Experimenting with waveguides using the LiteVNA

[coppercone2]

Bought a new build plate for 3d printer, since the old one keeps loosening even though its torqued real hard. I got it used and cheap even years ago so I guess that maybe it was a bit dodgy to begin with. I don't want to experiment with trying to fix a ball joint, its probobly quite difficult. Set screw + ball joint =

Will be painting two slightly deformed horns after dinner, will get some pictures

[coppercone2]

well primer made this go so much better. For the first one I was not sure if primer would help on the 3d print, since its plastic and stuff, but it helped big time, but the mg chemicals spray paint can is still trash that sprays little globs of paint

[coppercone2]

oh yeah the results are much more clear with the lower frequency, on just the SWR indiana jones test I get 0.4 SWR lower across the board (the launcher is 1.5 and with antenna it goes to 1.1). With the higher frequency it was not so clear. It might be a cal issue of course, that stuff is so touchy.

[coppercone2]

These are the WR62 horns. I want to try to make a RF gasket for them by mixing 2 part silicone with silver powder. Quite the ordeal to make these lol. I forgot to thicken up the walls, and the print bed was messed up, so I had to file them a little but they do work (warped and a offset near the end). I have a new bed coming for future work. Still better I think then what I saw in some old air force training videos on youtube, they look like they were playing stick ball with those wave guide parts.

For gaskets, what do you think the best procedure is for making them? I thought maybe to just make the mold, pour silver powder in there, level it off with a playing card or whatever after tapping on it, then pour silicone on top so it seeps in and fills the gaps, fully cure, trim with a fresh razor blade and demold. Anything more then that I think would be hard, you would need to dry ice or maybe liquid nitrogen lap it I think, if you wanted to sand the silicone nice.. or make a two piece mold I guess, but that is alot of work

[joeqsmith]

Warp is an understatement and the file marks are clearly visible.  You may be better off with sand paper.   Why waste the expensive coating on the outside?   Because I processes so many horns together, some have over spray but it wasn't my intent.   

I haven't considered using gaskets for the connection points.  I've shown you how flat the printed parts are and ran your flashlight test. Hardly a concern for me.  With some of the experiments I ran, I have been using the choked flanges with O-rings to get them gas tight.  Lots of playing around but haven't anything to show for it yet.   

Did you run some sweeps to get an idea how well they radiate, or are you only looking at the VSWR?   If you are, post some photos of your setup.   

[coppercone2]

Its two effects

1) Its too thin and it warped, possibly when I ultrasonic cleaned them in that lab soap powder.
2) the print bed got loosened during the print and it started offsetting when the pull force got too high (near the end of the print the horn grips the plastic better), so each time it got set down on the last 20% of the build, it was like, rotated slightly in a random direction (5 hours into build), so it started looking like the damaged pyramids of egypt with big 'steps'

I torqued the bed alignment screws hard as I could, to the point of damaging the set screws, but it still loosen up. New print bed should come in soon.

Idk if its bad tolerances on the machining, damage, cheap screws or what, but I see that mechanism has a spring pop out of it when you open it up for trouble shooting so F that


They need a collet on the ball joint of the print bed like the 'smiths vise", instead of set screws. I saw someone build on on youtube for a special grinding setup, but of course that joint probobly costs more then a top end 3d printer.

So to clarify some of those marks that go side ways not lenght wise are missing material because the printer offset, but there are some file marks. It needed very aggressive cleanup.

[coppercone2]

As for more photos I am waiting on some wave guide parts so I can do a better test

[coppercone2]

The the summary on this page for wave guide joints is good if you need a refresher like me about the different joint types and what practical effect your choice of joints has.

https://globalinvacom.com/products/flanges-flat-groove-and-choke

[joeqsmith]

As for more photos I am waiting on some wave guide parts so I can do a better test

How are you measuring the VSWR and radiation pattern today?  Are you using a VNA?  If so, which?   

I saw that RS 40GHz VNA on ebay for $5k a couple of weeks ago.  Looked like the drive was dead but it booted up and physically appeared in good condition.  Didn't look abused.    I checked used prices for working ones and seems they were running around 40k.   

[coppercone2]

its the 8510C that has 20GHz gen and 18GHz reflectometer, and the 7mm cal kit with the sliding load and air line, months to fix the system and restore it (the sweeper had the leaked alkaline and I ended up combining two sweepers and redoing a silk screen on the mother board to make one functioning one)

But the cool thing is I have 2 couplers that go WR62 to APC7, so I can try to calibrate out wave guide now by putting the coupler at the end of the setup. But it will take a while to fly that from all over the world

I think I saw that VNA too or something close, but its down right scary to work on those machines IMO. I  think they make you bald with too much exposure to the open chassis with PCB removed (from stress). I cut down when I started lookin in the mirror dealing with that in my spare time. I tried to clean up all the connectors, grease everything, replace suspect parts, and reseat all the components I could find (like screw down caps and even the chassis shield elements that slide around and crap like that), to make sure that the contacts are all good.

It had demons too, it was crashing for some reason, but I dunno it seems to have stabilized some what.


