3
$\begingroup$

TL/DR: How does someone make a good horn antenna system without using expensively made parts? Welding thin steel for this is not something even pro welders can do (it's a specialty app.) What are DIY options then for joining sheet metal. What is the most careless thing that could still function, even fn well (like 300 sheet metal screws poking inside.)

Beware, this is long: I don't expect to match the efficiency of radar components. I made one out of copper/solder about 20 years ago using 2 pieces of copper and it had a short waveguide section for a probe or loop antenna. It was maybe 22" aperture, I remember getting all the dimensions using a C++ program, drawing the piece in autocad and hijacking the plotting machine to print it out at work lol (it wound up being far from the printed accuracy) ... Copper has skyrocketed in cost since then and it's heavy and deforms easily. ... On the other hand, thin steel requires an exceptional welder I don't even know where to find, and have asked all around before. For very narrow gauge, as in the sheet metal used for the waveguide in a microwave oven I think it's a process involving a roller disc inside the waveguide making spot welds at the closest interval possible and doing multiple passes until it's continuous.

Literature(s) about horn antennas and waveguides just say "smooth" and "continuous", no dings or warpage either. Not what a little bit of that might do. What for example would a rivet's profile do to the inside of an otherwise mirrorlike inside wall with exact angles. Not that getting it right the 1st attempt is ever expected, but I seek thoughts and advice from others before wanting to invest my time and resources into a project. I've seen an example of someone making a horns joining a corner with adhesive backed foil tape (?! they had no seeming concern about this nor data on performance, Maybe they even knew what they were doing I just doubt it.) I don't have any good examples for reference or really any examples of similar steel horn antennae. I think of all the such things involved with how the steel overlaps and maybe using a brazing technique (no familiarity with brazing) or something like closely drilled pairs of holes with twisted wire clamping edges/faces together. L shaped edges and the like create a problem inside with a valley/furrow that could be mitigated maybe a little by filling it with conductive epoxy or solder (solder in reality would be hard to do, it running all over the place or being lumpy.) I've never seen a homemade antenna that was any good besides ones made from copper, and those 2 were smaller than the one I did.

Improve this question
$\endgroup$
5
  • 1
    $\begingroup$ I agree about solder, it's a standard way of joining steel plate - gutters are sometimes made this way. Don't worry about small protrusions like screw heads or rivets. Do worry about small errors in dimensions caused by folding the metal not allowing for stretch around the corner. $\endgroup$
    – tomnexus
    Jan 14, 2020 at 5:10
  • $\begingroup$ Thanks for the advice. I kind of suspected that irregularities like that might not even come close to ruining performance; but nothing I read before alluded to that and spoke only of getting very fine tolerances. Most commercially made waveguide parts and feed horns are very finely made and the prices reflect it, Then there are DIY examples I've seen and certainly respect although they're not thoroughly evaluated or compared for performance against slight variations-- one actually was (reddit); it was mostly for an experiment to 3D print a small horn, painting the surface conductively (works!) $\endgroup$
    – Owen
    Jan 23, 2020 at 1:59
  • $\begingroup$ I should say that every one of the DIY examples used copper, and the 3D printed object covered with a copper-filled paint. Google has targeted me (2 days ago) with an ad for this steel horn antenna: fairviewmicrowave.com/images/productPDF/SH0118N.pdf , min. gain 11dB, range 1 to 18GHz: Its shape looks advanced, with a sort of fold in the center of two opposite sides, I guess to accomodate down to 1GHz given the width and height (16cm x 20cm, big end, no small end given.) I know of 'folded waveguides' and something like that in a variable phase shifter; this the first horn I saw. $\endgroup$
    – Owen
    Jan 23, 2020 at 2:53
  • $\begingroup$ "variable phase shifter", that is a voltage-variable phase shifter with low hysteresis, non-latching, shaped like a hockey puck with a ferrite core and coil around it for voltage control. I forget the typical control dV/dt, I knew the inventor who had mostly in mind that it would be used carrying power in phased array antennas. $\endgroup$
    – Owen
    Jan 23, 2020 at 3:11
  • $\begingroup$ You're seeing a Double Ridged Horn, which is a different beast and not something you can make at home. Usually covers a wide frequency range. The ridge and feed shape is critical to gain, pattern and VSWR. A pyramidal standard gain horn made of sheet metal has a much narrower bandwidth, but can be made to dimensions and will work to spec. $\endgroup$
    – tomnexus
    Jan 23, 2020 at 5:23

5 Answers 5

Reset to default
3
$\begingroup$

I'm no expert on sheet metal fabrication and hopefully someone will come along with a more confident one, but in general I would expect that soldering/brazing is the correct answer.

