Selecting Best Antenna Type for WWVB

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
WWVB broadcasts on 60 kHz. I am not sure there is a "best." Just asking for opinions or experience. Interference is an issue locally, as my oscilloscope blocks the signal if it is within a few feet.

After searching I have found 3 or maybe 3-1/2 choices I need to decide between:

1) Shielded loop. Several designs using bicycle wheels, electrical conduit pipe, copper water pipe and so forth. This one uses shielded RG58 or RG59 cable:
http://www.febo.com/time-freq/wwvb/antenna/

2) "Conventional" air-core loop:
http://www.arrl.org/files/file/QEX_Next_Issue/2015/Nov-Dec_2015/Magliacane.pdf
See, page 16. Note: made from ribbon cable.

3) Ferrite-loop:
http://prc68.com/I/Loop.shtml
http://webpages.charter.net/ekyle/wwvbant.pdf

4) Ferrite-sleeve (counts as one-half):
Sorry about the you tube. The construction article I saved is lengthy.

I did not find any source that compared the various versions. The ferrite sleeve is quite expensive, if you use American-sourced ferrite (type 33 ferrite alone is >$100). However, the other three options are not particularly expensive. For either the shielded loop or simple air-core, I would probably opt for a 2 to 3 foot square.

My question: Any experience or opinions on the first three types listed?

Regards, John
 

DickCappels

Joined Aug 21, 2008
10,153
I have had limited experience in this area. Speaking of area, a loop antenna's sensitivity is proportional to area. Ferrite loop antennas get you more effective area than an air-core loop antenna of the same dimensions. The lowest cost in terms of signal-to-noise will be with either an air core loop, or if in a noisy environment, a shielded loop.
 

Papabravo

Joined Feb 24, 2006
21,158
The research on VLF antennas shows that they need to be physically very large to accommodate the long wavelength of the signal. The wavelength is 5000 meters. I'm not aware of a good way to evaluate the alternatives except by performing an empirical experiment. The results might also depend on where you live and what kind of soil you have.
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
@Papabravo

Agreed, there are undefined variables. The home is made of logs in the far Western corner of Lorain County (about 40 miles West of Cleveland). Area is flat, but surrounded by forest. Overall, it is basically flat farmland, but I am on a bank-- many 100 ft above the Black River. Lots of windows in the room in which this will be used. I do not want an exterior antenna. The small ferrite-core antenna that comes with various "atomic clocks" will catch a usable signal late at night (e.g., 2 AM), but during the day, reception is very poor and unusable.

This article impressed me as it makes clear that the equations can be reduced to area and mass: http://traktoria.org/files/radio/an...vlf_receivers_with_air-core_loop_antennas.pdf (See: Eq. 3.14). Does adding a ferrite core actually reduce the area dependence for sensitivity, or does it simply make construction of smaller, resonant antennas more convenient? (See: second link under ferrite-loop in my original post)?

Unfortunately, this is my first time building an antenna. All of my previous experience has been in using commercial antennas. So, I am pretty ignorant of the field.

I am leaning toward the RG58 shielded loop, but have not ruled out an air-core antenna made from ribbon cable, assuming I can find 40-conductor ribbon cable with easily soldered conductors. I do have a way to solder such wires that tests well with smaller ribbon cable with bare copper conductors (rainbow type), but when tested with typical computer gray cable, the wires do not solder well. My suspicion is that the samples of gray cable I have are plated aluminum, not copper.

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
I decided to go with the shielded RG58 design (link #1 above). It was only $0.14/ft and is supposed to be solid copper. I hope. At least the wooden frame will be usable for both designs, should I have to make the air-core version.

