The “Mystery” Crystal Set

After becoming involved in an interesting discussion of the Mystery Crystal Set coil, and how to go about measuring its inductance, I decided to wind one and take some measurements. The details of the analysis are discussed on the page: Calculation Methods Part 1d. I've summarized some of the information from that page here.

I've also created an Online Calculator for determining the coil parameters.

After analysing the coil, I had no excuse not to build it into a complete set. So the final result is pictured below.

The Mystery Crystal Set, was first described in the July 3rd, 1932 Brisbane [Australia] Sunday Mail, and is unique in the way that its main coil is wound. The original Mystery Set circuit diagram is reproduced below:

One of the more interesting characteristics of the Mystery Crystal Set is that there is no apparent signal path between the antenna and ground. The antenna is connected to the primary, while the circuit's only ground point is connected to the secondary. This set makes use of the inter-winding capacitance to complete the signal path.

The coil has the primary and secondary wound side by side (i.e., bifilar). In the original design, a 3 inch (75 mm) diameter coil form was used. The primary was close-wound 12 turns. Then, after the 12th primary turn, the winding continued with the primary and secondary close-wound side by side for the next 25 turns. The secondary was then terminated, and the primary then continued on for another close-wound 13 turns. Thus, the primary has a total of 50 turns, and the secondary has 25 turns.

By means of a two position switch, the antenna could be connected to either end of the primary winding, giving either a Selective or Broad tuning characteristic. The Broad setting also gives more volume. I've reproduced the schematic as it originally appeared. However, it should be noted, that if the coil primary and secondary leads are located the way the diagram implies, then the Broad-Selective switch positions are actually backwards. I simply wired mine to be permanently in the (true) selective configuration.

It can also be reconnected so that the ground is connected to the stator of the second section of the variable capacitor (the antenna connection is to the rotor/frame). In this configuration, it is similar to an MRL#39 crystal set. This gives a further improvement to selectivity at the expense of sensitivity. So far, I haven't found this configuration to be any advantage on the medium wave band, and I expect that is because of the relatively low Q of my coil. However, in this configuration, it picks up signals in the 49 meter shortwave band, which are of some interest.

Note, that in the original coil, the secondary is not exactly centred. In order for the secondary to be centred, the total number of primary and secondary turns would have to be an odd number. When I wound my coil, using an empty vitamin bottle for the form, I made the primary 51 turns so that the secondary could be centred exactly.

Coil specifications are as follows:

1. Number of primary turns: 51;

2. Number of secondary turns: 25;

3. Coil diameter: 76.76 mm;

4. Total coil length: 42.85 mm;

5. Coil form material: HDPE, εR =2.3;

6. Wire type: #24 AWG magnet wire, polyamide insulation: εR=3.5;

7. Estimated effective εR: 2;

8. Conductor diameter: 0.5106 mm;

9. Wire outside diameter including insulation: 0.5588 mm;

10. Winding pitch: 0.5638 mm.

11. Estimated inter-winding capacitance: 1500 pF

12. Primary Inductance LP: 187.1718 µH

13. Secondary Inductance LS: 58.1545 µH

Aside from the modern materials, this is very close to the specifications for the coil in the original Mystery Set article.

For the detector, I tried several different germanium, and Schottky diodes. The best performance was with a type 1N191 germanium diode. It was even better with several 1N191's in parallel. In the end I paralleled three 1N191 diodes. I did not bother with the 0.001 µF capacitor after the detector. It serves no useful purpose, other than to increase component count. The cathode of the detector is wired to an output jack on the front panel. I use an audio matching transformer between the detector output and the headphone. The audio matching unit and headphones are shown below:

The matching unit consists of a Bogen T725 audio transformer, and two 6-position rotary switches connected to taps on the transformer. One switch selects the input tap, and the second switch selects the output tap. Also included is a 'Benny' which is used to adjust the detector's DC load to match the AC load. The control on the far left is a 100 kΩ potentiometer for adjusting the Benny's DC load. The headphones are Western Electric Model D173014 sound powered phones. The elements have an impedance of 600 ohms each, and are wired in series, for a total impedance of 1200 ohms.

As a side note, rather than the traditional 1/4" phone jacks, I've used RCA type phono connectors, as they are less expensive, and seem to make more reliable electrical contact.

Here is a rear view of the receiver.

A 6:1 reduction drive is used on the variable capacitor. Even with the low selectivity, the reduction drive still makes tuning much easier. The antenna and ground connection screws are shown at the back on the base board. The underside of the base board has been hollowed out using a router, and the wiring and detector diodes are located there. The vertical wood dowel (with the knob on top) is used as a support for the leads coming off of the coil. This is shown in more detail in the following front view:

Now, I must confess that these are rather bad construction techniques for a crystal set, if the intention is (and it should be) to get the maximum possible Q.

