Schumann Resonance Receiver
The current design incorporates a number of changes that (can) lower the frequency response down to the ULF-ELF range. The design intentions are as follows:
- build an eField receiver
- building large coils requires tools I don't have
- I am more familiar with eField techniques than hField
- ball antenna provides high capacitance for size
- 10" metal sphere ideally provides about 16pF capacitance
- stainless steel gazing balls are relatively inexpensive
- avoid distortion and maintain low noise to detect millivot signal
- maximize dynamic range in the head amp (buffer) to handle large levels of hum (on the order of volts)
- adapt high-end audio preamp design techniques
- high resolution ADC will be required to get good detail on the signals of interest
- remote power for the head amp removes the need for batteries in the buffer itself
- build to be relatively portable
The current buffer
As of 8/2004 a new buffer has been employed with excellent results. This buffer tries to hit the design objectives outlined above: large dynamic range and low distortion with low noise. The buffer is housed in the housing used for the first buffer.
The yellow arrow points to the Teflon (PTFE) standoff used on the ultra-high impedance input to the buffer. A PDF of the schematic for this buffer is available here.
The first buffer
The original inspiration for this receiver was a high impedance buffer designed by Scott Fusare and published in "The Lowdown" (see the article here).
The original buffer is shown below. It is built directly on a piece of PC board, with pads punched from scrap PC board and superglued down to support components. The connecting cable is a 100' roll of telephone wire from Radio Shack. Note the strain relief at the lower left of the photo—a large screw with a cable tie securing the cable. The buffer is remotely powered from the receiver.
A schematic of the original buffer is available here. Please note that in my schematic, R3, the 100Meg resistor is a real resistor.
Other options
If you are building a more conventional FET front end, you can create a very high value resistor with a pair of NPN transistors. Scott Fusare introduced me to this technique, and he says it was suggested to him by Ken Walsh. This is an economical approach to creating an extremely high value resistor. You can see Scott's article on this approach here.
I did
experiment with this approach. Initially, I used a 2N3904
pair that Scott sent me to test. A pair I made made from
2N4401 transistors is shown to the right. In my tests, the
2N4401 tests at about 100 gigaohms – and initial field
tests looked promising. However, Scott warns that too high a
value can result in severe problems with fair weather
currents causing significant offset voltages on the FET gate.
He says that the brand of transistor seems to make a big
difference. For example, he has measured Motorola 2N3904s at
over a teraohm.
Note that this kind of resistor is definitely NOT a replacement for a real, good quality resistor (e.g. Victoreen). So, don't try to run these "resistors" at high voltages. But they should work adequately for home-brew ion chamber amplifiers and the like. Scott Fusare details to a friend how to measure the the bias current of your front-end – effectively in circuit – here . I really appreciate Scott sharing this material! Finally, this approach is recommended only if you don't have a 10 to 100 gigaohm resistor handy. If you do, use it!
The antenna/buffer assembly
The
main support is 1 1/2" PVC pipe. The ball with the pipe is 6'
long. Deployed, the unit is a little shorter, standing over
5' tall.
The buffer is housed in the small plastic box.
The 10" stainless steel gazing ball is mounted on a PVC adapter with silicone rubber caulk. A teflon jacketed wire through the PVC pipe connects the ball to the buffer. The wire to the ballis attachedby a screw that goes into a hole tapped in the ball. The screw connection is treated with GB Ox-Gard in hopes of minimizing future oxidation. The legs are hoe handles drilled on one end to accept the screws. Details of the arrangement are shown below.
The broomsticks are drilled and mounted via screws with wing nuts to an aluminum bracket made from 1" x 1/8" flat bar stock. The bracket is formed around the PVC tube and drilled to accept the broomstick screws and a mounting screw. Once in place, the bracket is further secured using JB Weld.
Experience has shown that the single screw attachment to the down pipe is not mechanically sound. If I did this again, I would use bar stock wide enough to allow at least two securing screws. Also, I have had trouble getting the epoxy to hold.
A cut away diagram detailing the ball mount and connection is shown below.
Performance
Currently, the signal is run through the DC-coupled into a an opamp that provides a little gain and offset control for the low pass filter to follow. The input of the opamp is AC coupled to provides some high pass filtering, which is desireable to help manage near DC transients from any movement near the ball (cats, crows, people, trees, etc.) as well as DC srift.
Right now, I am running the signal through a Wavetek 852 dual filter, set to cut off everything above about 55 Hz (at 96 db per octave!) and running at 0db gain. I monitor the signal out on my oscilloscope and feed it into Spectrum Lab on my PC.
The PC card in my machine starts to roll-off below about 15Hz. I compensate for this to some extent with postprocessing. However, the sound card input coupling capacitors can be increased in value to lower the break point.
Below is an example of a recording made in front of my house at a quiet moment with my previous preamp. It was completed at just after 10PM CST on 11/24/01.
Obvious is an almost always present signal at about 25 Hz. You can also see a signal at about 36 Hz that disappears right at 1000 seconds into the recording. Other examples are available for download from this page .
System Design - Planned
The next phase is to cancel a good proportion of the hum right at the input. (details to follow)
A number of references are available on the links page here.