Homebrew improved UnUn Design...

Status
Not open for further replies.

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
I've been working on creating a decent Unun that is good for full legal limit power, and provided good coverage from 1.6 - 30 MHz. Achieving that isn't all that hard but doing it in a way that will not present a hazard has been one of my main goals. Though I attribute some of the alleged problems to either pure fiction, or poor design. Nevertheless my intentions have been to build something that was:

I purchased a waterproof electrical box from Home Depot as my case, and it resembles pretty much every image I've seen of commercial Ununs. The link to the electrical box is here:

4 in. x 4 in. Junction Box-E987NR at The Home Depot

That provided a suitable outdoor housing (it's waterproof), with easy mounting options for mounting at the base of your antenna (vertical). It has plenty of room to fit my core choice, which is a pair of Amidon FT-240-61 ferrite cores. You can obtain these for about $13/ea from the fine people over at AllTronics:

Amidon FT-240-61 | Alltronics

Now since Home Depot annoyed me to no end, after getting the electrical box I went quite literally across the street to Lowes to obtain most of the rest of the parts. You'll need the following items (regardless of where you get them) ideally these will be stainless or alternatively zinc plated:

(2) 1/4-20 1" - 1 1/2" hex bolts
(8) Suitable 1/4" washers
(2) 1/4-20 hex nuts
(4) #4-40 screws
(4) #4-40 nuts

You will need an SO-239 connector I have been using American Radio Supply for my connectors and such, but you can get them from DXEngineering, or ebay or where ever you like. I personally love the prices over at ARS and he ships FAST!

SO-239 UHF-Female 4-Hole Panel Mount Coennector (ARS-0511A)

You'll also want a pack of ring terminals for 14-16ga wire with the ring small enough to be used with the #4 screws. You will also need 14-16ga ring terminals capable of be used with 1/4" bolt, I got mine from DXEngineering at the same time I purchased the radial ground plate. They have them 20 for $5:

http://www.dxengineering.com/parts/dxe-225rt-20

You'll also want 14ga stranded (You can use solid core enameled wire, but you'll want to put PTFE tubing around it, so just avoid the hassle) PTFE insulated wire. There's lots of sources for that, but since you don't need very much (like 4 feet worth or so), I bought a tiny quantity from ebay at:

10 Feet 14 Stranded PTFE Insulated Silver Plated Copper Wire | eBay

14ga will handle maximum power, and the PTFE insulation will handle extremely high temps without catching fire, it is after all teflon and they put that in frying pans. Lastly and for safety concerns you'll want to obtain some 3M 27 Glass cloth electrical tape, to wrap the FT-240-61 cores with. This will provide a little extra insurance against fire and the like in the case of high loads, shorts, and bad impedance matching. You can get that from Amazon for about $15.

What you'll want to do regardless of the wire source, is cut the wire into two (2) five foot sections. Here you face a choice, bifilar or parallel wire. To be honest I've done both and have personally not noticed much of a difference. In fact all the photos I've seen of commercial Ununs are all using parallel winding, so if that's easiest for you, you can do it. Otherwise you can create your own bifilar winding by taping the ends of your 5 foot sections, and then gently twisting the sections around each other. You don't need to make the twists tight, a nice loose job will do. I've personally used parallel winding in my core but you can do whatever you like. You will want to leave 4-5 inches on both ends of your winding because you'll need to make some connections when you're done, and if you leave them long enough you can avoid having to make superfluous soldiering that otherwise could of been avoided. I also recommend marking the two ends of the wire that will be your positive wire so you can identify it easily after you're done wrapping the core.

At this point, you'll want to wrap the cores with your 3M 27 glass cloth tape. I've wrapped both separately then wrapped them as a unit. But you may just want to wrap them as a single unit from the start and avoid the extra taping. To be honest I have no idea which is better. Theory would say wrapping them as a unit is better due to concerns over altering the thickness of the cores. But I'm not sure a thin layer of tape is going to make THAT MUCH difference overall, in fact I've read that the differences are pretty negligible. The fact of the matter is, it's most likely to effect the far ends of the range the unun will cover, and I'm pretty dubious of anyone operating much there. I know I don't operate much at 30MHz nor at 1.6MHz.