John
 

Papabravo

Joined Feb 24, 2006
21,158
Michigan, Ohio, Indiana and Wisconsin have adequate ground conductivity. Trying to receive VLF signals in the daytime will always be problematical. To get an indication of what you are up against try receiving distant AM stations on the lower end of the Broadcast Band 520 kHz to 600 kHz. If you can't hear those stations with the antennas you are considering then you have no chance with the WWVB signal. I can tell you the WWV also broadcasts on 5.0 MHz., 10.0 MHz., and 20.0 MHz. Those frequencies provide more than adequate signals all year round in the daytime with modest antennas.
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
It is a long story... And, I am aware of the other frequencies WWVB uses. I actually bought a Realistic DX-302 last year to listen to shortwave. I picked up a really neat South American station, but returned the radio for another reason. Anyway, I have designed a PVC rack to hang my 100 feet of RG58 onto and am scrounging up parts for the rack as we write. Everything else is "in the mail." Hopefully, I will have an update around the weekend, or by next Tuesday.

For me, this is really fun and the only deadline is the literal one. Fortunately (unfortunately?), it looks like we are having an early Spring, unlike the past 2 years. My play time becomes much more limited in Spring.

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE:

I built the antenna described in the first link, post #1. I used about 100 ft of RG58; frame is 36" on the diagnonals, 11 turns, gap at the top.
upload_2016-3-5_10-28-54.png

Everything here is just crudely put together for temporary testing. Here is the "tuning" circuit. The mica caps total about 23100 pF; the egg-slicer goes to 1000 pF. I very roughly tuned the antenna resonance yesterday to 60 kHz.
upload_2016-3-5_10-35-8.png

The red proto board is from SparkFun. It has a 16F1829 on it for decoding and other things, like reading and programming an RTC. The green pcb is a WWVB receiver from PV Electronics in the UK.

I captured WWVB at midnight local time with the ferrite loop that was supplied with the receiver. Then switched to the shielded loop. Signal was strong and small changes in capacitance ((say, +/- 1000 pF) made little difference. I had to retune it a little at about 0700 local to recapture the signal (local sunrise is about 0700), and it has been working continuously until 1100 local (1000 Fort Collins time). Unfortunately, the receiver does not provide a signal strength output. I will be monitoring it through the afternoon and early evening.

I would like to tune it better and characterize the antenna. Loop inductance measured 286 mH with an Wavetek LCR55 meter, which gives a calculated capacitance of about 24,600 pF.

What I want to determine is the antenna's Q factor and impedance. A very rough estimate of the former just using free air resonance and visual estimation from the o'scope as I scanned the frequency shows it to be about 15, maybe a little higher, but I don't trust that measurement, as the frequency generator was not stable. The problem I am facing is that most equipment and descriptions do not cover such long wavelengths and electrically small antennas.

Any easy suggestions?

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE#2:
I wanted to prove whether the more complicated shielded loop gave better reception than the ferrite loop, but I do not have a frequency analyzer or other specialized RF equipment. I ordered one of the cheap SDR dongles, and if I can get that to work, then I will probably get the upconverter from NOOELEC, which will allow me a poor-man frequency analyzer. That is in the distant future. Today, I made a little amplifier for the signal. Some of the coupling capacitors are a bit odd, but those values are what I had on hand. The basic design comes in part from an article on a VLF receiver (http://sidstation.loudet.org/hw-en.xhtml ) and a few other miscellaneous files I had.

The as built schematic is attached as a PDF below (NB: A fixed resistor of 82k "R10" was substituted for R9).

I then made a short loop of wire terminated with a 51 Ω resistor on the end of a length of RG58 coax. That was attached to my frequency generator and functioned as my transmitter. Directionality of both antennas was obvious. The shielded loop subjectively gave a stronger signal from further away. Here is the oscilloscope tracing of the source (yellow, channel1) and output from the antenna amplifier (magenta, channel3). Vpp for the source was 1.48V and for the antenna signal it was about 5V .

DS1Z_QuickPrint3.png

I did a little better estimate of the Q for the shielded loop. It was not a rigid set up, and slight movements made significant changes. I did get a value of around 40 when I was really careful not to bump anything. The difference between the two -3dB frequencies was not symmetrical about f0. I didn't expect them to be exactly equal, but they were quite unsymmetrical. With f0 = 60 kHz , the two -3 dB frequencies were 60.38 kHz and 58.98 kHz. I assume experimental error is the most likely cause, or perhaps I have a little too much capacitance in the resonant circuit.

