Previous Failure (dotted)
Giant Coat Hanger (Version 2)
Updated: 2026-08-25
My main antenna goes by a nickname: The Giant Coat Hanger. So-called for being a delta shaped SLOUS (sky loop of unusual size). It is strung between three very large trees spaced far apart. One is a sturdy white pine, the other two very ancient (and somewhat decrepit) black cherry trees. Each side of the triangle spans a gap of approximately 44m at an average 12m elevation.
It’s evolution spans several stages across many years. Should that be of interest, I spell out each of those stages below. Info on how I erected it is detailed elsewhere: LINK
For the Giant Coat Hanger’s current arrangement to make any sense, I first must admit of the dismal failure immediately preceding it. Certain aspects of this failed prior antenna remain significant despite its later deconstruction. Chief among these would be its feed-point, which I chose to reemploy for reasons that need explaining.
Said feed-point (subsequently recycled) was an MFJ remote auto-tuner mounted on brackets partway up the trunk of a particular tree. The tree in question is an ancient and widely spreading black cherry. Like many of its species, this tree rises up from the ground as a cluster of conjoined siblings. For a mounting location, I chose the far side of that bole furthest away from the house. This for its presenting a clear path by slingshot up to an overhanging, very high fork. For ease of maintenance, I mounted the auto-tuner at a height of 2m above ground, easily reachable by stepladder.
This very large tree comprised of several boles did not stand alone, being the central member of others like it all in a row, and so closely spaced that their roots could hardly avoid crisscrossing each other underground. Thus was I prevented from digging a trench in which to lay conduit for running coax quite so far as would have been ideal. I therefor dug in a straight line from just below the nearest basement window only so far as the nearer side of that tree. And so to this day does my short run of LMR-400 emerge from out of buried conduit to loop around a bole of that tree up the location of that MFJ remote auto-tuner on the opposite side. The expense and labor of that installation was nothing I wished to sacrifice without good and sufficient reason. The antenna itself, however, proved a different matter entirely.
It seemed like a good idea at the time. My excuses are these... I was then just returning to the hobby after a wholly unintentional 26-year QRT (the usual story: marriage, a kid, work, and loss of prior antenna due to re-roofing). I was just then concluding a much fraught 20-year marriage. Having surrendered a mortgage-free house in Kalamazoo MI for an apartment in Holland MI also brought to conclusion a most fatiguing, hour-long commute each way to the job I had accepted five years prior. Shortly before the lease was set to renew, my XYL-to-be (now XYL in fact, but then still a long-distance girlfriend resident in British Columbia) called my attention to a property listing in nearby Port Sheldon Township. It meeting both our requirements (my own being no HOA, a 90-foot tower restriction, trees and space for an antenna) I made the purchase.
Which brings me now to the particulars of that failed first antenna layout. Lacking funds for a tower quite yet (or since, as it turns out) I figured to put up as high as I possibly could, as large of a sky loop as I possibly could. I picked out four convenient trees. The feed-point tree being that already described. The other three, so I hoped, would allow for suspending my wire from one or another high fork so as to form a somewhat lopsided diamond completely surrounding the house.
Here was when I learned my lesson about the need to special-order perfectly spherical fishing sinkers for use as slingshot weights. This shape in favor of locally available teardrop shaped ones. I had on hand a quantity of teardrops in two sizes: 1oz & 2oz (28g & 57g). Right away, I used up all of those with aggravating misfires. Nor did either of the two local sports outlets have any spherical ones. Having put off construction until after leaves had dropped mid-November, and with the cost of overnight shipping on heaving lead sinkers being too dear, I had no choice but to return to local stores and keep on enduring misfires until I might have success.
Thus did I end up festooning that first tree plus also the second with snagged hanging weights, each brightly marked by its tuft of fluorescent pink fly-tying fluff. Rigging each of those first two trees consumed a whole day by itself, and thus the entirety of the single weekend I had budgeted. This with the weatherman predicting snow for the following weekend. A huge complication as those first two trees had offered much better high-limb targets than either of the remaining two I had chosen. The third was yet another black cherry tree, all of its higher forks discouraging by reason of tight placement. The second was an enormous white pine which I now judged to be nearly impossible.
