Trap dipole for WARC bands

or as part of

Diamond W8010 Multiband Dipole Modifications

SEE ALSO DETAILS OF MY HOMEBREW TRAP INVERTED-V FOR 10 TO 80 METRES

& MY WEB PAGE ABOUT TUNING COAX TRAPS

----------------------------------------------------------------------------------------------------------

The Diamond W-8010 is a HF Multi-band wire dipole using 6 traps plus two wire sections each side to cover the non-WARC bands from '80' to '10' metres (hence the 8010 coding). Strangely, this model code is a departure from their other models like their W-735, which covers just 3.5 and 7 MHz, and the W-721 which covers just the 7 and 21 MHz bands. By using separate legs and splitting up the band coverage between them, it makes the tuning (and design) somewhat simpler. One of the ideas that they did use in this design (as against other manufacturers' designs seem on the web) was the use of tuning "tails". This means that the wire lengths between the traps is always constant and the fine tuning is done by cutting/adjusting the "tail" lengths. Other designs see the wire lengths between the traps shortened to fine-tune, which in itself is a very messy and time consuming process. The 'extra tail' process - particularly those attached at the traps, means that if you cut too short, then simply add a new tail on a solder/crimp lug a little longer than the last one and start the tuning process for that band again. No joining wires !

DOWNLOAD : Original (english language) Diamond data sheet for the W-8010. W-735 and W-721 antennas (210KB in PDF format)

(All images - click for a larger image)

This is the basic layout of the W-8010, 3 traps on either side of a balun (supplied). The leg shown horizontally above covers 3.5, 7 and 21 MHz while the one showed angled covers 14 and 28 MHz.

One thing I will mention about the traps used in this design : they are LC traps rather than coax types. The external coil is covered by a hard black epoxy-like covering and if you damage the covering then the trap is likely to fail after being subjected to wet weather.

As you can see from the SWR curve at left, the bandwidths reduce as frequency goes down - that is normal - but paying attention to the numerical figures shows that the 1:5:1 SWR bandwidth on the 3.5MHz band is only 24KHz and is 60KHz on the 7MHz band. Unless you closely tune these bands to segment where you normally operate then the SWR may be unacceptable. On 14MHz, the 100KHz 1.5:1 bandwidth is not too much of an issue because most users would either have it set for the CW end or centred on about 14.200. If your transceiver has an automatic ATU either internal or external, then it won't be too much of an issue anyway - simply the efficiency of the antenna will drop as you operate further away from resonance.

Don't get me wrong : if you need a physically short antenna to operate on 80 and 40 metres then this is a good choice. All that is meant is that YOU have to carefully tune it for YOUR needs in YOUR installation.

The very ends of the dipole legs terminate on plastic insulators and to provide for the adjustment tails, the wire length is effectively somehat shorter due to the initial 55cm fold-back, some of which is used for the knot.

Each band is tuned by trimming a "tail". On 10, 15 and 40 metres, it is an extra wire attached at the relevant trap. On 20 and 80 metres, it is the tail at the end of the dipole legs. The diagram at left shows how much you can expect the centre frequency to move at each band for a one centimetre (1 cm) change ON EACH SIDE.

Note that the nylon rope supplied seems to be a marine-grade version and should last for many years.

One of the things that annoyed me about the W-8010, apart from the narrow bandwidths on 80 and 40 metres, was that I still needed to run a separate dipole structure to cover the 10 MHz, 18 MHz and 24 MHz WARC bands. Previously I had used a "ray dipole" that covered all bands down to 40 metres ( i.e. 28, 24, 21, 18, 14, 10 & 7 MHz) but it had to come down for the W-8010 to be put in place. Could I add another wire leg ( or legs ) to each side to add these bands ?

At first I thought about adding these extra bands as separate dipoles and thus convert the W-8010 into a combination trap/ray dipole structure. That would have meant another 3 wires out each side and even though they don't need much separation and can be fabricated such that the higher bands 'droop' from the lowest band, they are a bit of a problem to tune. The interaction between the elements because of the close coupling means that adjustment of one dipole in the ray causes de-tuning in another - so around the loop you go - yet again !

Trap dipole for WARC bands :

The solution was to create an extra trap dipole for the 10, 18 and 24 MHz bands - surely that can't be too hard.... I guess that I should mention that I don't run high power on HF - particularly on the WARC bands - and the typical power level is 100W PEP. That meant that I didn't have to plan on building traps with a huge power rating. As I had a reasonable supply of Belden RG174 miniature coax on hand, I created some traps using that cable. Even so, these should certainly handle 100W+ PEP, maybe up to 400W PEP - I just have no need to try it at that high a power level ! I contemplated using RG58 but at the 100w power level, it just isn't necessary.

