Experimenting with waveguides using the LiteVNA

[joeqsmith]

Part 5, a touch of quantum physics and demonstration of a homemade Electron Paramagnetic Resonance (EPR) spectrometer.

[joeqsmith]

The link is to a paper on capacitive vs inductive iris coupling to a cavity resonator.  It describes the pros and cons of each.   My homemade EPR spectrometer is using an inductive iris.  I had tried a few different geometries including capacitive coupling.  I was using copper foil tape that I had cut the slits with a razor blade and then sandwiching it between two sections of waveguide. 

https://dspace.mit.edu/bitstream/handle/1721.1/4939/RLE-TR-106-04711125.pdf?sequence=1

[joeqsmith]

The piston was originally a solid brass bar.  This caused problems when trying to tune the cavities resonate frequency.   I had shown an old HP adjustable short that I was originally going to use for the cavity but the screw pitch was too course and it had too much backlash.   The piston it used was a far better design.  It used finger contacts and was silver plated.     I have now changed my cavity to using a very similar piston.   The tuning is now very stable.   

The iris is the next thing that I need to address.  The adjustable thumb screw is just too sloppy.   The plan is to change to using a micrometer for this as well.   

[KE5FX]

Part 5, a touch of quantum physics and demonstration of a homemade Electron Paramagnetic Resonance (EPR) spectrometer.

That is one seriously awesome video.   

[joeqsmith]

New iris with micrometer stub and finger style piston installed.  Tuning is more stable  now.   

[coppercone2]

http://ro.uow.edu.au/cgi/viewcontent.cgi?article=2946&context=theses

might require a macrovna

very interesting paper, and it shows where that weird waveguide plasma jet thrust thing the Chinese made came from, and describes the history of similar apparatus. a truly enlightening paper

he gives essentially blue prints on the end

and then there is this refined document with much less detail that was made further on
https://ro.uow.edu.au/cgi/viewcontent.cgi?referer=&httpsredir=1&article=1472&context=engpapers

[joeqsmith]

Next I would like to try some sample tubes actually made specifically for EPR.  I suspect with the higher purity quartz and thinner wall, it would help improve the sensitivity. 

https://sp-wilmadlabglass.com/product/suprasil-epr-tubes-5mm-od/

[coppercone2]

I finally tried to plate my 3d printed horn antenna with janus green B, SPS, PEG4000 copper plate.

It worked very well all things considered.

I bought some larger bottles to store the solution (media bottle)


The biggest problem is the dimensions of the conductive paint, I am still working on getting palladium strike working to plate these parts without conductive paint. The tin chloride I got is severely degraded (only 20% of it dissolved).

But I am very confident that it can be done, so long you build a fixtured plating tank and precision bend plating inserts.

I think around 8-12 GHz you can get very high quality parts made from the plating.

its actually smooth and feels like putting copper against your skin and even has the same drilling problems as copper does with nasty burrs.

Glad I decided to keep with trying to plate it instead of trying to do crazy metal casted parts. I was certain I was going to get some flaky mess but it turns out really nice if you manage it correctly.

[joeqsmith]

Show some good photos of what you have.

[coppercone2]

I am waiting to do the 8-12 GHz horn because this one has a busted flange that I filled with JB weld, and its above my test frequency, it only has 3 mounting holes as a result of this. I am going to make up 2L of solution and do a better plate on a higher wave guide size that I can actually measure on VNA


1) I got a 4 inch ID copper tube segment, that is ~0.05% phosphorous, which is ideal for electroplating, compared to random scraps I use that I can form into a proper electrode
2) I got 5um polypropylene fabric coming in mail to make anode bags with
3) make up 2L of plating solution so I can plate without the antenna sitting on bottom of tank almost coming out of water (that is, it can be suspended instead of moored)

there is 100% copper coverage the black is a failure of the kool-amp silver plating to get into the crevices. all defect appear to be because of the poor paint job (it was painted with spray paint squirted into a cup and brush painted after being swirled for a while to degass it and periodically had more paint added to make it liquid, real bootleg)

If you don't do 100% job on the kool amp, it leaves silver oxide (black) on the "not fully rubbed" parts of the surface, my brillo can't get into some of the channels in the corner etc


my fresh tin chloride should come in soon so hopefully I can get away from that god damn conductive paint crap. you could sand the paint down before you work but sanding it was driving me crazy, it keeps clogging sanding sticks and you end up with a 'zoolander' situation where you have 500 clogged manicuring products in the trash


I also bought a 24/40 glass vacuum filtering funnel (non frit but with holes), fine filter disks and a 2L erlynmeyer heavy walled flask so I can filter my baths through a vacuum system using a aspirator to get crap out after a run, and hopefully for the larger tank I can use my battery powered perestalic pump to pump out the electrolyte instead of trying to pour big containers into funnels

I also have my berylium copper fixture springs (curl it up into a spring and clamp it on the object undergoing plating) and titanium tweezers to make the whole process a little cleaner. I still need to get a squirt bottle for distilled water, It got really messy when I kept trying to pour it into a normal funnel to filter it out and swapping containers etc

[coppercone2]

hmm I started to rub on it hard with a coin to "burnish" it and the surface popped off, then I basically peeled the whole horn like a heavily armored chocolate

the plating is fairly strong but the adhesion to the paint is poor

I would not call it a fail, but you do need to be gentle with this method.