(solder in reality would be hard to do, it running all over the place or being lumpy.)

From what I have seen of soldering large metal parts (with a torch, not an iron), with the proper flux and temperature for the materials involved, solder should not be lumpy — it should flow into the joint and make smooth junctions under surface tension. (At least as long as you don't have drips of solder running away from the heated work area into a cold area; and if that does happen, you can then either apply heat to melt the drip, or grind it off.)

Improve this answer
$\endgroup$
4
  • 1
    $\begingroup$ Soldering is better, as brazing requires heating the joint up to a red heat. That might very well warp the horn. $\endgroup$
    – Mike Waters
    Jan 14, 2020 at 2:05
  • $\begingroup$ Grinding down lumps as mentioned in other replies I don't imagine would be too difficult. There is an extension shaft for dremel tools, like a gooseneck lamp, long enough to get inside the small end of the horn and a waveguide. AND/OR cutting a pencil style iron off near the handle and fixing it at an angle to a post. It can be inspected inside with a mirror-on-a-stick (?) -- endoscopes are actually getting to be about $80 in the same auto parts store you'd get such a mirror (I don't think any mirrors they sell are actually small enough or long enough to get in a 7cmx3cm waveguide.) $\endgroup$
    – Owen
    Jan 23, 2020 at 1:22
  • $\begingroup$ Mike Waters, I know I really need to learn about soldering steel. I always thought no solder would stick to it-- at the same time, somehow I know I believed that but not realizing how it contradicted how I always filed down the tip of an extremely cheap soldering iron when it fouled-- so I could tin it if I did it quickly on the exposed soft steel (using either paste for plumbing or I think sometimes liquid flux and mostly electronics rosin-core lead solder. It was 20W) $\endgroup$
    – Owen
    Jan 23, 2020 at 3:31
  • $\begingroup$ @Owen Learning about soldering is off-topic for this site, either in questions or comments, but there's probably some good tutorials or videos just a Google search away. :-) $\endgroup$
    – Mike Waters
    Jan 23, 2020 at 23:16
3
$\begingroup$

For very narrow gauge, as in the sheet metal used for the waveguide in a microwave oven I think it's a process involving a roller disc inside the waveguide making spot welds at the closest interval possible and doing multiple passes until it's continuous.

People reviewing TIG welders on YouTube routinely weld razor blades together. Any thin steel that's heavy enough to hold a horn shape without deforming under its own weight can surely be welded by a skilled welder with a decent machine. TIG welds can be quite flat right off the machine, but if you need it absolutely smooth you can grind/sand/polish the seam until it's smooth.

Look around for a metal fabrication shop in your area — if you're not sure where to start, stop by an auto body shop and ask them to point you in the right direction.

Improve this answer
$\endgroup$
3
  • $\begingroup$ Or look for a metals outlet and check their corkboard for fabricators looking for small projects. $\endgroup$
    – Chris K8NVH
    Jan 14, 2020 at 16:46
  • $\begingroup$ I should ask around a lot more than I have been it seems. Thanks for the input. $\endgroup$
    – Owen
    Jan 15, 2020 at 4:24
  • 1
    $\begingroup$ Razor blades, wow. $\endgroup$
    – Owen
    Jan 15, 2020 at 4:36
2
$\begingroup$

For really fine control of soldering, use a soldering gun as a resistance welder. Just cut off the end of the heating tip, so that instead of a loop you have two big copper wires coming out of the gun. Press them onto either side of the joint and pull the trigger. The current passing through the metal will quickly heat it to the point where it will melt the solder. It's much easier to control than a torch.

