3D Printed Resin Dielectrics - mmWave Game Changer for antennas?

3D Print UV resin dielectrics are changing the way experimenters like me design and build microwave and mmWave antennas and components in our home labs and workshops.
Rogers corp recently announced Radix (TM) a new 3D printing UV resin for SLA/DLP printers which has a hugely improved performance as a dielectric for creating lenses and beam-forming elements for microwave and mmwave antenna systems. The feature sizes that are possible means we'll be able to create 3D printed graded-index or GRIN lenses for focussing and refracting microwave and mmWave signals in the range from a few GHz up to at least 47 GHz.
The material has a dissipation factor/loss tangent of around 0.004, which is an order of magnitude better than most UV resins. It beats the ABS filament from Preperm on water absorption and looks like it is going to make possible things we haven't even thought of yet.
www.anycubic.com/collections/...
www.elegoo.com/collections/3d...
• Development of a 3D Pr... Fortify
Garcia and Chisum, arxiv.org/pdf/2003.09004.pdf
Images from Rogers Corp rogerscorp.com/advanced-elect...
and Coonrod's Corner • Coonrod's Corner 2021 ...
More detail about the Luneberg lenses at 3D Fortify's site 3dfortify.com/3d-printed-rf-d...
and on their KZbin seminar • Development of a 3D Pr...
Other images and scribble are my own. You can tell, they're a bit rubbish.
Chapters:
00:00 WAIT! WHAT?
00:39 Graded Index 3D Metamaterials
02:01 Rogers Radix(TM) UV 3D Resin

[@joshmyer9]

FWIW, my Saturn can get 10um layer heights in theory, and is typically used at 50um for normal quality. Its pixels are roughly 50x50, but the new Phrozen mini 8k is 22x22um, also with 10um layers. This may change your calculus for what's practical a bit.

[@MachiningandMicrowaves]

I guess it depends on the resin a bit, they must have had to make compromises on how well it sets to get the dielectric stuff printable, but if they are talking publicly about W band, that implies wall thicknesses of about 20um. I'm working on some solid machined lenses at 122 GHz/2.44 mm wavelength but they are just solid, no graded-index stuff and I can't machine voids obviously. They said 85 GHz is being tested, but that's using the printers from 3D Fortify, so I don't know what level of performance they can get.

[@christianbowes7907]

Agree, 150um is in line with FDM printers . 10-50um is typical for resin layers

[@bdykes7316]

This is revolutionary. It has a low cost of entry, a short time to test designs, and fundamentally different design constraints.

[@MachiningandMicrowaves]

Interesting times indeed.

[@engrenage]

a possible workaround to water absorption would be to soak the 3D-printed wave guide / antenna in liquid paraffin ; I've used this for oil-cooling (immersing computers in a tank) precisely because its dielectric is close to that of air. for 3D printed parts this would definitely require a vacuum tank. oily liquids are usually hydrophobic so this should help. come to think about it, if the dielectric of a printed material is low, it should be able to improve it by soaking the part in an appropriate fluid. some years ago I read about metal 3D printing through sintering ; since the parts were sponge-like and very fragile, post-processing would include dipping the print in molten copper : thanks to capillarity the copper would fill in the voids between the nodules, resulting in a hybrid material with decent mechanical properties. a similar approach could be used for wave guides and dielectric properties.

[@MachiningandMicrowaves]

There's a lot of interestng work going on with dielectric materials for sure.

[@tissuepaper9962]

RF engineering is one of those disciplines that is "indistinguishable from magic" per the old adage.

[@MachiningandMicrowaves]

"Dabbling in the Occult and messing with Things from the Dungeon Dimensions since 1968" should be my slogan. I really hope this 3D print dielectric lens tech is going to be affordable for home experimenters.

[@Bianchi77]

Keep it up, thank you for sharing :)

[@MachiningandMicrowaves]

This dielectric print thing is going to have some very interesting applications, we just need to imaging them...

[@igorb4650]

Whow, cool stuff! Looking forward for some rf magic.

[@MachiningandMicrowaves]

Here's hoping the resin is affordable and the high refractive index material appears for sale

[@ivanlam1304]

Perhaps you could 3D print the equivalent of Optical fiber using some of these materials and this could be an alternative to metal waveguide

[@MachiningandMicrowaves]

Dielectric waveguide has been used for a long time but indeed you can get self-focussing effects like in optical fibre using gradient index printing techniques

[@Ivan.Wright]

That's pretty cool. What do you look forward to trying out with this new technology?

[@MachiningandMicrowaves]

Moonbounce communications needs an extremely clean antenna pattern so you don't pick up the noise caused by the ground, buildings and trees being 280 degrees Kelvin warmer than the cold sky around the moon. Being able to suppress the sidelobes of the antenna feed better should increase the signal to noise ratio on receive. I'd like to try omnidirectional antennas with very compressed vertical patterns for rain-scatter and snow-scatter experiments and try making Fresnel Zone Plates and.... well, things that I don't even know about yet. I need to go and get some sleep!

[@gregdownunder]

A friend of mine wanted to know where yesterdays video is on this magic stuff. I was half way through it an now is gone. PS I know nothing about radio eng or machining but enjoy your content

[@MachiningandMicrowaves]

Bit of a glitch last night, I couldn't uncheck the notification box and as the vid isn't finished, I pulled it sorry. Should be published around 17th but I have some others to put out first. It's 3.15 am here and I just finished making an X band radome and mount, but sadly no time to film that one. I have to be at work in five hours! My new studio seems to be working well, but still some minor usability issues to fix

[@smellycat249]

I wish I was a 10th as smart as anything you just said. Stay in school kids.

[@MachiningandMicrowaves]

Everything useful that I learned in the last five years came from the fabulous folks on KZbin. What a stunning resource it is. Also a cesspit, but at least it's not F***b**k! OMIGODS I am overexcited about that resin stuff. SO MANY IDEAS!!!

[@auxchar]

​@@MachiningandMicrowaves I've got a Chinese mini lathe and a 3D printer, and some programming ability. I don't know a whole lot about radio, but I know enough to know I'd like to learn more. Got any ideas for something interesting I could get started with for one of those little TV tuner USB SDR dongles?

[@PaulDominguez]

The Phrozen Sonic Mighty has a build vol 20 x 12.5 x 22 cm. As someone that lives in a rural area with limited to cells service, I'd love to see ideas for 3D printing or machining 4g/5g antenna 😁

[@MachiningandMicrowaves]

No news yet about the availability of this material. Back in the day, I made a mesh reflector in a roughly parabolic shape using stressed rods and fishing line and fixed a 3G MiFi dongle at the focus to get 2G internet service at a rural location a long way from the nearest cell. I have to do something similar at home if I lose the VDSL fibre service as we have no cellphone service here. I put a MiFi on an antenna mast and raise it above the rooftops. One option is to use a slightly curved sheet of mesh up a pole at 45 degrees and a printed or machined lens to focus the signal, so you get a reflection off the sheet nice and high above the local clutter. "Flyswatter" antenna. Problem with 3/4G is that the wavelength means lenses and reflectors need to be at least 8 wavelengths to start working, preferably more, so as 1800/2100 MHz, that means around a metre is the smallest effective size.

[@PaulDominguez]

@@MachiningandMicrowaves thanks, I'm started learning about antennas when my DSL service was terminated in my area. It's been an expensive endeavor. I built a cellular router with Raspberry Pi and EM12G modem. Testing different antenna and service. Great channel. So glad KZbin recommended it

[@C_Dana]

Love your channel but what happened to the latest video of yours titled "RADIX part one v0.4"? Can't access it anymore.

