[NLRS] Waveguide transitions

[Baker, Donn B]

Hi John,
I see that Jon, W0ZQ, as provide some answers, but I want to stick my 2 cents worth in too....  Comments inline below.

73 Donn
WA2VOI/0

> -----Original Message-----
> From: John P. Toscano [mailto:tosca005 at tc.umn.edu]


SNIP

> I did find a couple of postings of waveguide to SMA adapters 
> that don't 
> specify a waveguide type/size, but say only that the 
> waveguide opening 
> is approximately 3/4" wide by 3/8" high (that would be 0.750" 
> x 0.875") 
> in one case, and 1/2" x 1" in another case (0.500" x 1.000"), 
> and 3/8" x 
> 7/8" in yet another (0.875" x 0.375")
> 
> The dimensions of WR-90 are allegedly 0.900" x 0.400" while WR-75 is 
> supposed to be 0.750" x 0.375".  I say "supposed to be" and 
> "allegedly" 
> because my only source for the dimensions of waveguide is a web site 
> found by google:  http://www.wa1mba.org/wavegd.htm  and I 
> have not taken 
> the time to find a more definitive source, but it's probably right...
>
You found a good source.  The table data is correct, and Dick's write ups are great.  The problems you mention (above) are the result of people (surplus sellers) not having the faintest idea what's important re: waveguide.  And they don't care/want to know.  (Can't blame them, I guess.)

"WR-xx" waveguide specifies the inner dimension of the broadwall, in inches.  Thus, WR90 is 0.90 inches across.  WR90 has 0.40 inches as the height of the guide.  You will see this very often quoted as 7/8" x 3/8" because that's typically as close as can be measured with common rulers. As Jon pointed out, the OUTSIDE dimension of WR90 is 1" X 0.5"

WR62 will be 0.62 (actually 0.622)" or 5/8" x 5/16" (sometimes 3/8") as its usually quoted.  WR75 will be .75" x 3/8"  These are just because people don't measure correctly. (A note: there ARE "reduced-height" waveguides, but the diffence from the standard is dramatic... usually at least 1/2 the normal dimension.  All of those you quote above are for "standard" waveguide, just sloppy measurements.)

There are several sizes of waveguide that will work on 10GHz:  WR112, WR90, and WR75.  WR62 will work (with somewhat high attenuation) at 10.368, but it will NOT work at 10.250 (WBFM).  In general, all three sizes can be intermixed in a system (inlcuding WR62) if you're careful, and will accept a higher SWR. (Or provide tuning screws to do matching.)  WR90 is PREFERRED, but use what you've got.  (I'm going to be using a WR62 waveguide switch on 24 GHz, which is WR42 waveguide.  It will take a bit of care, but should be just fine.)


> Anyone know if these surplus waveguide adapters would work @ 
> 10.368 GHz, 
> since they don't seem to be either WR-75 or WR-90?  In 
> particular, two 
> of these surplus waveguide transitions seem to have a very different 
> width-to-height ratio than either WR-90 or WR-75, although 
> the third one 
> I mentioned is pretty close to falling between WR-75 and WR-90 in 
> dimensions so I imagine it should work.
> 
Yes, they would work, BUT... you should stick with the "correct" transitions.  That is, if you are going to use WR90 for a feed horn stick to a WR90 transition.  Also, the choice of coax connector is important.  I would recomend that you stay with SMA... and NOT 'N'. More down below re: coax. 

One additional though... consider making your own transition.  Its a simple job... there are lots of published "designs" for SMA to waveguide transitions.  They can be very simple, and there's bits around to make them from.  A couple of URLs:
<http://www.50mhzandup.org/pg98.pdf>
   This is the Rolls Royce !  But they don't have to be this "complicated"

<http://www.newsvhf.com/trans.gif> 
   Very typical... can be tuned for best match, etc.

A piece of WR90, 1.5 to 2" long, a chassis-mount SMA connector, a piece of brass 1/2" by 1" (to close one end of the WR90), and a piece of .040 - .060 ot thicker brass 1-5/8 x 1-5/8 for a flange, a little solder, and you've got a transition.  The SMA needs to be 1/4-guide wavelength from the closed end (0.367"), and the probe needs to be about 0.3" long (includes the wall of the waveguide).  It'll work.  If you want to try building one, let me know... I'd be happy to help, as would any of the others, I'm sure.

> One of my thoughts was to build a horn on the adapter, in 

Two different things... a transition, and a horn.  Both will have flanges. 

> fact, I even 
> thought it might be interesting to try building a horn in 2 or 3 
> sections that could be stacked according to how much gain vs how much 
> beamwidth you wanted in a given situation, and maybe point 
> that forward 
> (away from the dish) with a coax relay to select either the low-gain 
> horn for easier pointing at strong signals or the full dish for 
> high-gain at the expense of more difficult pointing.  Of 

With strong signals, you DON'T NEED easier pointing !!  (KISS!)

> course, this is 
> probably more effort than the idea deserves, but if only time 
> was not in 
> such short supply for me, I'd love to play around with it, even if it 
> only taught me once again to KISS (Keep It Simple, Stupid!).
 
The feed horn for a parabolic dish antenna MUST match the dish it will be used with.  The feed horn's job is to properly supply the RF energy to the surface of the dish for the maximum efficency.  For the 18" dishes we're using, this means the horn must have about 10-11 dB of gain, and have slightly different gains in the E and H planes.  In keeping with KISS, if you want a horn antenna, build one.  If you want to use an offset dish, use it.  You'll have two different antennas.  Personally, I don't think its worth the effort.  The largest PRACTICAL horn will have 18-20dB gain. It'll be maybe 10" long, and 6" square at the opening.  The 18" dishes have 30-31 dB.  Again, as Jon pointed out, the 18" dishes are easy to use with a little practice.  An 18dB horn is too small to bother with.

A very good source for background on dish feeds, etc., is Chapter 6 (I think) of W1GHZ's "Online Antenna Book."  Google either the title or Paul's call.

One of the reasons I like to use the Chaparrel feed horn that comes with the DSS dishes is that its already a match for the dish its being used with.  Yes, there is the conversion necessary, but its an easy job.  (Well, kind of.)  Building the optimum WR90 feed horn for the dishes is also easy.  One advantage the Chapparell has is that its already properly placed at the dish's focal focal point.  The other horn requires a bit of work to get it there.

> And while I have the attention of some 10 GHz experts, I'm 
> wondering if 
> you have any words of wisdom about the jumper from the transverter to 
> the adapter.  I have some pieces of RG-58 (ugh), RG-142, and LMR-200 
> (all of them are RG-58 size) to which I can easily attach SMA male 
> connectors.  But the losses at 10GHz for any of them seem 
> awfully high. 

Jon pointed out ".141" as the cable to use.  Correct.  And, use AS SHORT A RUN AS POSSIBLE.  Mount the TR switch/Transverter/etc., UNDER the feed arm of the dish, and use 3-6" of cable to go from the TR switch to the feed.  The beauty of an offset dish is that there is NO BLOCKAGE of the dish by the feed.  Anything UNDER the feed and such is out of the antenna pattern.
 
DO NOT use any of the "small" flexible cable coaxes (those you mentioned).  There is no need to, and they suck at 10GHz.  More of a problem than the cable loss is putting connectors on... ANY imperfections will greatly increase the loss.  This isn't 40m where anything will work.  .141 cable is available from DEMI, and surplus, but many of us have a stock of it around.  Ask when you need some.  Connectors are avaiable from Digi-Key at reasonable prices, and are VERY easy to install.  When used with .141, they're actually rated to 26 GHz !  One comment: I disagree with Jon re: bending .141.  It's NOT that forgiving.  Minimum bend radius is 1/2" or more, and you CANNOT straighten that radius out and re-bend it without having the losses go up.  1" bend radius is better.  There is also some "semi-rigid" cable floating around that you can wrap around a pencil, etc.  ONCE.  And don't do it unless you have to.  Don't plan on using RG-8, etc., either (it, and 'N' connectors are rated to 18GHz, but....)

A good thing to do (my opinion) is find some pre-made SMA - .141. "jumpers" FIRST, and THEN plan the layout of the transverter/dish mounting, etc. 

> 
> 73 de W0JT
>