I used a Heathkit QM-1 Q-Meter to measure the Q of the primary (tank) side of the coil. The results are shown in the following graph:

With a maximum Q of just over 180, it's certainly not a big league set, and not surprisingly, the selectivity is not particularly good. However, the sensitivity is better than I expected, and I have been able to pick up several stations more than 1000 km distant.

I entered this set in the 2012 Radioboard DX Contest, and following is my reception log:

Log  Date        Time   Freq  Call   Dist  Power  Dist/  Location

No.              (CST)  kHz   Sign   km    Watts  Power  City, Province/State

---  ----------  -----  ----  ----   ----  -----  -----  --------------------

001  2012-07-24  23:55   860  CBKF-2   17  10000    300  SASKATOON, SK

002  2012-07-25  00:01   650  CKOM      9  10000    300  SASKATOON, SK

003  2012-07-25  00:03   600  CJWW     12   8000    300  SASKATOON, SK

004  2012-07-25  00:05   540  CBK      94  50000    300  WATROUS, SK

005  2012-07-25  00:12   900  CKBI    126  10000    473  PRINCE ALBERT, SK

006  2012-07-25  00:14   990  CBW     687  46000   2210  WINNIPEG, MB

007  2012-07-25  00:16  1160  KSL    1326  50000   4233  SALT LAKE CITY, UT, US

008  2012-07-25  00:35  1340  CIBQ    406   1000   2030  BROOKS, AB

009  2012-07-25  00:40  1050  CJNB    140  10000    525  NORTH BATTLEFORD, SK

010  2012-07-25  00:43  1260  CFRN    497  50000   1587  EDMONTON, AB

011  2012-07-25  00:47  1140  CHRB    515  46000   1657  HIGH RIVER, AB

012  2012-07-25  03:48  1280  CJSL    428  10000   1605  ESTEVAN, SK

013  2012-07-25  03:53  1130  CKWX   1201  50000   3834  VANCOUVER, BC

014  2012-07-25  03:57  1310  KNOX    834   5000   3382  GRAND FORKS, ND, US

015  2012-07-25  04:00  1320  CHMB   1199  50000   3827  VANCOUVER, BC

016  2012-07-25  04:06  1330  CJYM    117  10000    439  ROSETOWN, SK

017  2012-07-25  11:08  1090  KBOZ    791   5000   3208  BOZEMAN, MT, US

018  2012-07-25  11:27  1500  KSTP   1269  50000   4051  ST. PAUL, MN, US

019  2012-07-25  11:32  1540  KXEL   1542  50000   4922  WATERLOO, IA, US

020  2012-07-25  11:47  1460  KLTC    646   5000   2620  DICKINSON, ND, US

021  2012-07-25  11:52   910  CKDQ    475  50000   1516  DRUMHELLER, AB

022  2012-07-26  00:01  1570  CKMW    689  10000   2584  MORDEN, MB

023  2012-07-26  00:31   940  CJGX    313  10000   1174  YORKTON, SK

024  2012-07-26  00:35   950  CFAM    702  10000   2633  ALTONA, MB

025  2012-07-26  00:37   960  CFAC    513  50000   1638  CALGARY, AB

026  2012-07-26  00:40   980  CJME    240   5000    973  REGINA, SK

027  2012-07-26  00:48  1110  KFAB   1464  50000   4673  OMAHA, NE, US

028  2012-07-26  01:03  1030  KTWO   1030  50000   3288  CASPER, WY, US

All station locations are in Canada or the United States (US locations are indicated as such).

The Dist/Power number is the Distance divided by the Logarithm of the Transmitter Power (minimum value is limited to 300), and is often used as a guideline to determine the relative difficulty of receiving the station (the higher the number, the more difficult), though this is a very rough guide, because many other factors also affect reception.

The following map shows the locations of the stations received.

It is important to note that I have three powerful local stations at 600 kHz, 650 kHz and 860 kHz. Because of this, I was not able to pick up any distant stations in this frequency range, even when I used a series of wavetraps to attenuate the locals. This is where a more selective set would prove its worth. Still, I can't complain about the results. Though I must say that the good performance is largely due a decent Antenna/Ground system, very sensitive headphones, and good geography (my location is great for DX).

As mentioned above, it's possible to configure the antenna/ground connections as an MRL#39 set, which unintentionally made it sensitive to shortwave stations. I logged the following shortwave stations in this configuration:

Log  Date        Time   Freq  Call   Dist  Country, Comments

No.              (CST)   kHz  Sign   km

---  ----------  -----  ----  -----  ----  ------------------------

001  2012-07-24  23:45  6060  ....   3850  Cuba, Radio Havana

002  2012-07-25  03:14  5935  WWCR   2360  USA, Nashville, TN; Religious

003  2012-07-25  21:05  7385  WHRI   2970  USA, Cypress Creek, SC; Religious

004  2012-07-28  02:47  5000  WWV    1285  USA, Ft. Collins, CO; Time Signal

The Mystery Crystal Set, seems to be a good beginner's set, considering that it does not require any exotic components, and is quite easy to build.

Back to: