OBSERVATIONS FROM SCOTLAND                            28 May 2008                                GM1SXX

I've been doing the numbers for a new triband loop.  It soon dawned on me that with very little work, I could make the loop function on five bands.  That's right, five bands from an aerial that's only 2M square.

Here are the results of the Scottish Jury.

The above figures assume a 1.9M square loop made of 15mm copper water pipe with soldered 'yorkshire' elbows at 7M above ground level (in my attic) with a good quality high voltage variable capacitor, preferably a suitably rated vacuum variable.  I scaled the figures for 100W of drive (voltages and currents).

You may think the 0.86% efficiency on 160M is lousy. This is not so. Unless you have a huge amount of real-estate to put up a 132foot quarter-wave), or 264 feet for a half-wave, this represents a good compromise.  Furthermore, it requires no ground system whatsoever.  It's notoriously difficult to get any sort of decent efficiency on 160M with all types of small aerials.  The laws of physics intervene.

The efficiency on 80M easily exceeds that of most verticals, especially the trapped sort, and that's assuming that such a vertical has an excellent ground system.

On 40M the efficiency is a commendable 51% while on 30 & 20M, it is very good indeed.  All this from a square of metal tubing less than 2M by 2M in size.  

The secret, should there be one, of efficient magnetic loop aerials is in having the lowest possible ohmic  losses  while using the best tuning capacitor you can afford.  Ordinary tuning capacitors are unsuitable for even QRP power levels. 5W = 1.8KV.  Apart from the real risk of flashovers, the wiping contacts are just too lossy.  Take a quick look at the circulating current on the various bands in the table above to see why.  You really don't want to have any wiping connections whatsoever.

As a minimum, you might use a wide spaced butterfly type. These have no sliding connections. If you cannot find a suitable butterfly cap, you may want to try making one. Several excellent designs exist.  Google is your friend.

I decided on a different route.... EBay.  I found a seller in the Ukraine auctioning 10KV@50A rated vacuum variables with a capacitance range of 7.5 - 350pf, positively ideal for my newest aerial.  One is now winging  it's way here from Russia. The cost was very reasonable, under £50 including postage.

If you check out the chart above, you will find that it can cover 20-80m with only the vacuum variable in circuit.  For 160M I plan eventually to fit a heavy duty  RF relay to switch a 1000pf '10KV 'doorknob' in and out of circuit.  The capacitor will carry the best part of 40 AMPS of RF current on 160M and of course, the relay contacts need to be rated to handle that. You should always remember that mag-loops carry serious currents and voltages and your construction should be as good as possible.

Don't be tempted to use heavy-gauge braid to connect the loop to to the capacitor. Although braid can carry a lot of DC current, it's construction using many small wires of small surface area makes it a poor conductor at RF.  Think 'skin effect'.

Use a copper connecting strap (as wide and short as is feasible) to connect the loop to the capacitor.  Use a blowtorch, and not a soldering iron to solder the strap to the loop. The joints need to be as perfect as you can manage. The free ends of the copper straps can be wrapped around the capacitor connections and a Jubilee clip used to put light pressure on the copper strap for a good connection.  Don't overtighten the clip, you could easily crack and destroy the capacitor.  It only needs to be tight enough to avoid slippage.  You may be able to scrounge  off-cuts of thin sheet copper from a plumber or an engineering shop.

I looked at using other sizes of water pipe for this aerial but eventually went back to 15mm. While the losses are greater than with the larger 22MM pipe, the difference in efficiency is about 0.2 S'Units equivalent.  Thicker pipe does lead to lower circulating current because of the greater surface area to handle the skin current but you rapidly reach a state of diminishing returns.  

If your thing is portable operating on the lower bands (160 & 80) it would be an interesting experiment to make a 'fat loop', one using tube of  4 or more inches diameter. Do please remember that the 'tube' can be fabricated from copper sheet over a non-insulating former and so it's possible and feasible to cover a large plastic ring with copper sheet so long as the joints are adequately bonded together by soldering.   Increasing the surface area reduces the losses. Fatter is better! 

If the aerial is to be used outdoors, the feed and tuning capacitor must be in a non-conductive waterproof enclosure. 

The drive motor to turn your capacitor should have an interference suppression capacitor and be fed via a couple of ferrite sleeves.  Dress the control cables straight down and take them off the loop along with the coaxial feedline as close to the centreline of the loop as possible.  You can use a fibreglass rod or some other good quality insulator and ty-wrap the wires to this.  I place two sleeves at the motor and two more on the control wires just below the feedpoint.  The coax feed line to the loop also has two sleeves fitted just below the feedpoint

My drive motor is a 12V geared type made by MAXON. It has a 148:1 gearbox and will drive the tuning capacitor through a piece of plastic sleeving to provide electrical insulation from the capacitor.  This is important and it should be good enough to withstand 10KV.  The motor/gearbox is mounted on an IKEA chopping-board which is securely ty-wrapped to the loop by several ties on each side. 

For the feed, I'll use a pair of FT200 cores at the base of the loop (directly opposite the tuning capacitor) with two turns of heavy insulated stranded wire wrapped through them to drive the loop.  The co-ax is just connected directly to the transformer winding and a couple of ferrite sleeves slipped over the connection as close as you can to the feedpoint. A single Ty-Wrap stops them from slipping down the co-ax..

73 Al.

GM1SXX

Back to OBSERVATIONS