And so, very reluctantly, I compromised on height. Bringing out a 30-foot (9.14m) extension ladder (purchased earlier for mounting bat houses), I screwed eye bolts as high up as I felt comfortable reaching. From these did I rig the two remaining antenna corners. This did loop end end up height at two corners, not-so-high at two others. Like so with its midpoint average directly over the house.
A single month only did that arrangement serve. The problem was that deadly menace RF inside the shack. At original design height, it might not have been a problem. But I simply could not be sure. Whatever solution I might come up with had to wait until spring. Yet when spring did come, I had other distractions. A pressing deadline at work had me committed to putting in unpaid overtime every evening and also on weekends. As a test engineer on salary, that can’t be avoided sometimes. Usually not a big deal as I can ask for comp-time later. Plus also in this case, I got rewarded with a really nice bonus. A real drag however, in that meanwhile the trees had leafed out.
Thereby did my winter-long QRT suffer an extension until the following November. By which time no longer was I willing to suffer the risk of stringing wire directly over the house. I hauled down wire and rigging both from all but the feed-point tree. After much consideration, I came up with that triangular path still presently in use.
The feed-point might be better arranged as a 1:1 balun wound on a largish Mix 37 ferite. I think about upgrading to this on occasion. One day I will get to it.
My own feedwires connect directly to the far end of a short, mostly underground, run of LMR-400 coax. The reasons are as below...
The MFJ remote auto-tuner proved problematic almost at once. This on account of its memory feature, which could not be remotely cleared. As is common with most all auto-tuners, its design too was an L-Match. The tuning parameters of which are seemingly less broad than for a T-Match. Thus did it happen that the memory feature worked against me. Conditions would change (raindrops clinging to 138m of wire, trees leafing out in sprint, those leaves being dry or wet) but the cursed auto-tuner, for having memorized a frequency, would not readjust. Which then required clearing the memory.
But that could not be done from inside the shack. Instead it was required to open the remote box itself. Whose cover was held on with six non-captive screws and nuts. Then it was needful to depress three internal buttons in sequence. These located on a standoff daughter board mounted upside down! Which operation required using a dental mirror to even find.
Very tiresome, to say the least. For a shorter antenna whose wire did not run very close by and through trees, likely that MFJ model works very well. But it just wouldn’t for me. And so, I began leaving it off, tuning instead from inside the shack using an MFJ T-match I’ve owned ever since my days as a Novice. Without which manual tuning method.
Other brands are better designed. I now own an LDG RT-100 remote auto-tuner (purchased for portable use). The RT-100 allows for compelling a fresh sweep in search of a better match than provided by its memory feature. And you can do it from the control position inside the shack. Just thought I should mention.
So once again the time frame is mid-November. The year’s first, very light snow had melted. My chosen triangular path presented a somewhat greater challenge than had the prior diamond loop. Firstly, the total size would encompass an area nearly twice larger. Nor were any of the prospective three spans entirely clear of other trees standing at least partially between.
For triangle span 1-of-3, after not too many tries, I threaded a spherical sinker and trailing line horizontally at height ENE through the outer-edge canopy an intervening black cherry tree so as to extend my wire still further beyond and so reach corner-point rigging already waiting in a high fork of black cherry tree number three.
For span 2-of-3, I threaded the wire right through the middle of the largest black cherry off all SSW to corner rigging extending out at height from a sturdy white pine of medium girth at the base.
For span 3-of-3, I shot a sinker trailing line over top of a midsize birch tree WNW back toward the antenna-feedpoint black cherry tree.
Here, however, I found myself confronted with a geometry problem which I had put off trying to solve until now. How to thread my line through the canopy of the feed-point tree back to that same high fork from which the antenna originates?