I don't really plan to operate much on these WARC bands ( 12 metres, 17 metres and 30 metres ) but that old proverbial Murphy has it that if I don't prepare for it then sometime real soon, I will need to - and then it will be in a hurry. If nothing else, it is a good technical exercise in antenna experimentation.

For those who aren't familiar with it, the Diamond BU-50 balun is just a 50 ohm 1:1 wideband balun (1.8 to 40 MHz or so) so if you are just building a new antenna (rather than modifying one like a W-8010), use any a 1:1 balun that covers at least 10 to 25 MHz. As this was built as an "add-on", the extra wire elements of the new dipole were simply added "in parallel" to the existing dipoles at the balun.

Many of you who build multiband wire antennas will have found the Coaxial Trap Design software by Tony Field VE6YP - { download: coaxtrap.zip & read the HELP ! } and it seemed like a reasonable starting point. I checked my Iocal stocks of PVC conduits then put those diameters into Tony's calculator to see which gave the form ratio closest to the optimum value of 0.45. I then jotted down the dimensions of coil turns, coil length, and coax length for each band (18 & 24 MHz) using that coax. I am not going to tabulate the exact trap details here because you will probably have different coax, or a different PVC former, and you can just as easily work out your own using the software, but the general format was calculated as follows :

18.1 MHz traps : 42mm dia PVC former, 4.4 turns of RG174, total of 2 required

24.9 MHz traps : 42mm dia PVC former, 3.4 turns of RG174, total of 2 required

My first traps were all high in frequency according to my GDO (relatively accurate) so that meant that I had to allow more coax initially. In the end, I used about 8 to 10 % more than the coax length given in the calculator to make sure that it would at least reach the target frequency - so keep that in mind... Remember that you can shorten the coax used in winding the trap - but you can't lengthen it !

The second idea came to me while I was tuning the traps. You need to expand the turns (i.e. pull apart !) to get to the actual frequency you want, remembering the extra coax length allowance. Now in my readings, there are articles that say the trap should be tuned lower than the band and others that say mid-band. (One tuning method gives slightly more "gain" than the other.) Let's ignore that for now. The expansion of the turns is not easy when you have drilled one hole for the coax to start from, and then another the correct distance along for the coax to 'end' through. The start hole is fine - just the "end hole" idea needs to be modified. Instead of drilling just one hole at the "other end" of the PVC former, drill a series of holes close together, forming a slot. Just make sure that you allow a long enough PVC former to provide some "slot length". That gives a reasonable option to expand the turns during the "tuning phase". Just make sure that you either wind good quality insulation tape over the turns OR use some other fixing method to stop the turns from moving after the fine tuning process is complete. Self-amalgamating rubber tape would be even better - the stuff you should use on all of your exterior coaxial connection joints instead of plain old insulation tape - as it won't ever come apart, even when exposed to some pretty horrid weather..

Everyone will have a different method of terminating the trap - wire loops through the ends, screw terminals (like mine),..... so the final trap construction is up to you. You do need to reasonably accurately tune the traps and the two ways that I have done it are :

(1) by GDO (Grid Dip Oscillator - or transistor dip oscillator, FET dip...) - this is the quickest and easiest and provided the GDO frequency calibration is close, then that is sufficient. Simply loosely couple the GDO coil to the trap and locate the dip indicating resonance;

(2) by a RF signal generator and diode probe to a voltmeter - Use a series resistance of 5K to 10K from the generator to the top of the trap (other end at common ground) and use a low capacity/high impedance diode probe at the same junction to read voltage -- preferably on a analogue meter (not a DMM). Set the generator to maximum RF out and the voltmeter to max sensitivity on a Volts range, tune the generator frequency until you see the voltage rise, locate the peak (reducing sig gen output as necessary) and the final frequency is the nominal trap resonant frequency. The RF probe must be very low capacitance or the final generator frequency will be lower than the true (unloaded) trap frequency. You can also use a spectrum analyser in place of the diode probe - however it must be fed via a series resistor of 5K to 10K to reduce the loading on the trap. If you are using this method, the quickest way is set the generator to the desired trap frequency and expand or contract the trap winding to peak the output voltage.

This is actually a 40 metre / 7 MHz trap but it is easier to see the slot at the bottom in the PVC former in this photo, and, as they say, a picture is worth a thousand words.

