Prototyping a LF Ferrite Rod Antenna

[jFarley]

Using the Standard Loop amp (with Tuning Diodes) to hear NAVTEX transmissions did not appear to be a viable option, so I have been doing some prototyping to see what would work.  It's antenna building time, and away we go!

I have come to use the following methodology to design resonant loops:

1) Define the frequency coverage and Q required across the band.
2) Define the Tuning Means (diodes?  air-variable cap?) to be used and get a firm handle on the capacitance range available
3) Define the winding (air coil?  ferrite rod?) and the required inductance.

I decided that I wanted a ferrite rod loop which would cover the LF range from ~150 kHz to ~530 kHz.  I decided on ultimately using a dual section 365 pF variable cap; nice units are available as new construction and cheap.  This would require a winding inductance of around 1500 uH.  I had four R33-050-750 rods in the junk box (well, 3 and 1/2 and 1/2, since I dropped one), and one of these would be great for the project.

Capacitor Q would most likely not be an issue down at these frequencies; strong evidence for this comes from the crystal radio guys, in particular, from a guy named Ben Tongue, who has published a number of detailed studies on the components typically used in a crystal radio build.  Incidentally, Ben Tongue was a founding partner of Blonder-Tongue (remember them?)  With regards to cap Q, this is a must read:

http://www.bentongue.com/xtalset/24Cmnts/24Cmnts.html

Q of the inductor is very much an issue, and again I turned to Mr. Tongue for some insight:

http://www.bentongue.com/xtalset/29MxQFL/29MxQFL.html

If anything is gleaned from this, Mr. Tongue dispels the prevailing myth that "for best Q, the winding should cover the entire length of the rod".  I have verified that at least in my winding best Q is attained at 40-50% rod coverage.

I built several prototype coils, and am currently using the following for this loop:

The ferrite bar is fitted with a length of 1/2" PEX pipe as a coil form which has been slitted down its length to fit over the rod.  On this was wound a balanced winding of 105 + 105 turns of 7/44 Litz (that's all I had other than 660/46 which is a tad too big).  The final winding will be either 40/44 or 60/44, both on order.  The net inductance is around 2900 uH because at this time I have only a 250 pF air variable under motor control out in the Antenna Farm.

Here is an image of the prototype on the bench:



To be noted: many published designs use the full extent of the winding to feed the following amp.  In my RF environment, this was unmitigated disaster!  The strong MW locals (local WMVP 1000 in particular) were causing all sorts of problems for the amp.  Instead, I took the amp feeds from +/- 5 turn taps on the winding, and this has cleared up the problem nicely.  It's probably too conservative, and this will be revisited when the real Litz comes in and new coils are wound.

The amp is just a balanced U310 push-pull amplifier running drain currents of around 10 mA.

I got the entire schmear mounted outside yesterday and patched in to the motor driven cap, and have begun testing.  Results shortly!

[jFarley]

I powered up the loop last night at 518 for an overnight test while capturing with Yand (the NAVTEX decoder I use).  All of the Canadian and US Area 4 stations seemed to be there in the morning report with good to great copy.  No GRL or Carib stations were reported.  Everything went swimmingly until...

When I came down in the AM, I noted that the loop had detuned gradually and was peaked at around 508 kHz.  Temps dropped here to about 11F overnight, and the tempco of the rod had bitten me in the arse!  I will either have to drop the Q or temperature compensate the winding if I insist on doing overnites with this.

But it's a good start.  Reception quality while I was listening live was equal to or better than the Wellbrook, which is not too bad down there.  As I type this I am comparing reception of the ferrite rod against the Wellbrook during daytime NDBs.  Again, the ferrite is equal to or better than the Wellbrook.  Without a low-pass filter on the SDR, the SDR is full of spurious junk due to strong MW locals when using the Wellbrook.  So far, no junk has been noted with the ferrite rod.

BTW, the rodd was rotated for max reception from S GRL.
Q at 518 kHz was measured at ~350.

[ChrisSmolinski]

Very interesting results, jFarley!

Is the antenna directional at these frequencies, and if so, did that affect what NAVTEX stations were heard?

[jFarley]

It should be.  It was very directional in the house when rotated against noise.  At this moment, it is not mounted on the mast with the rotator.  Hopefully I will do that when I get the final winding built and get the rod in a waterproof housing; right now it's housed in a Hefty trash liner.  I see multiple freqs here with more than one NDB, and I am excited about the possibility of splitting them off each other's back.

I ran across this video the other day.  It is extremely interesting as it shows the now discontinued Palstar LA-30 in action:

http://www.youtube.com/watch?v=3Pji41kzYF8

I note a couple of things:

1) The directionality seems very good at LF, as does its sensitivity.
2) The Q is good but not great; you can see this in the approach to tuning peak, and the apparent lack of audio sideband cutting.
3) WTF good is a knob on something if it works backwards!  I realize that there is a planetary drive on the tuning cap, but really!!!!
4) Off peak rotation it hears house noise; that is hardly surprising.  Getting that antenna outdoors and away from the house is a must.

[ChrisSmolinski]

I had a directional loop antenna for LW and MW use many years ago. As I recall it worked OK inside, but this was in the early 80s, when we had far fewer offending devices in the house.

No wonder we see so many of the older loops on eBay today.

[ka1iic]

Hey Mr. jFarley. on the rod you dropped...  use some crazy glue and put it back together (if it didn't shatter that is).  I've done it dozens of times and it works for me ;-)

"I love it when a place comes together"

[jFarley]

That would work mechanically, true, but it tends to create an air gap, however small.  If there is a chip-out, forget it.  Any winding over or near that gap will see higher losses, and if you are shooting for best Q cheapness don't pay.  Besides, I found another couple 7.5" rods while cleaning the lab area!

Rather, I am using the shorter sections for another project.

[ka1iic]

You are correct of course :-) but me being cheap... and not having major lab stock anymore <sigh> well you know where that goes for sure :-)  Just a comment...

Great on the longer sticks... use um ehhh?  Good luck with the project... :-)

I'm thinking of making a box loop for the BCB... I've made a dozen or so over the years and they work for me... imagine making one for VLF... never tried that... hmmm...  I would but the noise level here at those frequencies is way out of line... arrrggghhh  later dude :-)

[jFarley]

Things are moving along nicely.  There are at least 5 technologies involved in a remotely tuned ferrite loop.

1) The rod, its windings, and the capacitor used to tune to frequency.
2) The amplifier.

I spent most of the winter evaluating different rod and winding prototypes, as well as the amplifier.   Those resulted in the logs which for the most part were posted here:  http://www.hfunderground.com/board/index.php/topic,19868.0.html

This was a pretty ugly but effective setup, held together with alligator clips and electrical tape as seen here.



These results inspired the confidence to build a far more robust working model which will be used to evaluate:

3) Means for precisely tuning the mechanical capacitor from about 150 feet away.
4) Interfaces to the loop
5) Loop electrostatic shielding

Resulting logs to be posted here:  http://www.hfunderground.com/board/index.php/topic,20896.0.html

This incorporates a far more advanced method for tuning the loop, and the guts of the tuning train can be seen here:



So far it's been a lot of fun with very encouraging results!

[ChrisSmolinski]

Very impressive, both the loop design and the logs. I look forward to your next update!

[Rafman]

Very impressive... I was just reading some of your designs in reprints of Fine Tuning "Proceedings"...

I also want to thank you for finding Ben Tongue! I worked under his brilliance at Blonder-Tongue in R & D in 1983. He taught me many things I retain till this day! However, I never thought I'd see a website with his brilliance on display.

Keep up the great work!!!

Rafman

[jFarley]

Thank you for the kind words, Rafman; I really do appreciate them.

The Crystal Radio guys are a very interesting bunch.  It is very old school technology, true, but to build a good crystal set you really have to do a number of small things just right.  Mr. Tongue seems to have a detail oriented mind, and reading his pubs has saved me a ton of time and experimental builds.  That you worked for him at BT is an incredible addition to the It Really Is a Small Freaking World File Folder.

On the back burner: some day I'd like to put together a crystal set and see if I can copy any of the AM pirates with it.  That would be cool.

[ChrisSmolinski]

FYI - the www.bentongue.com site seems to be down as I type this. However the pages are available on www.archive.org

[jFarley]

Ben Tongue (of Blonder Tongue fame) passed recently.  I am glad that his work is archived.  I have copies of the stuff I need archived here.  Some of his work is not matched on other sites.