[@MachiningandMicrowaves]

I accidentally sent out notifications for a draft version. I'll be publishing the finished version on 17th Feb, and part two soon afterwards

[@MachiningandMicrowaves]

I hope to get two more videos out shortly, but I had some urgent jobs to get out and haven't had time to set up to film them because of building my studio. I'll have to do a quick video on the setup, it might be something to put on the second channel. I wanted to get the studio absolutely right so I could show finished parts and do explanations much more effectively.

[@jhawk6554]

I know some of these words!

[@robertwatsonbath]

Bit rusty on this stuff so bear with me - most equivalent dielectric approximations (Maxwell-Garnet etc) require that the size of the inclusions (spherical air-filled voids) be much smaller than the wavelength of the wave *in the surrounding medium* (Rogers magic stuff). Since the wavelength in a material scales by a 1/sqrt(Dk) don't you need a better than the numbers you suggest? So at 24GHz with Dk=2.8 your 625um is more like 374um?

[@MachiningandMicrowaves]

I think the rule of thumb of 1/20 of a wavelength probably includes that effect as a fiddle factor, finger-in-the-air sort of estimate. It might be 1/30 or 1/15, or something else entirely. Modelling it would be a computational nightmare, especially as the material becomes more sparse with Dk less than 1.2. Sadly I missed the live seminar, otherwise that was one of the questions I was going to raise, like where does their estimate of maximum frequency vs feature size come from. I suppose we could ask them as a follow-up question, but at the moment, I'm coated with oil and steel swarf, trying to extract a long sharp bit of it from the palm of my hand. It went through my glove while I was CLEANING MY LATHE. Too dangerous, this cleaning lark, that's five cleaning injuries so far this year, it's a dangerous sport.

[@robertwatsonbath]

@@MachiningandMicrowaves Makes sense. My usual rule of thumb for EM is 1/10 wavelength in the material. There are some magic sizes of inclusion to avoid also (resonance effects) otherwise doesn't work an an equivalent dielectric (did some work on scattering form "spongy" hailstones years ago, air/water/ice mixtures).

[@engrenage]

@@MachiningandMicrowaves so much for avoiding jobs with grease and noisy machines XD

[@icebluscorpion]

WOW 😮AWESOME 😍! DAMN it always comes to resolution though 😒. Need a better display for my printer😑

[@MachiningandMicrowaves]

So many ways to spend money!

[@icebluscorpion]

Oh yes definitely 😅. Especially when it's worth it 😏

[@JeffreyBoye]

4pi steradian coverage electrically steerable mm wave disco ball... in space?

[@MachiningandMicrowaves]

Got it in one!

[@johnnycash4034]

Well well well Which printer would you say is a go?

[@MachiningandMicrowaves]

That's the tricky question. Need to look at a wash/cure unit as well. I would be driven by my need for casting patterns for aluminium and bronze rather than for the dielectric stuff. I fancy a DLP job, but not going to jump in yet, there was a crowdfunder for an Anycubic DLP, not sure where that got to. Elegoo Saturn is £490 or so with reasonable build volume, but it's the whole ecosystem that would influence me most, slicers and wash/cure and offers on resin and reliability and taking a view on where the company is going. Or just use a pin and blindfold

[@SgtSeth]

Could this be used for VHF/UHF ?

[@MachiningandMicrowaves]

The issue is one of scale. A typical lens is around 10 wavelengths across, and even a dielspike needs to be around 0.7 wavelengths across and 4-5 wl long, so it gets very tricky below 3 GHz. Lens mass and cost rises roughly with the inverse cube of frequency, so a 1 GHz lens would weigh around 1000 times that of a 10 GHz lens and be ludicrously expensive. The region from 5 to 50 GHz is where these lenses work well in terms of practicality and cost. You could build a UHF version using lego bricks or some sort of dielectric fluid in a shaped container, but metal antennas rule below 1 GHz

[@evzone84]

I love your videos but I just hit the porridge transition point for my brain. This is all ελληνικά to me.

[@MachiningandMicrowaves]

I'm just fascinated by the graded-index approach to beam forming. Such a geek. Sad, innit!

[@rhama8751]

What???