I am gonna find out if you can swab the barrel of  a wave guide by putting never dull wadding on a stick.Don't think I want to be paying for that rhodium plating


seriously my stomach churns thinking about what you might run into that R&S unit lol.. serenity now, potentially add another $50 for antacids

[joeqsmith]

   I've had to repair most of my equipment and maintain it.  I don't get too worried anymore.     I was thinking how much time I would waste just trying to get a service manual and software for it.  Then I am guessing license keys would need to be regenerated.  Get that all done then figure out how to align it.   Maybe a nightmare or a very good deal.   

Do you run the 8510C standalone or with a PC?   What calibration does it support?

[coppercone2]

It runs stand alone right now, I have a computer in the other room with the card and I was meaning to make a hole in the wall to connect to it.

I assume you mean this menu?

[coppercone2]

and this is the standards that I can use

[coppercone2]

and this is the ones I can't use yet because I need the different reflectometer

[joeqsmith]

So for waveguides, does it support the offset short, shims and such?   For the cal standards, does it have a database for the ones HP sold?   If so, I wonder if you load up one of the waveguide kits, how it compares.   Just curious.   

[coppercone2]

not sure yet need to do more research. there is this thing but its too expensive I assume that you just put those things on and do the same cal routine

https://www.ebay.co.uk/itm/382213470112

[coppercone2]

lmfao I remembered  I left some corn in the little steam oven last week and forgot to eat it because of that manufacturer. at least its not smelly

[coppercone2]

you would be looking at the 11644A calibration kits from HP for waveguide cal with the 8510C. At least thats the higher frequency one. The lower frequency one must exist too.

I don't think the disk ever changes the menu options that you see, it just loads values for them. Unless it makes custom standards, which you can add, then I don't know. That may be a whole other feature I am not familiar with. I think I heard about capacitance standards and some weird stuff, not sure how it interfaces programming wise.

[coppercone2]

and for the paint, you should see how much paint dripped on the floor because of the shitty spray can. This was NOT a neat process. It sprays about as good as putting water in your hand and flinging it at something


But I wanted it double sided just incase some paint flaked off inside or something. And if I do metal plate them, it will look much cooler, I could just put gloss or epoxy on the outside

[joeqsmith]

Nice kit.  Looks like they still offer it.   Check the very last page of their catalog: 
https://www.atecorp.com/atecorp/media/pdfs/data-sheets/CMT-Waveguide_Specs.pdf

Yes, what I have shown is basically my poor man's clone of the X11644A with a few other bits.  Do you have to load the cal kit from disk, or is it built into the system?  Can you change the coefficients and save them back to disk? 

Looks like the hams use 10.368 GHz and 24.048 GHz.  You seem to be trying to experiment in the middle of it but with posts about WR42  are you planning to get more accessories for the 8510 to use it at the 24GHz?

I plan to stay around 10GHz as that seems to be a sweet spot for parts availability and price.  To use the LiteVNA with my extender to measure S11 at this frequency, it really needs a new amplifier, coupler and adapters.  I tied a test the other day with some really bottom end adapters and was amazed just how poor their insertion loss was.   You get what you pay for.    That old coupler is spec'ed to 8.3GHz but I have shown data up to 10GHz even with the connectors and low frequency amplifier.           

[joeqsmith]

For the paint, maybe you could get a cheap airbrush and just cut the old cans open and pour out the contents into a sealed container.   OEM may offer some suggestions on thinners and setup.   

[coppercone2]

I don't think its the right kind for that unless you reformulate it. It seems to like bubble and stuff when you spray it out, I had a few spots that were not painting and what I did is sprayed the paint into a little tray and swirled it around a bunch so it thickens up and then you have a few seconds to work with it before it cakes up, in a viscous form, that has low mobility so it wont migrate off the spots that the primer fell off after sanding due to surface tension or whatever the hell is happening there.

Maybe if you can spray it out, dry it out, then add regular paint thinner, it might work? I think like there is butane or something that is dissolved in the whatever to make a unusually volatile solvent. Maybe its just me but it does not appear to behave the same as other paints. I think it dries much quicker. I dunno I guess the SDS does not say anything weird maybe I can drip just a bit of acetone in there to adjust it after it stands out for a little bit and try it.

Maybe its the same, I guess I never tried to paint brush spray paint because its not worth while since there is no silver in it. But the liquid seems weird in the tray lol

[joeqsmith]

They offer thinners for manual spray applications.  Have a look:
https://mgchemicals.com/category/conductive-paint/conductive-paint-thinners/

They call out AR, which I assume is what you used.  Wouldn't hurt to ask them.   

All the test horns were PLA.  I had no problems with the primer adhering to the plastic.  All of them were primed except for one.    If you wanted to try to manually spray them, you may want to ask if there is a better coating for your particular plastics.   

[coppercone2]

Hmm the thinner is IPA based but the paint is acetone based .I guess it works with everything and they just want acetone in the spray one because it evaporates faster?

[coppercone2]

I got a wave guide tuner to try to see what will happen to VSWR with it. should have the setup in like 1.5 months 

i got alot of captains, pilots and drivers im depending on for this setup