Aggravatingly, it proved quite impossible to do. I tried from standing on the roof. I tried from atop a step ladder in the yard. Once again I festooned that tree (this time on the opposite side) with an array of fluorescent-tufted sinkers. In the end, I gave up.
Admitting defeat, I terminated the would-be loop instead as an end fed. Bitter-ending the wire at a ceramic egg, I rigged a draw cord, pulley and ballast on the house-side of black cherry tree number one, diametrically opposite the starting point. An almost-loop, then, its start and finish separated by the width of said tree’s canopy.
In that configuration, what it amounted to was an overlong inverted-L on 160m, and a bizarrely contorted long-wire on all higher bands. Even so, still it functioned amazingly well. A crossbreed antenna wholly without pedigree, but at the same time, quite an awful lot of it. Over the course of that following winter, still it more than proved its worth, even such as it was.
|
|
Having suffered no frustration at all in making contacts, I decided to leave well enough alone, pondering only once every now and again how I might close the loop. During this interim I had joined an early morning round-table net on 160m. Scheduled for weekdays between 5AM and 6AM, I would set my alarm so as to join it every day before work. Soon it became my favorite on-air activity.
Owning no amplifier as yet, my IC-745 (ostensibly a 110W rig) showed only 85W on both the MFJ T-Match‘s forward power meter, and also a Diawa 2-needle meter. This while most others on that same net were running 300W, 500W, even in one case 1.5kW. Illinois, Iowa and Wyoming all heard me well enough. This while stations in New England heard me only just barely. Fearing any change might make things worse, I left the antenna just as it was until after I might acquire an amplifier, a span of several years as turned out.
Meanwhile I did erect a 48-foot guyed mast, trying out an ad-hoc folded monopole. For a while I experimented with enhancing soil conductivity by spreading graphite atop its minimal 8-wire radial field to a distance of 24 feet (7.3m). On the 160m band, no phase of my experiments gave improved reports on the 160m net verus the still-open Giant Coat Hanger. On those days when the net moved to 80m the monopole did seem just slightly better. On all higher bands the folded monopole proved noticeably less good. And so I switched back full-time to the still-open GCH.
After some years, at the continuing suggestion of the aforementioned 160m round-table net operator in New England plus a couple of others, at long last I purchased an amplifier. An RM Italy BLA600 model, selected specifically as being full-QSK via PIN diode. This on account of shopping around for a new rig, one likewise having full-QSK via PIN diode (not a slow and audibly clacking relay as in the IC-745). Full-QSK via PIN diode being the full consensus suggestion of other ops on the 160m CW round-table net.
Having no longer any concern for the strength of my own signal in the 160m round-table net, I was now free to try out closing the loop. By this time also I had received the advice of David Ring N1EA that feeding via a vertical V (as opposed to parallel open wires) should prove no issue at all.
And so one weekend in December of 2022, I added sufficient of wire to the terminus for extending it downward at an angle to meet the grounded flat braid.
As of this writing, I have employed Version the Giant Coat Hanger as my primary antenna for most of a decade: most of seven years as an open loop, a further three properly closed. These are my experiences of its two sequential phases.
For resorting to a manual T-match indoors (post removal of remote ATU), it was needful to make a chart: settings for C1, L, and C2 for each band. Thus did I discover very sharp tuning. In order to achieve any match, I needed a very deft hand on the dials. For the lower bands, my chart required plural rows: settings for both the top and the bottom of each CW band. For 160m, a setting for the middle as well.
Thus did I operate for the first several years. For having a chart, tuning wasn’t all that much of a problem. Sometimes a chart setting might prove spot on, other times requiring a minor tweak. At first I would do it at minimum power: 10W from an IC-745. Later I installed a coax switch allowing to tune with miliwatts via a RigExpert AA-30. Which remains my habit today.