In this view, you can see the screw/nut trap termination plus the blue crimp lugs (subsequently soldered) with the incoming, outgoing and "tail" wires. In this version, I used 4mm metric screw/nut hardware but could just as easily have used 3/16" whitworth or other convenient thread series. For extra strength, you could use a screw that will traverse the whole former then use it as an anchor by tying the incoming/outgoing wire in a loop around that before going to the terminal lug. Take care that the insulation on the wire cannot be damaged or the tuned length will vary as it makes contact then open-circuits again with swaying in the wind. You probably need to place an insulating sleeve over the thread inside the former to ensure electrical isolation. If you can't find anything else to use, trim a piece of outer sleeve off RG58 coax and use that as it will survive the sun's UV quite well. This technique would involve using a few more nuts & washers and a longer screw (eg 55/60mm long for a 42 mm PVC former) but take the stress off the wire joint. Ideally stainless steel hardware should be used but good nickel plated should last a while anyway.

Having now built coaxial traps for a few bands and incorporated them into physical antennas I have noted that the exact frequency of the trap does not seem to be unduly critical. What must happen is that the traps for each side be tuned to the same frequency !
Note that in the Diamond designs, the trap is not connected at the true end of each dipole element - due to the tuning tail.

Effectively, the extra WARC-bands trap dipole is similar to the original ones except the traps are coax rather than LC series construction.

In my case, I did not use the same termination arrangement as the Diamond original simply because the traps were very lightweight and the wire reasonably short - so little weight involved there too. I simply terminated the wire lengths to the traps in a crimp lug (subsequently soldered) and they attached to screws/nuts through one side of the PVC former.

The wire lengths for the elements are as follows - but note that if your traps are significantly different to mine, some length variations will occur on 18 and 10 MHz :

The tuning tail lengths for 18 and 24 MHz start at about 20 to 30cm and are trimmed back from there. Just a tip - don't cut the tail on the 10 MHz element wire (length R) - wind the piece below the termination insulator back along itself a section at a time and recheck the centre frequency. Most importantly, adjust BOTH sides of the dipoles similtaneously, whether it be a tail-cut or a wind-back..

I used standard plastic egg insulators for the ends of the 10MHz dipole segment and then just some reasonable quality nylon rope down to the fastening points.

Here's another hint for success :- keep the extra dipole a reasonable spacing from the other two legs to reduce inter-leg coupling. The original Diamond brochure indicates a 30 degree angle/spacing between the wire legs and if possible the extra dipole segment should be spaced another 30 degrees away. Note that it doesn't really matter if the "30 degrees" is horizontal or vertical - just so long as the legs are separated... For instance, all of the legs might terminate on a single mast pipe on each end with the 80 metre end at the top, the 20 metre end lower down ( eg 1/3 from the top) and finally the 30 metre end (this new dipole) lower still (eg 1/2 way down) on the masting pipe.

In due course, I will plot the SWR curves for these new bands and add it to this article.

The final outcome : an all-band HF trapped dipole antenna covering the following frequency bands : 3.5 , 7 ,10 , 14 ,18 , 21 , 24.9 & 28 MHz, one balun & one coax feeder.
In my installation, it is formed as an inverted-V with approximately 90 degrees between the legs & given the centre height and the downward tilt of the wire legs, it should be nearly omni-directional.

This article has intentionally not been a "hold my hand" style because everyone does things differently. You may have different PVC, different coax,.... and even the mounting method (flat dipole / inverted-V etc) and all will affect the final outcome. The critical info here is the wire lengths in the table above as it gives a starting point for those who want to build their own version and have the ability to wind up and tune some simple coax traps. Of course, the same concepts can be applied to making up your own trap dipoles for other bands.

My aim was to provide info/incentive to others by letting them know how I did it, and why.

The experience is creating this add-on has led me to other coaxial trap projects - for example: see my HF Field Day ideas page

If you build one of these add-ons, let me know how you got on....email to Feedback Form

-------------------------------------------------------------------------------

02 Oct 2009 : Just an addendum of interest regarding the Diamond W8010 - a warning if you want to take it that way.....

From an incoming email :

Yesterday I was worked by the W8010 antenna through lamp PA 1,5 KW (GS-35B russian lamp) and breaked traps on 10 and 15 meters band. So big power was deformed traps and SWR and resonance points was go far up. From 28.500 to 30 MHz, from 14.170 to 15 MHz , from 21.160 to 21.400 -))

With interesting I find and read your article http://www.vk4adc.com/w8010.php . With this thinking I have the folowing question: can you tell me measured inductance of each traps in W8010 ? I want modernize the traps for working with 1.5 - 2 KW RF Power. Unfortunately when my traps was undeformed I did not measured the L.

