[NLRS] Waveguide transitions

[Baker, Donn B]

> -----Original Message-----
> From: John P. Toscano [mailto:tosca005 at tc.umn.edu]
> Sent: Tuesday, May 18, 2004 9:11 PM
SNIP 
> I hope you won't mind a few more followup questions.  And 
> since I know 
> I'm not the only 10 GHz rookie on this list, I figured it 
> would be okay 
> to send the questions back to the whole list instead of just to you.

Of course not.  It how we learn !

> Baker, Donn B wrote:
> >>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. 
> 
> Do I have my terminology wrong?  My (mis?-)understanding was that the 
> transition was the piece of hardware that converts from waveguide to 
> coax, and that to make a horn feed, I would go to the W1GHZ 
> web site and 
> get the program that draws a horn template, transfer it to a sheet of 
> brass, cut, bend/fold, and (NOT staple, spindle or mutiliate) 
> but solder 
> the sheet of brass into a 3-dimensional horn, and attach that to the 
> waveguide end of the transition to make a feed horn.  I guess I don't 
> understand your comment, so can you tell me what you meant by 
> it?  Just 
> trying to understand as much as I can here...

This is just a matter of "the way things are done."  Not a biggie.  You COULD build exactly as you describe, but that way you have a horn on the end of the coax-waveguide transition.  No flexibility.  On the other hand if you build a coax-waveguide transition AND a separate horn, to change the horn you simply unbolt the two flages, and bolt on the other horn.

 \
  \         | |
   \--------| |-|
                |  "your way"
   /------------|
  /
 /

 \
  \     |  |     | |
   \----|  |-----| |-|
        |  |         |  "my way"
   /----|  |---------|
  /     |  |
 /
     Flanges (2)
     (bolt together)


> > With strong signals, you DON'T NEED easier pointing !!  (KISS!)
> 
> OK, I am willing to take your word for it, and will KISS off 
> my failed 
SNIP 
> > 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.
> 
>  From Doug's earlier comments, I'm starting to think that 
> maybe I should 
> remove the two 90 degree SMA elbows from the T/R relay also, 
> even though 
> they were factory installed by DEMI and I would hope that 
> they are aware 
> of how easy it is to introduce loss at 10 GHz.  

Trust me, Steve DOES know.  The 90s that he uses are OK... and don't introduce that mush loss.


>I see that some folks 
> have done this, and some haven't.  I was planning to mount the 
> transverter as you describe in one manner or another, but now I'm 
> thinking that the existing 90 degree elbows may be a problem (even if 
> they are pre-fabricated SMA-to-SMA elbows installed by DEMI). 
>  Perhaps 
> two straight pieces of ".141" with an SMA connector on each 
> end (instead 
> of an SMA barrel adapter) to move the relay out a couple of 
> inches from 
> the transverter and horizontal, so that the common lead could 
> come away 
> from the relay with a very gradual curve over to the feedhorn.  

Would work, of course, but it will be a lot less stable mechanically (even though I'm NOT a fan of the way Steve mounts the relay.  I added a spacer-block between mine and the back of the transverter.  Hled everything in place with 4-40 bolts.

>This 
> thought comes mainly from Doug's comments about his problems 
> with this 
> part of his system, and a comment at the last breakfast from Bob that 
> he's never seen a 2-watt DEMI unit that actually put out 2 
> watts.  Maybe 
> there's close to 2 watts at the transmit port, but too much 
> loss at the 
> output of the T/R relay due to the elbows...

No, the fault is almost CERTAINLY that the "2-watt" specification is for the amplifier running at 13.8 VDC, and NOT at the 12.5/12.6 of a lead-acid battery.

> 
> As it turns out, I already have about 3 feet or so of 
> Semi-rigid 0.141" 
> that I picked up at Aurora several years ago for an unrelated project 
> that never went anywhere, so I guess I only need to scare up some SMA 
> males for it.
> 
> It looks like Digi-Key has some at $5.62 and $3.99 each, and DEMI has 
> some that are $1.50 each.  The ones from DEMI (Johnson 
> 142-0694-016) use 
> the center conductor of the semirigid without any sort of pin as the 
> central contact, although the catalog page that their site points to 
> mentions a different part number that also has a center pin (which I 
> presume DEMI doesn't normally stock).  I don't know for sure 
> which style 
> Digi-Key sells, but from the drawing I *THINK* they have a separate 
> center pin.  Any words of wisdom on the choice of SMA's for 
> .141 semirigid?

Do NOT use ones with separate pins !!!  Use ONLY the ones which use the center conductor of the .141 coax as the center pin of the connector.  The Johnson part number is 142-0694-016 for the nickel-plated connector.  You do not need to spend the extra money for Gold plating.  These are the connectors that Steve has listed (on the detail page, below).  You don't want the solder-type with separate pin because of the difficulty of making the solder connection correcly.  If you are using flexible cable then you need the solder pin, of courser, but you aren't so don't use them.  Besides the others are cheaper.  NOTE: The two-hole flange-mount connectors listed (SMAFLAN-2P) are good for making transitions.  And at a good price !!  They have long enough pins and teflon insulation that you can cut off what you do not need.)

> 
>    http://dkc3.digikey.com/PDF/T042/0232.pdf
>    http://www.downeastmicrowave.com/PDF/smass.pdf
> 
> 
> > 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.
> 
> Sounds reasonable, but now THREE of you have said that sharp 
> bends are 
> bad, and sharp bends followed by an attempt to straighten out 
> the stuff 
> are still bad.  So I guess I am a bit leery of buying some pre-made 
> jumpers of questionable history, given my limited resources 
> for testing 
> them.

Pre-made jumpers are almost always commerical/military surpuls.  They will have been made correctly, and will be OK with bend radius.  When we bend the stuff ourselves, without proper tools, the bends kink, ans loss goes up.  Yes, testing is a good idea, but usually not a problem with the pre-formed surplus jumpers.  The trick may be to find one(s) that bend the way you want (need) them to bend.

73 Donn
WA2VOI/0