After wrapping your cores, you're ready to begin winding the wire around your core, I've done a lot of research on how many turns. I've read everything from 15-18 turns, but honestly that's an unruly mess I've done it. It does improve your low end response slightly but it's a real pain in the back side. I've been using 12-13 turns with excellent results, and suggest you try the same this will give you pretty good response from 1.5-1.6MHz thru 30-40MHz. You can read all about this here:

http://webzoom.freewebs.com/g8jnj/33ft Verticals and Ununs.pdf

He did a lot of research into these devices and I decided to take my 12-13 turns directly from that document.

Also one of the things you'll see on some of the Do-It-Yourself unun designs on the net if you google is that they use a superfluous piece of low loss coax (RG-8X etc). Well there's a much better way to design this thing than that, also avoiding the coax component you'll avoid additional soldier connections that can go bad, or break or whatever over time. My design ultimately uses only ONE (1) soldier joint, the rest are lugs, which also have their problems obviously but diagnosing the problem with 3 lugs is a relatively easy problem and can also be easily fixed and done relatively quickly.

Since it's late at the moment and we're in the middle of a tropical storm (or certainly getting gusty now), I'll post up a wiring diagram, and some photos in the next day or two so you can see how this all comes together. I'll be assembling the Unun from scratch (sans the extra core and the 3M 27 tape which unfortunately I don't have on hand yet .. they're both on order sadly). But you will get the basic idea of how it all goes together.
 
Last edited:

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
I forgot to mention in my parts list, that you will also need two (2) 1/4-20 wing nuts. Though the wing nuts AREN'T mandatory they make things much easier once it comes to attaching your Unun to your antenna and ground system. You can skip the wing nuts if you'd prefer and add an additional two (2) 1/4-20 nuts. My personal preference is for wing nuts, so that is what I used. I also highly recommend, that if you're constructing your own home brew antenna that you obtain a 3rd "set" of 1/4-20 gear (Nuts, Bolt, washers, and wing nuts) for attaching to. But that is beyond the scope of why I'm writing here.

Update as well, we lost our internet (thankfully not power) when Sandy came by South Florida so I wasn't able to post until today. Based on the current shipping estimates I should have all the piece parts for this unun next week and will photograph the assembly. In the mean time this is a basic wiring diagram of the unun I will be building. Please not, the windings in the diagram are NOT how they should be on your actual core, on the core they should be tight and right next to each other (or bifilar) not all sloppy as my diagram indicates. I simply couldn't do better with my drawing program.

One last note.. The red wire, heading to the radiating element from the core... The one hoping the gray "common" wire. There is NO attachment between those two specific wire at that point. In the actual build (which you'll see) the red radiating wire goes directly to the bolt attachment, and nowhere else.

 
Last edited:

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
I've had fun researching best practices when building these things, this is sort of the culmination of what I know at this point. Who knows, in a year I might have a better idea and design (maybe using Iron instead of ferrite cores). I do know that this thing should have no problems handling full legal power, and have great response (because of the two cores). This unun will be driving my home brew 33ft vertical (which may become a 43ft vertical before too long). I admit I did cheat some on the antenna side of things, in so much as I am using the radial plate from DXE as well as their wire kit for connecting the Unun to the ground and the antenna. Mainly because I couldn't do better for less money. The radial plate while not exactly mandatory sure makes things easier when it comes to running your ground plain.
 

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
I'm just waiting on my 3M 27 tape to arrive before putting this project together and uploading photos of the assembly here. That should arrive tomorrow, it was supposed to be here today; BUT with Sandy wreaking havoc on the east coast one day late is understandable.
 

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
Due to some time constraints today, these images aren't as complete as I would have liked them to be, however you should get a pretty rough idea of how to assemble your own. You will need a 5/8th drill to cut out for the SO-239 in case you needed that bit of information. You will also see little bits of tape along the the length of the conductors, that is/was to keep my pair of wires aligned relatively straight. I used 3M 27 tape for that. I didn't link directly to these images as they are.. Rather large. You'll also notice that there are two (2) cable ties, I used those to secure the ends of my windings so I didn't have any issues with the turns coming loose. From a 10 foot length of wire, I folded that over and cut them to produce two (2) 5 foot lengths. After winding I had about 18 - 24 inches left over (approx.).

Two FT-240-61 Ferrite cores:
Click

Same cores taped with 3M 27 Glass cloth tape:
Click

Here's the cores wound with 14ga PTFE stranded wire:
Click

Here is the whole thing assembled:
Click
 
Last edited:

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
Final thoughts on this project... If I had to do this again with this specific design, I think a 5" x 5" or even a 6" x 6" box would be nice, though everything does fit pretty nicely in the case I got a little more "wiggle" room would be nice. I also might be inclined to wrap the cores with the tape not once, but twice or, alternatively wrap the cores separately than together. I also recommend using soldier flux on the SO-239 connector center pin, it helped. I used Kester #2331-ZX flux, though I despise the pen tip delivery system I got this from Radio Shack on eBay. The pen works great for connectors and is rubbish for coax braids (unrelated to this project).