Many thanks !

Andrey, (callsign deleted)
Russia, Belgorod

-------------------------------------------------------------------------------

My emailed response :

Andrey

You cannot measure the inductance of the traps because the capacitance built into the construction will cause measurement errors - you can only measure the trap resonance frequency and you know what they will set up for be anyway. My W8010 is up in the air so I cannot just measure the traps.

It sounds like you could have had the insulation on the enamelled wire break down and you now have shorted turns on the faulty traps. You could try to remove the black 'baked-on shroud' and count the turns on each one as they will have to be replaced anyway.

Diamond's power rating of 1.2KW PEP is probably too high for their method of construction, and should probably be more conservative at around 500 watts. If you were putting 1500 watts of carrier into it, or had high compression on voice/SSB, it simply isn't designed to dissipate the heat due to losses in the traps. I used to sell these from my business and usually warned customers about using high power linears with them.

You can still use the basic W8010 construction and build up coaxial traps (using suitably rated cable) and place them at the same positions - however the lengths will probably need adjustment on any band lower than the replacement trap on any given wire. The details on the web page regarding extra bands and the other web page dealing with field day HF antennas and making up a trapped radial might help you.

73

Doug

-------------------------------------------------------------------------------

Added 24 Oct 2009, from an incoming email :

Evening (morning) Doug,

Great to read the information on your site about the modified 8010 dipole.
I use (and cloned) this dipole already for over 15 years, with great enthusiasm.
At the moment it is in use with a SB220 feeding it, no problems with power, only one of the traps shows a little brown coloring.
Once I used a w8010 clone during the JOTA and it was producing sparks of over 10 cm at the end of the 20 metre legs, great for the scouts! So make the ends not too sharp, bend them back.

During a overhaul and cleaning I measured the traps as follow:

trap N (wnd/len) L(uH) Q Rs f(dip MHz)
20m 13(16mm) 8,5 130 5,8 31
40m 17(22mm) 13 118 4,8 38
80m 68(87mm) 80 220 7,8 23
Core PVC 40mm, wire 1mm lacqued, length=wnd.length+30mm

{ eg 20m trap, 13 turns over 16mm of former length giving 8.5uH inductance, a resulting Q of 130, effective series resistance of 5.8 ohms and a self-resonant frequency of 31MHz. }

I use these values to clone the antenna for friends.
My construction notes are only in Dutch...:( { but Google Translate will make them readable, - or - see below }

Please pass the values to Andrey (or put them on your site)

Next week we are going to build a number of them on the national meeting of VERON.

Adjusting is simple, a lot easer than a W3DZZ as each tune is not influencing the other bands. Only the ends are a little sensitive the height and moisture. The 20m ends are most difficult to tune as there are some (parasitic) dips on 15 and 13 MHz.

Your extension for 10/81/24 differs a lot from the w8010 design. The 8010 works with coils as L and ens wires as C bringing the extension in resonance at a certain band. This works perfect on high power as the high voltage is only at the end of a adjustment element. Calculation for this are found on www.smeter.net and described by Jerry Hall K1TD (ex K1LPL) in QST sept 1974 and described in an article by Rick W5RH.

You use bandstop tuned LCs like the W3DZZ combine with the nice tuning tails lucky the traps haven't a that high Q so the voltage at resonance is reasonable. I tried an 18 MHz additional dipole, but it ruined the 20m tuning...:(

Together with a friend of my, PA2G , we build this antenna and published the experiment in our league magazine, a summary of this you can find on http://home.versatel.nl/bartsimpsons/HamRadio/antenne/homemade_loaded_coil_dipole.htm

{ local rough copy of the translated web page here }

I'm the guy on the photo with the white shirt... Here in Holland the houses have only a little garden (mine is 15m) so this antenna is ideal as you can bend the end and run 80 to 10 m without tuning from your city garden. Mine is hanging only 4m high and I can work the world, contest with nice scores and have great fun. The balance in the dipole also limits the nearby interference, a problem that is strong with G5RV and other assymetric antenna's.

Good luck and a lot of fun,
perhaps see on the band once..

73'Thieu pa0m@amsat.org
PA0M and swl NL199 (editor swl coloumn VERON league magazine)

Thanks Thieu. I am sure many of the visitors to the web page will find your information helpful in either repairing or cloning the W8010. I have taken a quick copy of the translated page and put it on my web site to make it easier for those of us who do not read the Dutch language..

-------------------------------------------------------------------------------