Experimenting with waveguides using the LiteVNA

[coppercone2]

for that slotted guide, I think I would put some thin gauge brass springs (like a caliper) under the movable part, so when you insert it, it spring loads itself into position. I am not sure if you did that, but the mechanism sounds a little loose. Spot welder seems useful, my mind goes towards repurposing the leafs in a banana connector to act like a slide spring.

I think I also have some section of brass finger stock some where (RF gasket) which I never used because it was annoying to attach to stuff, but since my spot welder finished I have been meaning to try it for something.

I think there is alot of fun to be had with etching thin brass parts for making weird stuff (i.e. diy rotary joint).

#1 obstacle to working with sheet metal : clean all the work surfaces to more then 5x5 inch space 

[joeqsmith]

The problem with making things more complex than they have to be for a one off experiment, it wastes a lot of time.  The bits I have put together just have to work well enough to demonstrate the basic ideas behind them.   

The original WG consisted of just the slot.  The Teflon sleeve kept it centered.  I added the copper guides to help keep it aligned so I could add a pointer and scale to make the measurements.  I wasn't even going to have a scale but rather just mark it with a grease pen and measure it with calipers later.    Then I wanted to show the ferrite and needed to be able to take my hands off the probe. 

The best solution, hook up the microwave counter with it's GPS reference and measure it directly.  Forget all that mechanical stuff from days gone by.  But the series is about waveguides.    

[coppercone2]

I think there are fast general techniques with standard parts/blocks to make these sort of things work better with minimal effort but there is no good guide on how to do it. I wrote it because someone might follow you and if they know some neat trick to add features with 2 minutes extra work using clever equipment/etc, it might make evolution occur, maybe someone can make cheap advanced DIY machinery if there is a few possible improvements listed... and it gives someone a reason to repeat your work with small modifications because its easy to just think there is nothing left to do there and its done unless you wanna shell out hard core with the castings, heavy duty electroforming and machining. There is an appeal to light duty brass working making this stuff, especially if it can be made better. I think its actually accessible without having to get serious training in 'related fields' .

[joeqsmith]

I think outside of making some very simple parts like I have shown in my experiments so far,  you would be very limited using the techniques I've used.  I think for the these lower frequencies, 3D printing has some merit.   You may be able to control the dimensions tighter than beating on brass with a hammer,  make more complex shapes, include standard flanges allowing a hybrid approach.   

The attached link is a paper on directly plating conductive plastics.   One example they show is a horn.   The surface finish is very poor for the band they are attempting to work in.  We have some different plastics on-order now and plan to compare a few different approaches.   If I can find a combination that I like, maybe we can try printing some more complex shapes.       

https://people.duke.edu/~bjw24/Publication98.pdf

I'm not so sure we couldn't plate the higher resistivity ESD plastics as well using higher voltages.   Lot's of experiments to try....

[mr ed]

Mcrowaves are fun even beyond what the vna can do. Some uwave things do cost but gunnplexers can be had new for $100 on digikey, used much less. You can also make waveguides and horns by hand if so inclined. True, performance wont be similar to a Sage gold plated machined and polished guide but you can still experiment plenty. Attached are some home made horns for 25GHz which work well enough. I use solder paste, brass shim stock, tin snips and a torch.

[joeqsmith]

Nice looking horns.  Measuring the speeders in the neighborhood? 

Found an oscillator under modules but no gunnplexers.

https://cdn.macom.com/datasheets/MACS-007800-0M1R00.pdf
https://www.digikey.com/en/products/detail/macom-technology-solutions/MACS-007800-0M1R00/12627111

[LaserSteve]

Take a look at " SWR Autotesters"  on Ebay.

Steve

[mr ed]

Gunnplexer   MACOM   MACS-007801-0M1R10

https://www.digikey.com/en/products/detail/macom-technology-solutions/MACS-007801-0M1R10/12627587

[joeqsmith]

Thanks.  Here's the datasheet.
https://cdn.macom.com/datasheets/MACS-007801-0M1R10.pdf

[joeqsmith]

Printed horn after wet sanding and primed next to metal horn.   New filament in foreground.   Lot's of prep work left.

[joeqsmith]

The plan is to start with the two short PLA horns from the on-line site I linked.  One with the raw printed texture, the other wet sanded.  Both will be coated.    Then I plan to try copper.   Depending how this works out, I start working with the larger horns then onto other plastics. 

[coppercone2]

I forgot I had a WR42 adapter. I will try a WR42 horn

oh god, free cad programs. I might resize it after I get a degree in computer science.

[joeqsmith]

Nice. Post some pictures of your setup once you get going.

[coppercone2]

the problem is going to be figuring out how to resize this crap. IDK why they don't put a percent resize thing, instead its gotta be some crazy shit

Actually I think I had a program that has that on the other PC, I will look at it some time.

[joeqsmith]

the problem is going to be figuring out how to resize this crap. IDK why they don't put a percent resize thing, instead its gotta be some crazy shit

Actually I think I had a program that has that on the other PC, I will look at it some time.

Resizing, software?  Must have something to do with the adapter you mention.     

[joeqsmith]

I was asked about using automotive copper head gasket dressing to coat the waveguides as a cheaper source.   I have a few different ones I use.

The problem is the carrier is not going to allow the copper to form a bond and it will remain an open.  The material also stays soft.  It may be cheaper but I doubt you will get any performance from it, not to mention, you may not want that sticky goo all over your equipment! 

[joeqsmith]

Even after several hours, you can see the copper gasket material is still soft.  Once the engine were run, some of the dressing I use will harden but I don't think our plastic is going to like those temps.

[joeqsmith]

Mixing up some solution to attempt to directly copper plate our PLA filament.   Better to start with something small that I know we can plate, like a nail.   Sanded it down and cleaned it with IPA.   After rinsing with water, looks like copper.  Note the end of our wire...

[joeqsmith]

Before attempting to plate a large horn, better to run some tests on a section of filament.   This conductive filament is over 1kOhm at an inch.  So I am using a 50V power supply.   I started out around 40V for about 5 minutes.  The problem is the plastic gets warm enough to get soft and would most likely deform.   

You can see though it does plate.  I measured about 20mm of raw vs plated and the resistance is about 1/3. An improvement but still poor.

[joeqsmith]

Back into the bath running around 20V for another 10 minutes got it down to 20 ohms.   Still not great.

[joeqsmith]

I reattached the wire around where the plating had started and continued at 4V for another 10 which got me down to 4 ohms.  Then another 10 minutes which got me down to 2 ohms.   

Shown uncleaned, you can see the shiny copper where the clip was attached.  Maybe the conductive paint performs better.

[coppercone2]

Well there is no horn for that wave guide size ready for download, just two examples.

Good luck with the plating BTW. I figured out I need a hull cell to get any consistent results and that is a ways off.  That is a complicated operation IMO to monitor the solution quality also, you need lots of stuff for good results.

[joeqsmith]

Now I'm with you. 

The coating arrived and I sprayed down the smaller non-conductive PLA horn that I had wet sanded and primed.   Pain to spray into the rectangular section and may try brushing.   Making complex parts this way will require a split.  This part has been drying for about 7 hours but it's humid and cold here.  It will need more time to cure.  Still for the first coat, it seems promising.   

I also sprayed the raw printed part. The coating adheres much better to the primer.    Depending how these two compare will determine how I proceed.   

[coppercone2]

gold plated probes would be good for testing weak substrates like that. And for tests on samples you can glue copper foil pieces (electrodes) to different parts of the circuit and then plate it with wires coming out, so you can can probe directly under neath. I am suspicious of that measurement being off because of high contact resistance.  EEvblog has a video that shows how much better gold probes are for testing dodgy surfaces.

I am not sure if those are gold probes or if its reflection from the copper, but its good information for other people that might have missed it.

[joeqsmith]

https://probemaster.com/8000-series-kits/

Still, not a good meter for measuring low resistance but the question is how it radiates compared with others.