.9 67.2 2.13 MΩ 55.8 71.0 74.5 71.0 2.48 MΩ 57.1 73.4 76.8 75.7 3.19 MΩ 56.3 76.3 76.3 78.3 3.97 MΩ 55.5 77.1 76.3 80.7 4.38 MΩ 53.4 75.6 73.2 80.7 5.00 MΩ 51.2 74.5 71.4 80.8 6.67 MΩ 47.0 71.9 67.4 78.6 10.0 MΩ 39.2 64.0 57.6 74.0

Graph 2
Power in the load resistor (in pW) as function of the load resistor and used diode.
The results of table 7 are shown in this graph. Conclusions:

Of the tested diodes, the 5082-2835 (one diode) gives the most output power at low signal levels.
This diode gives the maximum output power at a load resistor of about 5 MΩ.
The other diodes give less output power, and have their power peak at lower load resistance.

Improvements on detector unit 1

The maximum sensitivity of detector unit 1 does not occur when all impedances (of LC circuit, diode and load resistor) are equal.

The sensitivity and selectivity of the detector unit can be improved, by using a single 5082-2835 diode instead of a dual HSMS-282K.
And besides that increasing the load resistor from 1.6 MΩ to 5 MΩ.
In that case the output power at low signal levels increases with a factor 1.52   (80.8 pW / 53.3 pW  see table 7) or + 1.81 dB.
At high signal levels, the output power increases with a factor 1.09   ( 447 nW / 411 nW  see table 5) or + 0.36 dB
The Q factor (selectivity) at low signal levels increases with a factor 1.50   (150 mVac / 100 mVac  see table 6).
The Q factor at high signal levels increases with a factor 1.745 (3.49 Vac / 2.00 Vac  see table 4).

Back to the index