Basically, if you try to burnish it hard (scratch with coin), the area under the stressed area will un-glue from the paint and peel off, but I don't expect the wave-guide to fall apart if you handle it reasonably, but its sensitive to point stress, so adding force spreader beneath the flange bolts might be a good idea if you over tighten them


I did however only gently wipe down the surface with alcohol before plating, it was not de-greased properly. In some spots, the paint peeled off, granted the paint might not have been adhered great, so maybe I can say its possible to adhere the plating well enough to make the paint the weak spot, not the copper paint interface. 

[joeqsmith]

While we were not successful, you may have better luck:
https://www.eevblog.com/forum/rf-microwave/experimenting-with-waveguides-using-the-litevna/msg4888817/#msg4888817

I obviously gave up on the idea of plastic parts.  It's one thing to make an antenna but I think to make something like a magic T is going to be difficult.

[coppercone2]

I think your current density is way too high or something, mine looked like that when I was going 6x the advised current limit because of a NASA-esque calculation fail

For burn out, I doubt you can do it. Only dissolving it (microfluidics type 3d printing) IMO can leave a copper shell really intact.


You have carrot going on, I think thiourea makes it carrot color, mine got like that when I added "tarn-x".


I need to remake my solution again, very clean, and try it at low density. I think if you get the mixture correct it should actually plate out copper colored. The problem is when you start getting high rates, it has a weird crystal structure. Its kind of like tearing a thin cut of soft styrofoam (the white one of beads). Its not gummy enough and you can actually tear it.. I think the more you stay at the minimum rate, the more ductile it is. It very much reminds me of the wrappers on chocolates when torn. I think its basically 'pre stressed' material. It would make a great wrapper for high end wines or something lol. I always wondered what metal that is, I think its just a consequence of plating. If it was like soft ductile copper, it would not unwrap.


I still did not try stirring or temperature control, I need to know what it does without control measures to see how it behaves when its 'wrong'

[joeqsmith]

I think you misunderstood.  The part on the left was printed copper after sinter.  The part connected to the meter was the same part prior to sinter.   Note the shrinkage.  The large part would shrink to the WR90 dimensions.  Or at least, that was the idea.   

There is no current.  Rather a furnace and crucible. 

[coppercone2]

Oh I thought that was a thick plating and a attempt to burn out the filler like those 3d printed waxes. For getting controlled dimensions from embedded material... I don't even wanna think about it.

[coppercone2]

https://www.eevblog.com/forum/3d-printing/copper-injection-sintering-using-3d-mold/?topicseen

[joeqsmith]

https://www.eevblog.com/forum/3d-printing/copper-injection-sintering-using-3d-mold/?topicseen

This is pretty much what we tried to do but with a printed part.  I can't see using their setup to injection mold something like a hybrid T for example.  Then there is the problem of shrinkage. I doubt you could get a dimensionally good part.

[coppercone2]

probobly not but these results are starting to at least look pretty nice now

I was thinking maybe round things. Squares and rectangles probably always end up messed up

[joeqsmith]

I was thinking to try and experiment at higher frequencies.   There are reasons to do this with with that EPR spectrometer.  WR90 is fairly large.  I can't imagine trying to use this technique with even smaller parts.   

[coppercone2]

I still don't see how you are going to get ANY results, without electroless plating, for smaller frequencies. Just getting the paint to be correct on the apature prior to plating throws the dimensions out of whack. I think it might be the only good way.

but it is still interesting to see the technology got perfected that much


However, it almost seems like its a reasonable alternative to hot casting, because you don't need to pour molten metal, so safety factor goes up, like for the biconical antenna someone on the forum casted and then post machined, I can't blame anyone for not wanting to pour liquid metal and I think its a reasonable investment for lower frequency antennas.


personally though, doctors grafting skin off my ass might be preferable to wrangling with a filament printer  . Using the molding machine they made looks better but I am sure it must have plenty of demons not shown on video lol

[joeqsmith]

I still don't see how you are going to get ANY results, without electroless plating, for smaller frequencies. Just getting the paint to be correct on the apature prior to plating throws the dimensions out of whack. I think it might be the only good way.

but it is still interesting to see the technology got perfected that much

For the low frequencies I was playing with initially, that hobby brass was fine.  Even the soldered up PCBs worked well enough for the demonstrations.  None of that was plated or dimensionally good.  Surface finish was poor...  When I started to think about putting together a simple EPR system, I knew I it would be a waste of time to try and make all the components using such crude methods.   

However, it almost seems like its a reasonable alternative to hot casting, because you don't need to pour molten metal, so safety factor goes up, like for the biconical antenna someone on the forum casted and then post machined, I can't blame anyone for not wanting to pour liquid metal and I think its a reasonable investment for lower frequency antennas.

personally though, doctors grafting skin off my ass might be preferable to wrangling with a filament printer  . Using the molding machine they made looks better but I am sure it must have plenty of demons not shown on video lol

Start by sketching out how you would mold the hybid T using this method.   For something simple like a horn, I would just use copper plate.   For the X-band that coated, 3-D printed horn worked as well as the manufactured part it was modeled after.   See attached.