In due course I began to notice often reworking the same stations with consistently good reports. This while getting somewhat poorer reports from other stations at the same distance lying on different bearings. Already I had suspected such might prove the case, indeed ever more so the higher the band. For being end-fed and running some 135m-plus at an average 12m elevation, how could it not behave like a long-wire? Albeit one those horizontal portion forms an open, near-perfect Isosceles triangle... Hardly could it avoid exhibiting a great many lobes? And an equal number of nulls? Three such sets of paired lobes and nulls overlapping each other at an approximate 60°. Like so with some lobes adding, others subtracting. But with the overlapping nulls still remaining nulls.
After closing the loop my former chart of T-Match settings was, of course, no longer valid. In compiling a new chart, I found bandwidths to be notably wider. Single T-match adjustments afterward sufficed for the entire CW portion of higher bands. And fewer such settings were required on lower bands. On 60m for instance, a single setting sufficed for both Chan 1 & Chan 2, another such single setting served both Chan 3 & Chan 4. This as opposed each individual channel formerly requiring a separate setting all to itself.
As to potential nodes and nulls, these seem to now be fewer. Right away I found a wider variety of stations answering my CQs. This while one or two stations with whom I had used to QSO repeatedly now answered my QSOs rather less often. Which begs the question: have nodes and nulls flattened out? Or have they instead merely been rearranged? Intuition suggests the former. But there remain too many variables for much confidence in a personal opinion so obviously unscientific.
Apology for supplying only anecdotal data. Such an arrangement defies accurate modeling. I am unable to obtain exacting measurements. The heights at each corner vary somewhat. Also the ground beneath is uneven. And most intimidating of all, two of the three spans reach above intervening dense vegetation preventing use of my 100m tape reel without I construct temporary derricks at both ends.
For being a CW-only rag-chewer, I care chiefly about my all-time average QSO duration and little else: LOG . I do like DX, but do not hunt it, preferring to stumble upon new finds. I bail from pileups after just only two tries.
Know however, that on this antenna I have enjoyed to keep many a QSO going for well over an hour: LOG . Including several hour-plus QSOs with QRP stations. Such as the time I stumbled upon a op testing his just-finished magloop at 5W while it still hung from the ceiling of his garage.
My longest CW QSO to date was 189 minutes with K0AF a few years back. My furthest hour-plus QSO was with a station in Greece. Both of those last are on-record with the SKCC, said confirmations going toward my achieving their Marathon QSO Award: LINK . Find me ranked number 21 in that list. How all this might compare to others racking up hit-and-run contest QSOs to qualify for DXCC would be anyone’s guess.
Actually, I greatly favor AN-SOF over NEC for antenna modeling. But that point is moot, insofar as I lack a model of any kind.
The day may come, however, when gnawing curiosity impels I create a matched pair of abstract models if only to compare theoretical lobes and nulls. That day, I fear, continues to loom on a very far-distant horizon.
Sort of... A model anyhow of a theoretical isosclese sky loop, 44m per side at 12m hight above ground and perfectly level. This while fed by wires in a vertical V, one leg 2m above ground, the other reaching to ground. Also with the soil below perfectly level. All of which applies only very approximately to the actual GCH
For 80m and 160m, the patterns are omni-directional. Predicted for 60m were two very fat lobes oriented NNE and SE. Patterns for all higher bands me encourage me to model yet further with an eye toward attempting a hexagon. All this long while, I appear to have enjoyed good fortune by accident. This insofar as on higher bands some principal lobe or other just happened to orient toward europe.
But as mentioned, said model only vaguely represents the actual Giant Coat Hanger in situ. Which several nights of operating 60m while observing reports on the Reverse Beacon Network seems to directly contradict. Especially being picked up in Ireland, which direction had ought to be a deep null. I suppose next on my list had ought to be pulling 60m QSOs from my database to see how the total compares.
Not long before this writing I finally got around to something I really ought to have done at the very beginning: bonding my grounds.
Should it be of interest, I have parsed he following dates from my several pencil-on-paper logbooks.
Key #1: “thru ????-??-??” = until present date.
Key #2: Teal Color = continuity break in story line
The Giant Coat Hanger’s evolution and on-air use follows a somewhat broken timeline.
Equipment changes and virtual QRTs meanwhile occurred.