Regards,

John
 

Attachments

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE#3: On a lark, I pointed the shielded loop antenna at WWVB this morning and was surprised by the result:

DS1Z_QuickPrint4.png

That is the WWVB signal with the little amplifier. Definitely, an unplanned for event.

John
 

alfacliff

Joined Dec 13, 2013
2,458
make sure to leave a gap in the shield for a shielded loop. if you dont, there will be no rf except for feedline leakage. also, it is harder to pickup vlf signals depending on distance in the daytime then at night, look up vlf propagation.
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
Yep, night reception is one reason I am up at 3 AM playing with this thing and was my main motivator for building a better antenna for daytime use. The receivers I have are from "atomic clocks" and provide no relative signal strength indication. I was quite surprised with the improved signal, but around 7:15 AM, I still had signal sufficient for the receiver, but couldn't watch it on the scope as nicely. The new antenna provides 24-hour reception, but can be totally knocked out by turning on the microwave that is 40 feet away.

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
Hi Bertus,
I was aware of that version, but my shop is in an unheated barn, so I went with one I could make in the house. It is working fine. Today I got my old. stiff hands to lace the RG58 coil nicely. My real project is to measure drift of a few RTC modules (inexpensive from China) versus WWVB. I wanted to shoot for a few milliseconds precision using the "zero minute" transition that WWVB sends. I am using a PIC 16F1829. The firmware has been written and tested in parts with a simulated WWVB signal. I needed a reliable signal to get to the next step.

Thanks for thinking of this project.

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE #4
It is done and has been working for a few days. I laced the cables using flat lacing and put the tuning caps in the box at the base. Reception during the day is not a problem, but local thunderstorms are.

upload_2016-3-15_11-48-50.png

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
UPDATE 04/01/16

This is no joke. My SDR receiver arrived yesterday, and I have been busy listening to LF and even some VLF. At 60 kHz, the shield loop antenna gave a peak voltage of about -82dB. That was at 0500 DST (Eastern USA). Signal from the OEM ferrite loop antennas was buried in the noise at about -120 dB.

As for the actual project of measuring drift of a DS3231 RTC (Source: Banggood), its 1 Hz generated signal is about 0.1 to 0.15 ppm fast (about 1 second per 100 days). Next step is to try to trim that down. (See: http://forum.allaboutcircuits.com/threads/learning-rigol-ds1054z-quick5.122468/ ) for an actual oscilloscope trace.

John
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
I will definitely keep an eye out for it. The price, even including shipping to the US is certainly reasonable for a decent technical book.

The shielded loop I made is working well not only for WWVB, but for other VLF and thunderstorms. Yesterday afternoon, I was working on some code and all of a sudden, my scope went from 1 baud to a forest. Thirty minutes later, I heard the first thunder.

John
 

tracecom

Joined Apr 16, 2010
3,944
I am enjoying your thread. I have been a ham for more than 40 years and an SWL even longer, but inactive for much of that time. Recently, my interest has been rekindled. I am also interesting in RTC chips and circuits. Is your objective to create a very low drift RTC or what?
 

Thread Starter

jpanhalt

Joined Jan 18, 2008
11,087
@tracecom
My original objectives were twofold: 1) Something to do during the Winter; and 2) Improve my programming skill. This project got started with the availability of really cheap DDS and RTC modules -- DS3231 based RTC modules were about $1.50 at Banggood. That extended my experience with SPI and introduced me to I2C. Right out of the package, the DS3231 keeps pretty good time in human terms. That is, the seconds stay pretty good over the course of several days. One thing led to another and today, using WWVB as my reference, I can watch the drift in ms per day. I may eventually try to trim out some of the drift, but that is not a priority.

At my location, good reception of WWVB is pretty much restricted to night and early morning. By good reception, I mean good enough for my current software to read the sequence reliably. Although the signal can be easily read by a human during the day, small blips and glitches upset the timing. Revising that part of the code to make it more robust is what I am actually working on now. The earlier code used fairly crude error detection. I am trying to improve that in the current version. If that doesn't work, then I will just accept that I can't synchronize with WWVB 24/7. ;)

Regards, John
 
Top