Oh the holes at the base of junction box I used are #8 (I used #8-32 machine screws), that can be used along with a pipe clamp to attach your Unun to the base of your antenna system. The final parts list is as follows:

1 - 4 x 4 Junction box (link in first message)
2 - Amidon FT-240-61 ferrite cores (link in first message)
1 - SO-239 Panel mount connector (link in first message)
2 - 1/4-20 hex bolts 1 inch to 1 1/2 inches in length (I used 1" serrated flange type)
2 - 1/4-20 nuts
8 - 1/4" washers (4 per side)
2 - 1/4-20 wing nuts (or two (2) additional 1/4-20 nuts)
1 - 16-14ga #4 - #6 stud ring terminal
3 - 16-14ga 1/4" stud ring terminal
1 - Roll 3M 27 Glass Cloth Tape
4 - #4-40 Screws (for SO-239 panel mount connector)
4 - #4-40 nuts (for SO-239 panel mount connector)
1 - 10 foot length of 14ga stranded PTFE insulated wire (You could alternatively use 16ga)

Optional:
1 - Tube Kester #2331-ZX solder flux
1 - Red electrical tape (to mark plus (+) wire when wrapping (you could use a red pen too)

Tools needed:
Drill with 5/8" drill bit, plus drill bits for #4 screws and 1/4" bolts I also suggest a reamer for your 5/8" SO-239 hole. You'll also need a wrench for tightening the bolts and a phillips head and possibly a flat head screwdriver.

Hope someone finds this information useful, or at least informative. 73's!

Resized images for linking to forum, clicking the an image brings you to the full size version of the image.

Two (2) FT-240-61 Ferrite cores:


Both cores wrapped in 3M 27 Glass cloth tape:


Cores Wound with 14ga stranded PTFE wire:


Fully assembled:
 
Last edited:

prcguy

Member
Joined
Jun 30, 2006
Messages
15,344
Location
So Cal - Richardson, TX - Tewksbury, MA
Nice work. One test to see how good it is would be to make another then connect them back to back and measure loss across the freq range of interest. The goal would be under about .1dB per balunn,
prcguy
 

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
I thought I would add some final information to this thread. I provided really only one school of thought on the design of this project but there's obviously more than one way to skin a cat; or build a matching transformer. I thought I'd discuss wire choices in this post and cores in another post. This is going to be kind of long winded and I'm sure will raise the hackles of some people reading, ah well such is the nature of the internet.

WIRE
Obviously one of the key components to a matching transformer is wire, and there's lots of kinds of that floating around. So what kind of wire should you use? Well there's people that will say you should only use teflon coated wire for xyz reason, and there's people that say you can use any kind. The truth is both camps are right and both camps are wrong too. The wire you use should be capable of handling the load you place on it or into it. One of the key factors to remember in this project is to remember that the RF power being placed into a matching transformer is AC and not DC. That matters because the thickness of wire for AC and DC are different for the same load. Because simply put (and I mean VERY simply put) DC is always on and AC isn't. So that's the first thing to understand when gauging wire for your application you need to look at it's ability to handle AC power.

Another thing to consider is that work being done in a circuit generates heat. Because power (or watts) represent work energy being done within a circuit. The wires inside of your matching transformer will become warm under load, this happens due to loss. No matter how well you wrap your matching transformers core, and no matter how much attention to detail you make. The conducting wire is imperfect and a portion of power flowing through that wire will be converted to heat. The more power applied, the more heat is generated it's a fact of nature. Matching transformers work in part due to loss inside the transformer itself so there will be heat generated directly proportional to the amount of power (watts) you put in. In other words, work makes thing hot.

Lastly lets talk about the properties of the wire and it's jacket (insulation) a bit more in depth. Wire will generally fall into two broad categories they are "stranded" and "solid". A lot of matching transformers are in fact designed with both types. One immediate advantage to solid core over stranded is that attenuation in solid is lower than in stranded. One immediate disadvantage to solid core over stranded is it's harder to bend around a core. But having said that the attenuation problem in stranded is not as big of a problem over a short distance and the length we're talking about is around 5-6 feet. One of the other problems of stranded and is much more pronounced in lesser quality wire than in higher quality. That problem is breakage of the strands. Lets imagine you have a 19 strand wire and 3 of the strands break during your fiddling around, that is a loss of 15.7% of your conductive surface. Solid core doesn't suffer from this problem, there's only a single piece of wire to concern yourself with. Either you got it all, or you get none of it. But solid wire does suffer from fatigue to a greater extend than stranded wire does. Over time bending it will break it, it's also can break due to nicks and cuts. With respect to stranded having higher attenuation than solid at higher frequencies, the specific frequencies we're dealing with on HF (1.8 - 54MHz) aren't all that high and the losses are less drastic given the frequencies we're dealing with and the short distances we're covering.

With regard to the insulation of the wire.. There's a camp of people who will insist that the only wire to use is PTFE (Teflon) wire and/or enameled solid core wire. PTFE insulation is favorite for a couple of reason (1) It has good RF insulating properties and (2) It's virtually impervious to burning. I'm not sure why enameled solid core wire is chosen to be honest other than maybe some sort of tradition. I understand why solid core is chosen but enamel burns at high temps and is as likely to cause as many HV problems as PVC insulated wire. Typically people will buy solid core wire and shove it into PTFE sleeving but that isn't because it's overall "better" the number one reason I've seen quoted is availability. More places sell PTFE sleeving than sell small quantities of PTFE extruded wire and it simply becomes a matter of finding the part (wire) that fits your bill and your wallet. PTFE wire can exceed $2.00 per foot and even then you find yourself having to purchase $500 worth and that is just plain silly unless you are going into business for yourself. It's impractical for a home brewed project.

And lastly the most common insulation type you'll find will probably be PVC. It burns relatively easily when you're dealing with HV (high voltage) arcing. It can also cause arcing due to microscopic holes in the PVC insulation created during manufacture. The use of this type of insulation in an matching transformer shouldn't be completely disregarded however. It all goes back to the amount of power you're putting in. If you're only putting in 100 or 200 watts PVC stranded speaker wire might be suitable to the job. But care must be taken to prevent HV arcing problems between the wire and the core of your matching transformer. But more on that in a later post.
 

acyddrop

Member
Premium Subscriber
Joined
Nov 1, 2011
Messages
322
Location
Delray Beach, FL
This next section deals with the selection of the core for your matching transformer.

Cores
Core materials, that is the toroid at the center of your matching transformer; comes in many varieties and each comes with pros and cons. But for our purposes the materials fall broadly into two categories Iron Powder and Ferrite. The first question that often comes up is how large a core needs to be for a given amount of power and that isn't a straight forward answer. Back to the previous discussion with regards to work being done produces heat. The same holds true for your core material, as more is asked of it the hotter it will become. There are several factors that contribute to the heat generated inside of the core, those are: Cross sectional area, the type of material the core is made from, and lastly and more specifically the voltage and frequency at which we're operating.

Core saturation can be affected by AC and/or DC inputs (signals). Saturation will degrade the permeability of your core causing impaired performance or for it to out right fail. Saturation is measured in gauss. Ferrite has a lower gauss (around 2k) vs. Iron powder (around 5k). So it can be seen that Iron Powder can handle much high gauss before any significant saturation effect become apparent. However overheating of the core will usually occur before saturation when the given frequency is above 100KHz or so, and in both Iron and Ferrite core materials overheating will be the first cause of losses rather than the saturation issues.

Iron Powder cores are low permeability and are more suited to high power applications than than are ferrite material cores. Now this is where things get a little complex. Because ferrite cores require few turns for a given inductance, and when the same voltage is applied across the decreased number of turns the flux (gauss) will increase. But with that said, either material can be used in a matching transformer but there are trade-offs for choosing one material or the other. Ferrite cores will require you to apply fewer turns and will give you a higher impedance factor on a turn by turn basis and will couple better overall. However Iron powder cores will require you to use more turns and will give significantly less impedance per turn and will not couple nearly as well as ferrite. However Iron powder will allow you to apply more power and are more stable over a larger operating range.

At this time I'm choosing not to cover the various core materials that each type of core (Iron Powder/Ferrite) come in. Accept to say that if you're choosing a Ferrite core you will probably want a "Type 61" material or "Type K" (both have trade offs) and if you're choosing Powder Iron for your core you will probably want "Type 2" or "Type 6".
 
Status
Not open for further replies.
Top