Improve this answer
$\endgroup$
4
  • $\begingroup$ This sounds like a great idea. I was imagining that I would have trouble using a torch, as I did when using copper instead of steel. Thanks. $\endgroup$
    – Owen
    Jan 15, 2020 at 4:30
  • $\begingroup$ I did a quick search on YouTube for "soldering gun welding" and didn't find anything exact (but there was an interesting video of someone using 12 VDC from a small lead-acid battery for welding). $\endgroup$
    – rclocher3
    Jan 21, 2020 at 20:43
  • $\begingroup$ Wow, just at the existence of a cheap and portable electric ... My favorite lead-acid battery is a Cyclon, a sealed deep cycle series that's really heavy duty and available in many sizes of 12V, 6V and 2V single cells, the smallest I think being the X Cell, 2.17 Volts open circuit and the size about of a small prescription pill jar. There are lithium-ion powered jump starter boxes for cars now, but a very small one I saw (with my own car) could not start it with a totally dead battery. $\endgroup$
    – Owen
    Jan 23, 2020 at 1:05
  • $\begingroup$ I wondered about the possibility too once of using a microwave oven transformer backwards in a DIY welder, and definitely found a video. I only learned to gas-weld quickly for a class in school once and just never had any equipment to learn other methods and be experienced. $\endgroup$
    – Owen
    Jan 23, 2020 at 1:08
1
$\begingroup$

I recommend trying solder first, as suggested by a couple of other answers. That should give you a good electrical connection as well as some mechanical strength. If it doesn't go as well as you planned, put some foil HVAC tape over the joint. You should have enough inductive coupling that you won't need to worry about the conductivity of the adhesive.

Improve this answer
$\endgroup$
3
  • $\begingroup$ As I said in another comment I am just recently starting to know about inductive coupling such as this, realizing that there is an everyday example too sort of, in a microwave oven door. (I posed the question of how those work on Quora but the answers I got from people, all went to explaining something like the screen in the door's window and ignored my actual question (the "seal", that does not look like a seal, maybe being capacitive I wondered, often passed off as a 'metal gasket' too in writing-- but I only ever saw a conductive gasket on a Radar Range (60's door design obsolete c.'75)) $\endgroup$
    – Owen
    Jan 23, 2020 at 1:34
  • $\begingroup$ Before I really had any education in electromagnetism, I remember fixing an arcing problem in my family's late 1980's microwave, putting adhesive foil tape over a folded edge of metal inside that came close to another side of the cavity wall but had a gap of around 1mm. $\endgroup$
    – Owen
    Jan 23, 2020 at 1:41
  • $\begingroup$ ^^^^It could have been all the same piece and folded over towards itself. I don't have a total recall of it, 25 years ago (!) Covering the gap with foil tape seemed like the right idea and it worked $\endgroup$
    – Owen
    Jan 23, 2020 at 2:26
0
$\begingroup$

I would suggest overlapping the edges and using a adhesive such as JB weld. after the adhesive has set you can sand or file the seam to smooth out any squeeze out.

Improve this answer
$\endgroup$
5
  • 1
    $\begingroup$ Thanks! Is it not important to apply pressure and use an electrically conductive epoxy? There's a slew of companies making resins with special properties for different apps. $\endgroup$
    – Owen
    Jan 15, 2020 at 4:36
  • $\begingroup$ If you have enough overlap and you keep the adhesive layer thin enough, you'll have good high frequency coupling, even if the adhesive is an insulator. $\endgroup$
    – mrog
    Jan 22, 2020 at 17:21
  • $\begingroup$ I did not even know that steel can be soldered, with some kind of solder. Is "soldering" in this sense used interchangeably with "brazing"? I unsuccessfully tried brazing a tube in an older-style bike frame, almost white-hot with together a propane and tiny N2O-Acetylene torch from Radio Shack, I didn't prep the surface right for the brazing sticks. I used JB Weld next actually, just to see what it was like, and finally a custom bike mfr sandblasted it all and oxy-acetyl welded it all in <10 minutes. $\endgroup$
    – Owen
    Jan 23, 2020 at 0:29
  • $\begingroup$ MROG- This is exactly what I'm only starting to "get". It seems I've taken things I've read in the past to heart too much. But the question hit me recently even of "how does a microwave oven door complete a faraday cage without any contact." $\endgroup$
    – Owen
    Jan 23, 2020 at 0:33
  • $\begingroup$ How you learn this is beyond me, since I took EMag I and II for EE in college and specifically asked if we could learn waveguides but was told that was out of the scope of the course. (on oven doors^^^^ indeed they originally had a conductive gasket around the perimeter, and now use primarily a so-called "capacitive" seal that almost seems like it is a capacitor to high freq at 1st view but the whole thing is more complicated and only briefly described where I read about it.) $\endgroup$
    – Owen
    Jan 23, 2020 at 0:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .