I 3D Printed a HUGE Fresnel Lens
Вставка
- Опубліковано 10 лют 2025
- How I'm creating Fresnel Lenses and Focusing Mirrors for microwave and millimetre wave radio systems with my new SYIL X5 CNC machine, Bambu X1C 3D printer, Lasers and Electroforming. Why electrochemistry is a novel solution to creating microwave and mmWave antenna feedhorns. 5G and 6G wireless use these frequencies, but I'm using these tools to experiment with long distance communications and even Moonbounce.
Huge thanks to my Patreon and Ko-Fi supporters. I use those funds to pay for tooling and materials to machine custom antenna and radio system parts for experimenters who would otherwise be unable to get precision parts like these.
BUY ME A COFFEE? ko-fi.com/mach...
MY PATREON PAGE: / machiningandmicrowaves
Sorry to hear about your accident Neil. Wishing you a swift and full recovery. It's definitely the fun that counts. A good friend of mine always says "if it's not worth doing he'll probably have a go at it".
Remembering that my grip strength and dexterity are impaired is the main problem now. Things like glass bottles and full containers of bird food don't bounce well on terracotta tiles. I've put a mesh catch-net below my bench vice to catch carbide cutters that succumb to gravity so they don't chip, and I bought a whole box of market pens as they always disappear underneath machines when dropped. I'm doing the strengthening exercises, but it's nerve damage that's the problem. I just have to apply twice the effort that I would with the burned hand. It's totally recovered thanks to some magic membrane dressings. Nice bit of medical technology at work there.
I had no idea how much I have been missing your channel until today!
@@fretless496 i'm still getting back up to speed after my accident, but I'm hoping to put out some videos based on the 90+ hours of recordings I've made over the last six months. My work for television has rather pushed UA-cam into the background. I was hoping to get another video out on Friday because it's extremely clear that if I put one out on Sunday or Monday, it doesn't get any views so I'm probably going to wait until next Friday before I put another one out and then hopefully I should slip into a release schedule at least once every week or every two weeks
WERE BACK, WE LOVE AND SUPPORT YOU
I am totally humbled by the response from my lovely audience. I had a notification that I'd reached 50k subscribers recently. My mother is very proud. That's half way to getting a cheap plaque from UA-cam. Yay! Well, half a "yay".
one of those rare nuggets of information I've picked up from somewhere is that the reason for the layer of nickel over the gold is that without it the copper diffuses through the gold over time contaminating the surface...
Anyway, its nice to see you back.
It's a compromise. Nickel is an extremely lossy material at microwave frequencies, so I'd need maybe 7-10 micrometres of gold, which gets expensive! Using a micrometre or so of gold will mean that the copper forms a solid solution over time. Copper is a better conductor than gold at RF, so the only issue is that long-term, the solid solution surface might start to oxidise. I've never tried working out how fast the diffusion happens. Same would apply to silver under the gold I think. I sometimes use a chromate conversion treatment on aluminium (Alochrome 1200 or Surtec 650), and I guess I could use a suitable lacquer on the gold. It would certainly be interesting to find out the diffusion rate of copper through gold
Which becomes a problem when you want to use gold-plated copper, but don't want a ferromagnetic material in there
@@L1ama unless the gold is at least six skin depths thick, which is "only" 2.2 micrometres at 47 GHz, so it's not difficult to get that. At 10 GHz though, with 5x larger linear dimension parts, around 25x area plus needing 5 micrometres thickness, that's about 50 times the amount of gold, it gets silly very rapidly. Really the gold is only there as a thin protective layer to stop oxidation.
@@L1ama Typically palladium is used instead of nickel when goldplating PCB's when it has to be non-ferromagnetic.
@@Gin-toki Ah, good to know
I don't really care about the details of any of this s but I appreciate how intelligent, educated, well spoken, and experienced Neil is! Great visuals too.
This is brain food. 😂
Glad you are back, I'm ever waiting for the next tidbit of RF and manufacturing knowledge.
@@Wreaktifier I hope to get another video out next weekend. Fingers crossed (one one hand)
God help me. This was interesting as hell and I actually understood a fair amount of it. Ooww! Just read your injury description; I may need counseling after just that. Get well.
You are definitely being swept into the Vortex! I'm just having so much fun messing about with stuff, even after doing this for more than 50 years, there's SO much new and challenging things to investigate and play with. I'm just working on a vid about Backward Wave Oscillators (Carcinotrons) that's about as niche nerdy as it gets. I embrace my inner nerd shamelessly. Thanks for watching!
It’s exciting to hear about the possibilities in a modern fab hobby shop! Thank you for the video
for induction heating you will need a graphite crucible that reacts with the induced current, I have melted "stuff" that is non magnetic with such. They tend to burn in open air however but that can be limited by silicate or zirconium paints to limit oxygen exposure of the graphite. The paints also help insulate your coils when putting metal directly within them so less sparks and unhappy machine noises.
I did consider trying a MUCH higher frequency like 13.56 MHz and using that to generate eddy currents in compressed pellets of aluminium chips, but I suspect the oxide coatings would still be too strong to release the molten metal and it would need a centrifuge approach to get the molten metal into a solid cast lump. It all sounds like way too much effort given that I have two large bins full of milled aluminium chips and long stringy lathe turnings, and they have some coolant/oil mixed in with them. I did wonder about welding up a stainless crucible and trying to heat it in a furnace with an argon purge, but I think it would be better to find a local processor who could deal with the chips. Offcuts weighing a few ounces each melt fine, perhaps I should stick to melting those.
Indeed, piching up a close to impossible challenge is fun to do.
Glad you are doing better and happy to see you back here!
Hope your recovery continues!
W.R.T. the crumbly gold it rang bells…
Gold electroplating can be very hit and miss. I remember a problem with trying to electroform gold pillars for a novel chip bonding application. I think one of my former colleagues fixed it by using some kind of AC plating process, and managed to make fully dense gold reliably.
Good luck!
The trick seems to be pulses and reverse-polarity electrocleaning and a reasonable flow of electrolyte, plus going slowly and getting the pH and temperature just right. Also, sacrifices to relevant deities, incantation of the correct spells and crossing my fingers very hard. Getting the copper to fill right to the bottom of the into the narrow gold-lined grooves seems to be the hardest part
Welcome back, we were worried about you!
tbh, once you get above 23cm it all becomes black magic to me!
I've been tied up in the 500 THz region for a few months on a big project, but now I'm returning to the relative sanity of the Jiggerhertz Magicks
I had to come back and review your video again, lots of good info.
Thanks
I was looking at solar reflector to increase heat, your a good resource for me.
I accidentally melted some PLA at the focus of an 18 inch/450 mm Edmund Optics dish.. Bigger solar reflectors make scary amounts of heat if the figure of the dish is reasonable accurate
Good to see you back - I think a lot of people have been wondering where you were!
Glad you are making good progress and your injury hasn't dented your enthusiasm :)
Keep looking after yourself and very much look forward to the next video.
I think if you are going to adopt "Freedom Units" you should start expressing velocity as furlongs per fort-nite.
milligilberts per megaparsec are probably my favourite units, but I'm firmly in the acre-feet camp for lake volume measurements. My garden is a furlong by a chain, so it's easy to visualize that with a foot depth of water. OK, so it's also close to a megalitre, but acre-feet is much more human-scale somehow
@MachiningandMicrowaves - glad to hear you are on the mend. This is very interesting topic. Just one question - where is Amy?
@@thecasualengineer99 Aaah, one of the mysteries of the universe.... actually this video was used as a lecture at the Microwave Update conference in Vancouver and I wasn't sure if the audience was ready for a snarky digital avatar.
@@MachiningandMicrowaves ps - i noticed you have a Cutting Edge Engineering cup behind you, thanks for supporting Kurtis' channel
Great to see you back and sorry to hear about your accident, sounds like a bit of a nightmare. I probably only understand about 60% of what you're talking about but nonetheless, I do enjoy your videos.
Well, that's a good 10% more than I understand, so you're well ahead!
Use the milling machine as a shaper. The spindle can hold a fairly substantial triangular shaped bit with a tiny corner radius to do the final surface finish on the Fresnel lenses. This could be used to make a negative mandrel out of lacquer or wax which is then plated similarly to how vinyl records used to be made. The feature size of a vinyl record is in the micron range and the technology is well developed.
That's a good thought. If I can work out how to rotate the spindle to the correct angle, I could make a rotary shaper. Simpler if I had a 4th axis of course. The alternative is to get on with converting my Colchester lathe to CNC, then I can do all sorts of clever stuff.
Fascinating and clever! I particularly liked the idea of machining a sacrificial armature from aluminum, electroplating and then etching away the aluminum. Possibly too many problems with material uniformity as you mentioned, but I loved the out of the box thinking👍😁
Satellite manufacturers and folks making high vacuum devices can electroform stuff in their hundred million dollar labs, so why shouldn't I be able to do it in my zero-budget Secret Underground Mountain Lair? (codename "SHED")
Wish you a speedy recovery internet stranger! Hopoing to see you all healed and hopefully back soon. Take care
You have inspired me to play with RF.
No mention of a Phased Array antenna. They don’t have to be as complicated as Starlink’s Dishy McFlatface. Having said that, there is plenty here to keep us interested.
Reflectarrays are definitely in my future though. Phased arrays get very VERY expensive when they are a useful size, and you really need to be doing chip-bonding to get decent results above 10 GHz. I'm working on a mmWave imaging system using mechanical raster scanning, that should be amusing. The application I'm thinking of for phased arrays is for rainscatter and aircraft scatter from multiple fast-moving targets, but the disappearance of semiconductor power devices from the market over the last five years means I'm stuck with costs of at least $100 per active device, and I'd need at least 6x6 to make it worth the effort. That'll mean taking on a sponsor to help fund the project unless I win the lottery.
Regarding melting swarf -- if you're trying to induction heat it directly, no such luck, it's porous, hardly a path for current to flow between bits. Needs much higher frequency (low MHz?), or a starting "heel" to get it going (maybe just a plate or bar, but mind that it'll stand up or push out of the way if you're doing low frequencies or some kW of power). Or use a susceptor -- a graphite crucible is excellent here, and hardly burns at alumin(i)um temperatures (loss is noticeable at Cu temps however).
Or possibly, swarf could be pressed into a "cake", by sufficient hammering or hydraulic pressure, to heat directly.
As for cleanliness -- aluminum doesn't melt worth a damn into itself due to the surface oxide; you need a flux to get the crud to separate. This can be as simple as ordinary table salt (preferably eutectic NaCl-KCl, or NaCl-CaCl2 for lower MP, either way about half-and-half mix; mind, that must be anhydrous CaCl2!!). Best is to use just enough salt (flux) to loosen the oxide, making it easily removed, and coalescing trapped metal into beads that stir back into the bulk.
Downside is, salt normally absorbs some moisture, which reacts with the metal, forming more oxide (big whoop, we're washing that away anyway) and hydrogen. Hydrogen is soluble in molten metal, but crashes out as bubbles on freezing, just like the CO2 when freezing a soda! Inert dry gas (N2, Ar) can be bubbled through to coax it out (sparging), or a reactive gas like Cl2 used to burn it out (usually by putting a compound like Ca(OCl)2 on a dosing wand and thrusting that into the melt).
I'm pressing it into pellets using my 14 ton hydraulic press and a steel mould, but really not getting any worthwhile melting with the usual cheap far eastern toy induction heaters. I guess I need more compaction. I considered trying 1.5 kW at 13.56 MHz to see how that behaved, but I don't want to risk damaging my amplifier, and I have a large bin of chips to melt. Perhaps best just to send it off for recycling by the local folks who have an inert gas furnace. Perhaps I should try running a microwave oven with a graphite-lined ceramic pot and purge the oven with argon after plugging the vents and disabling the fan. Recycling is far simpler, and I might even get enough money from that for a packet of biscuits or a few coffees.
Nice to see another "Hoof GP" and "QRP labs" aficionado! 73 - Dino KLØS
And Cutting Edge Engineering Australia of course!
@@MachiningandMicrowaves Good taste in music too, UK Grim is a great album.
@@samthenerf My family come from Sleaford...
Simply amazing! Great work and looking forward to more. Gets me thinking about the DIY Perks and Shake The Future channels there on the other side of the pond. DIY Perks in regards to his nickel plating method he used with the Building the world's first breathing PC, little after ten minutes time, where I have been wondering about the effect of using the speaker and optimal frequency to potentially improve NiMH batteries performance when reconditioning modules cells so maybe some novel optimal battery charging method not written about or at least not seen by my eyes can be found. In regards to Shake The Future YT channel, IMHO he has been doing some great work with casting metals using microwave ovens. Neat methods where I have not read anywhere else as of yet detail such, though when I find the time will search for papers regarding. All the best!
@@jafinch78 Sounds like I need to do some research. The sort of research involving watching UA-cam channels. That's the best sort!
very interesting
glad to see you back on YT. I've missed the funny video comments. Hope you are well soon
Very nice video sir
you could cnc grind the parts, much slower , much more time, but very accurate
You could try rotating the part as you are electroplating it, maybe with a slight tilt or entirely horizontal for some gravity assistance. That should help clear out the debris and also help keep the plating very even as you have averaged out the inevitable variations in plating thickness.
@@foldionepapyrus3441 I'm using a rotary mount ensures the mandrel is rotated as well as getting a flow of fresh electrolyte but I think the entire problem is down to reducing the electric field concentration around the sharp edges at the top of each of the grooves which leads to a great deal, more deposition there than deep in the grooves. I know it's been done successfully in satellite antenna manufacturing and in university labs so I think it's just a question of slogging away and changing things until I get the recipe just right. It's basically a mixture of bucket chemistry and cake making.
@@MachiningandMicrowaves Yeah, almost all these processes are half artform to get just right.
And by total coincidence i piked up a satalite dish
Only a small one but its a fun start its comes in a suitcase and has a vacum cup mount
Or a tube mount
Very interesting. You may have addressed this and I missed it, but could you make that Fresnel reflector out of plastic and coat the surface with a radio-reflective paint or other material?
That's something I've tried and failed at, but I think the trick is either to use a metallized foil and stretch it very slightly, then put a slightly porous plastic form beneath it, coating it first with a glue film, then apply a vacuum to get the foil to adhere closely to the contours. For stepped shapes, it will be tricky, put for a large printed subreflector for example the print could have a VERY sparse infill and be very light and stiff. Using a ductile metal sheet that would stretch enough to conform to a paraboloid form would avoid the problems of creasing. It's probably feasible to use a conductive paint, then electroplate over it, but I haven't tried that yet. For more intricate devices, my Cunning Plan involves a two-stage Edwards vacuum pump followed by a turbomolecular pump and a magnetron source using some huge ring magnets so I can deposit a metallic film on the target, then electroplate with copper and maybe finish with a layer of gold. That's for next year when I can find a better vacuum chamber than my bell jar
@@MachiningandMicrowaves Well, it sounds like you have already thought through all this. I was thinking more of a sputtering technique to deposit a reflective surface. By the way, have you encountered the UA-cam channel Applied Science (www.youtube.com/@AppliedScience)? You two seem like kindred spirits. Best wishes!!
@KevinToppenberg Could you use electroless nickel? Then electro plate the thickness you need ? Remember when plating the "robber" configuration is very important to get an even thickness.
@@BenRasmussen-c3u Electroless copper can work but the adhesion is never really reliable enough. I've not tried nickel other than for decorative plating, because it's a terrible material at microwaves, being ferromagnetic and very lossy. If you laid down at least 1.2 micrometres of copper/silver/gold over it, that would work at 122 GHz, as after 6 skin depths, the current is negligible. At lower frequencies like 10 GHz, you'd need at least 4 micrometres.
Bonding UHMW
Use two-part epoxy or neoprene-based cement.
Clean both surfaces well, using a cleaner such as alcohol to eliminate even fingerprint oil.
Roughen both edges that will be bonded - the side of the plastic and the side of whatever you are going to adhere to it using 4-0 sandpaper.
Oxidize the surface of the plastic with an open flame, such as a hand-held torch, using just a couple of fast passes. Do not allow the material to blister.
Apply adhesive to both surfaces, and allow to dry until it becomes tacky.
Press the pieces together and clamp tightly. Allow it to sit overnight
Someone advised me about flame-treatment to activate the surface. I have a 65 kW propane torch! So long as the glue line is negligible and the material isn't highly polar or otherwise lossy, that sounds straightforward enough. I can of course use woodworking join techniques, wedges and dowels made from the same material to help hold it together. I need to find some slightly better cutters, perhaps single-flute, at least for the edges and for roughing, and I need to experiment with chip clearance using air blast or even coolant. Great info, thanks
For conical small shapes wire edm might be an interesting rabbithole to dive in, and the grooves might be doable too with some creative variation of edm (the machining kind, not the music, Google gets really confused if you ask it for edm grooves)
Skrillex, Deadmau5, seen them all.... Wire EDM is 100% on my roadmap for making fancy internal shapes like rectangular waveguides, but die-sinking using some sort of rotary vaned tool would be a mighty challenge with 0.5 mm grooves in graphite electrodes. There's probably a magic technique out there somewhere!
@@MachiningandMicrowaves having independent control over the top and bottom end of the wire also allows for all kinds of conical shapes so the horn without ridges should be easy. Getting the ridges in is the tricky part indeed!
@@rjordans The ultimate performance comes from having corrugations but also having a non-conical cross-section so the surface follows an elliptical curve or something similar. Another variation is where the axial cross-section is non-circular, probably elliptic. The key parameters are getting the sidelobe level below -40 dB and keeping the axial ratio as close to unity as possible. That takes a lot of computation and some very expensive software, so I'll be limited to the tools I can get low cost or free, but I'll be able to make lots of finished devices to test simply by parametric variation and wasting a lot of my time. Usually though when the CNC machine is running, I can be doing something else on the manual machines
Yes, yes, that’s all well and good, but what have you done with Amy?
Seriously though, I’m happy to see you’ve healed enough to make videos again. It’s very rare that a wizard explains their wizardry in a way that mere mortals like myself can follow along.
Ah, this vid started life as a lecture for Microwave Update 2024 in Vancouver. For complicated airline reasons I failed to get there, so I created a couple of videos to present remotely instead. I didn't think the esteemed audience would be ready for a snarky imaginary avatar, so AIMEE was granted a temporary leave of absence. It is HIGHLY likely that I may be blessed with the Return of the Snark for the next video.
Hi Neil, I've missed your videos. Best regards, Neil (we have a naming clash).
If you are looking for higher resolution 3d printing there are some larger format resin printers that, provided your material properties can be met, would proide high resolution/precision. It would still need to be printed in sections though likely.
@@patrickcallahan2210 I made a video about Rogers RADIX material and Fortify printers a while back. Unveiling the Future of Antennas and RF Lenses using Radix™ 3D printable material!
ua-cam.com/video/3YMRfw0uWlw/v-deo.html
Sorry to hear about your accident. Hope you make a full recovery soon. As for the video... I heard 'cake' Is there cake? :D
There might be cake
@@MachiningandMicrowaves Mmm caaakkeee...
The description of your accident made me shudder. I really hope you fully recover soon. The important question is did you get the septic tank properly cleaned in the end and who was the poor soul who finished the job?
I'd finished almost the whole job when it happened. I'd drilled a damn great hole through the P-trap siphon in one of the feeds to the tank, and put a load of Drano down the hole to try to clear the trap. That's what the boiling water was for. I lost my sense of smell a couple of decades ago, so sewage, rotten fish, roses and fresh-mown hay all smell of nothing. I can smell some small molecules, like lemons, coffee and the coarse unpleasant notes in cheap perfume. Sewage is just odourless sludge to me!
What kind of brightener are you using for your copper plating? PCB manufacturers are able to plate fairly high aspect ratio holes. You might try looking at different brightening solutions.
@@danielmatthews8475 That's definitely one of the variables that I'm experimenting with. So far I haven't tried varying the pH or concentration or make up of the electrolyte, but that's a nice easy thing to vary.
Maybe using argon atmosphere for melting chips would help? Also "Casting Crucible With Stopper Rod" can help.
@@staryduren I need to try some tests, but I think I might just ask the local metal recyclers if they can process aluminium chips
I’ve often wondered if a ‘figured’ convex lapping plate would enable mass production of pressed out/spun and stress relieved dishes.
The traditional spun dishes were made on a mandrel that allowed for spring-back, and it was a black art getting the figure just right. It's simple enough to make moulds for a carbon fibre dish, and then metallize the surface, but maybe combining a pressed thin metal sheet glued to a moulded rear support would be a useful solution for home-shop production. The Voyager spacecraft used an uncoated carbon fibre dish to save weight. A dielectric surface reflects a reasonable proportion of an incident wave. I like the mechanical complexity of a reflectarray though.
Wow Neil......I don't think we want to get into your personal affairs, but we get to grow fond of UA-camrs we follow and I was quite worried as some others I am sure.
Apart from my little accident, which meant I was unable to type, write or hold a screwdriver, and kept dropping soldering irons on my foot for several months, I was working 60+ hours a week on the BBC TV project with Professor Hannah Fry. That turned into a mammoth effort, machining over 250 parts and filling the Bin of Shame with a lot of failure, and just consumed my entire life. It was ENORMOUS fun though, and should result in a series of videos next Spring after the program is broadcast on BBC2 TV
I know very little about radio, but your list of techniques makes me wonder which combination of them will get you the result you are looking for. I am curious what are the limits for internal grinding. That should give very good surface finish. As a random thought, it might be worth getting in touch with some of the equipment manufacturers, CNC grinding equipment for example, in the hope that a new market would be appealing to them and a challenge is a good thing as who knows what the market could want next.
Grinding non-ferrous materials is not at all easy. I have been thinking about electrochemical machining as a potential technique, but I'd need to machine some titanium electrodes with very fine grooves. I've used internal grinding down to about 10 mm, never tried it on smaller bores, and only in carbon steels and tool steel.
Fascinating, have you considered the resin 3d printer tech for this - smaller feature size than possible of FDM and at least in theory you can blend in additives yourself to tune the material for RF work.
Also on the subject of practical for the home machine shop you can argue nothing and everything is - to make anything within the limits of your shed/garage and relative shoestring budget always requires plenty of effort, and ingenuity - so even the simple stuff will be much harder for you to make than send off to the professional. But you can do it all, even to the point of making functional silicon chips if you want to put in that effort. Its never about practical, its about if its worth the cost in time and money to you - which then folds back into the fun value.
@@foldionepapyrus3441 the problem with resins is that they're almost all horribly absorbent at microwave frequencies. I did a video awhile back about Rogers radix material which is used on Fortify resin printers to make ceramic filled resin antenna lenses. It's fairly expensive and needs heating agitation and regular wipes to prevent buildup of the filler. I'm to ensure even coverage. Whether any of the resins which I could print on an Elegoo or something similar have anything approaching a reasonable loss tangent needs further investigation as a couple of years of passed. I suspect that none of the suppliers publish the relative primitivity or lost tangent of their resins at micro frequencies, so I'd have to print some test coupons and measure the numbers for myself. Printing using a resin containing a high mass fraction of ceramic, dog bones or nodules would need some careful modifications to ensure that the right amount of stirring happened to keep the material homogeneous. I've certainly enjoyed watching the folks making their own chips in their home labs. I used to work in a semiconductor factory in the early 1970s making epitaxial planar transistors and microwave varactor diodes, so it brings back nice memories of a couple of summer vacations spent in silicon fab.
@@MachiningandMicrowaves I wouldn't expect most resins to tell you RF performance, considering how few of them even tell you some of the more commonly needed mechanical/chemical stuff. But there is a wide range of resins, so as long as its not the BPA element that almost universal across them all (though less so than it used to be from my research)...
@@foldionepapyrus3441 if I can get the mass fraction high enough using a high-performance ceramic infill then the amount of resin to cause loss should be reduced. The problem is some of the resin materials are highly polar and have a lost tangent several hundred times worse than a good diet material. It certainly worth doing some research, particularly for relatively thin lens structures where machining from solid materials like Rexolite 1422 isn't feasible
could you use as cast ceramic and then electroplate it????
@@BenRasmussen-c3u Any sort of casting, either chemical setting or sintering or molten material, has problems with warpage and shrinkage. If you can control warpage and allow for the shrink when making the mould, it can work. You can use PVD or other vacuum deposition methods or even laser sintering to get conductive layers on to ceramics.
awesome!
oh man, get well soon!
Missed you, glad to hear you are on the mend.
FYI - there's a US (I think) company, called "Servometer" that does electroplated nickel objects on sacrificial aluminium mandrels, exactly as you describe. Their primary material is nickel, as they use the shapes as "bellows" for various purposes. Fantastic stuff. Perhaps try contacting them?
@@AdityaMehendale that's a good point, I've seen companies that produce what appears to be copper Bellows using electroforming and I think that's the techniques that was probably used to make vacuum variable capacitors but the only place where I've seen the zincate process with gold used is in a paper by one of the US universities. There are several people making vacuum Bellows but they all have the benefit of very large radius curves so they won't have the same issues of high field gradient and concentrations at the sharp corners. I'll make some enquiries.
Sounds like a good excuse to get a CNC lathe.
@@lucaswilkins9217 Absolutely! I have written a post processor for fusion which drives the SYIL as though it was a Fanuc lathe, with the workpiece in the spindle and multiple lathe tools mounted in the vice. I do have a 750 W servo motor with 17 bit encoder that I could possibly press into service on the cross slide of my Colchester lathe, But as ever time is the limiting factor
What do you use microwave for? Honestly for hobbyist there doesn't seem to much in the way of actual utility. At least with hf you can communicate around the world.
@@Imaboss8ball Moonbounce, which is available any time the moon is visible here and at another location in another continent. Aircraft scatter, rainscatter (and snow and hail) out to maybe 800km, much further during tropospheric ducting events, digital TV transmissions, but also reflection from geology or built structures, pushing for distance records on the bands where oxygen and water vapour absorb signals. Also bistatic radar and fmcw chirp mmWave imaging. The key driver for me though is that by making high-performance equipment for other folks means they can have a lot of extra fun. As my late wife used to say about the traditional amateur radio bands when I told her of some fantastic contact I'd made, she would ask "did they have anything interesting to say?". also after 40 years of playing on the bands from 136 kilohertz through to UHF, and making contacts in over 300 countries on 1.8 MHz, I got a little burned out and decided to go back to where I started back in 1973, on the microwave bands. It's kept me engaged and then thrilled for the whole of my adult life. When I discuss how I play on the radio with other radio, I find we have almost no overlap in the sort of activities we enjoy and that's a good thing.
I wonder whether the reason for problems with copper electroplating is not that you have some sort of shielding effect at the bottom of those high aspect ratio trenches in your corrugated horn.
Yes, that's pretty much it. The electric field concentrates at the sharp corners, so there isn't much of a potential gradient in the bottom corners, so the electrolysis doesn't throw sufficient metal down into those areas. I suspect it might be possible to improve matters by reducing the concentration of the electrolyte and increasing the amount of turbulence by using an impeller rather than a bubbler to agitate the solution. I really haven't done enough experimentation yet so watch this space for new results
Neil, have you tried 3D printing an antenna slightly oversize and then "trimming" it to size with your CNC mill? UHMW is the acronym you were looking for. ;) With electroforming, I *assume* that the electrical flux(?) lines are like electroplating: the "flux lines" (if that's the correct term) must meet the surface of the part being plated perpendicularly, so outer sharp corners get heavier plating and sharp inner corners are extremely difficult to plate -- could that be part of your troubles when electroforming?
Also, some online companies like PCBWay have added 3D printing & CNC services (none of which I've ever used, but have seen some of their outputs on other UA-cam channels). I *think* they can produce the resin-printed parts. Other than that, I have *zero* 3D printing experience, so this is simply my best SWAG. Assuming that resin-printed parts can have finer details than the extruded filament parts can, you might get closer to what you need. Of course, this means that you'd have to farm out that portion of making your antennas.
Good luck and I'm waiting to see how you solve these particular problems!
Electroforming into deep narrow slots it's just playing difficult because as you say there are concentrations of field lines at the sharp top edges of the slots and getting any sort of field gradient down in the bottom of the slot is difficult.
Low voltages and low current densities help and there is definitely a sweet spot, it's just that I haven't found it yet. Machining 3-D prints is difficult as the layers tend to eliminate and the materials can be brittle using a abrasives and painting the surface and then fling it down is certainly a viable approach for mouldmaking finding materials that are excellent Dielectrics that are also viable for 3-D printing is tricky. metal 3-D prints are still extremely expensive and subject to complex shrinkage which you have to design out. Metallising 3-D printed surfaces reliably is getting easier but I would be concerned about whether the metal layer would continue to adhere to the printed material after temperature cycling and variations of humidity and physical flexing. I'm sure all these issues are fixable. The use of advanced materials in resin printers is certainly maturing rapidly but the costs of suitable machines and high-performance electric resins is still very much in the military/academic budget zone rather than home hobbyist
There's lot of knowledge available in the realm of PCB manufacturing when it comes to surfactants, additives to control deposition (mostly electroless) in holes and crevices, as well as current waveforms for electroplating to get an even copper deposition. Not my field of expertise so I hope someone else can chime in.
One of my Discord members has been working on PTH in his home lab with interesting results. I get pestered all the time to do sponsored vids by various PCB shops, but haven't taken them up. I'm following the technique used at a US University to get the zincate process tuned to perfection, but getting the slots deep enough is very challenging as the top corners of the slots tend to build up plating rapidly and eventually leave a hollow with electrolyte inside. The trick seems to involve cycles of electropolishing to remove dendrites that act as concentration points for deposition. The problem is certainly similar to plating a long, narrow PTH, but filling it completely with plating. I need an aspect ratio of about 2.5:1, perhaps 3:1, and that's where things go wrong. Bubbles don't seem to help, but I'm trying a pumped technique with a shaped electrode and swirling the electrolyte with a magnetically-linked pump
So, sort of a fresnel lens type dish?
@@Jemmet I'm not sure if Fresnel would recognise it as such. A proper Fresnel
lens uses internal reflections as well as refraction. That's probably a good generic name for it, but it could also be thought of as a sort of reflectarray perhaps?
First time viewing your channel -- very interesting content AND comments.
Could your part be split into 3 or 4 identical pieces, and then just machine them on your Syil VMC using an Air Turbine (brand) air spindle and a very tiny end mill? (then assemble parts back into the full conical part).
Do you have a 4th axis on your Syil? If you do, then all the movement could be in the A axis using an end mill the width of the groove.
Though... the results would be even better stepping (actually: positioning) down and over with a tiny end mill while still doing all the actual cutting movements back and forth with the rotational A axis (would reduce the "washout" caused by the end of an end mill that is the width of the groove).
I don't have the 4th axis drive - yet. I'd like a Renishaw-style auto probe first, but my financial resources are very limited! The 601BT30 air spindle looks like a perfect solution if I can find about £6k or win the lottery.
@@MachiningandMicrowaves They are a very reliable and accurate spindle... I used one in a previous facility where I worked, and never had a single issue with it.
How many rpm on your Syil?
I can do 18k sustained on the X5 3.7 kW spindle. It's limited to 18k, but rated for 20k. I'm looking at coolant-driven spindles, mostly because they might be less noisy than air spindles. 75k rpm looks easy but I think it only works with through-spindle coolant, which mine doesn't have. Maybe I can find a way to pipe coolant to the unit through the side of the BT30 holder, but I can't see how I could get it to work with the autochanger, and the X5 doesn't seem to have a manual change feature like some of the others. I'd have to mount it in the changer, then do a manual change and connect the coolant hose then continue the program, or just do a totally separate operation for the high speed spindle. Very clunky. You have to change the bearings and turbine impeller every time you change the tool in these low-cost spindles, but I'd only be using a single tool. I need to talk to the supplier. They suggested a starting price of less than £1k, which seems affordable if it is quiet and can run a 0.5 mm tool at 60k+ rpm on 8 bar coolant pressure. Hmmm, wonder if someone makes a through-coolant BT30 to BT30 adaptor?
Eh practicality is overrated. Glad to have you back!
Burns like that are no fun at all, and I can imagine how gnarly that was.
The specialist burns nurse looked at it, and said "that dead layer is going to have to come off". It was indeed very grey. She came at me with scalpel and tweezers and a determined look. I was expecting anaesthetic and needles, but she just said "this won't hurt a bit". She was right, I never felt a thing. Fascinating to watch someone slicing into you but feeling nothing! The new skin is beautiful and smooth and soft. I could be a hand model now, unless they wanted both hands in shot.
Aww, the Nokia FlexiHopper guts have disappeared from the shelf.
Gone on to greater things. Possibly it will reappear in an upcoming project!
and you could get cutters ground with a radius on the end with 2 or 4 flutes . I know someone in California that can make them for you
@@BenRasmussen-c3u I can get custom carbide cutters from Drill Service of Horley here in the UK as well. They can do any form I want, at a price. Good to know there are alternative suppliers out there
I was looking at using a parabolic reflector to concentrate sun light , your comment about light having many frequencies, really caught my attention, thanks for the insight
How about 3D printing a mold and then fill it with Silicone with Titanium dioxide added to it to increase it's relative permittivity. Or maybe even barium titanate.
Maybe you can also get close to a dielectric constant of 10 or even more.
@@GermanMythbuster Silicone is a very poor dielectric, so it would create an excellent absorber. However, maybe using a granular fill of alumina or TiO2 in a cellular print might just work! Interesting idea, thanks!
@@MachiningandMicrowaves Yeah I though of the Silicone more as a binder. But you are right why not directly fill a print with a filler, brilliant!
Neil, What about CVD or MCVD?
Definitely in my future plans. I have a nice roughing pump and a turbomolecular pump with a faulty controller, so I should be able to use a variety of vacuum deposition techniques. I guess I'll need to use a magnet-coupled spindle to rotate the target. I only have glass bell jars at the moment, I might try machining and welding up a chamber from stainless. So many projects, so little time!
@@MachiningandMicrowaves afaik getting high-vacuum quality welds is a pretty exacting process, depending on your motivations it might be worth just buying a chamber instead of dealing with the pain of that. Elasatomer-sealed rotary feedthroughs are also available for significantly cheaper than magnet-coupled ones, though I'm not sure what the tradeoffs are (possibly speed or torque limits?)
@@L1ama I don't need really high vacuum for PVD/CVD, I might get away with home made, but I'd like a proper chamber to do more interesting experiments. I keep watching the University Labs sales....
I was also going to suggest vacuum deposition. Just something I have always wanted to play with, for no other reason than I learned it existed, and it looks really cool.
@@justanum that's a good enough reason for me!
Isn't lens combinations viable with VHDPE refractors? Splitting the collimation power into two by adding in a lens that sits halfway between the main lens (the bigger front lens) and the radiating element should almost halve the thickness of the main lens. And the secondary "only" needs to be 2/3 the diameter of the main, and also only need about half the relative profile height... Should reduce the necessary base material weight (and machining time!) by a factor of almost two.
I only know nano- to micrometer EM optics (normal optical lenses), so this might have other problems? Always found antennas interesting though
The main constraint is that the Rayleigh distance from the source might limit how close the first lens could be, however it would be feasible to make that lens completely aspheric to suit the position, almost using it like a corrector plate. That certainly gives me some ideas about potential future experiments. The only other downside is that there is a potential reflection from each of the surfaces that the RF energy crosses. that usually requires a quarter wave plate or a set of quarter wave grooves in at least one surface of the lens. Easy enough to deal with when you're printing the lens, but also straightforward when you're making a lens from solid plate. The depth of the grooves is a quarter wave in the dielectric so not as deep as you would expect. another approach is to use a machine lens and an auxiliary 3-D printed lens made of high relative permittivity material. So many potential projects so little time!
@@MachiningandMicrowaves I have no idea what the radiating aperture in an mm-wave antenna might be, so I see your point :) -but combined with a waveguide, the projected (virtual) radiation aperture should get very small? decreasing the Rayleigh distance? Great work, keep it up.
I'm still at the WiFi can-tenna level for 2.4 and 5 GHz :)
@@AJBtheSuede For the Cassegrain antenna behind my head in the first minute or so of the video, at 122 GHz, the subreflector is at just a little more than the Rayleigh distance, but that's a VERY short focal length dish. On a more normal dish, it wouldn't be an issue
Have a look at Innegra fiber (white) laminate for radomes
The problem is always finding a resin that's low-loss enough. Blow-moulded material can be good, but if the laminate mesh is good, even polyester resins might work. I'll have a look at their products, thanks for the tip
@@MachiningandMicrowaves in their web page in the "resources" tab and then at the bottom of the page they have an application data note with electrical properties for radomes. The only drawback I see is that these poliolefin fibers can be a bit difficult to wet at first but it can be done. Vacuum bagging and oven or autoclave should delver best results.
@@miguelangelsimonfernandez5498 I'll take a look, thanks
are you 'ironing' your 3d prints to smooth them
I do on flat surfaces sometimes, but the finish is usually good enough for the frequencies in use. At 10 GHz, 1mm steps are almost invisible. I don't know if ironing would help with the layer steps on shallow curves, any idea if it works well?
@@MachiningandMicrowaves ive only done it once with great results on shallow surfaces, maybe a m6 tooling ball to produce large cusp raduis?ive never worked with bamboo slicer so idk the software side
I tried using a solvent on some of the HIPS material, but that went badly! Sanding with wet/dry paper under flowing water worked well, although the main reason I did it was to remove the rough bits along the joint line using polystyrene cement. I'll have a look at the Bambu slicer and see what it can do with the fancy dielectric materials
@@MachiningandMicrowaves i wonder if a tapered thread could be used as a substitute of the baffles of your horn, using a dremmel cutting disc to make the deep cuts ,the burrs pollished off with a cone of the opposite side
i don't understand radio stuff well what im saying could be nonsense
Hi Neil. If you ever need a sub 40 GHz antenna evaluated (and 3D patterned), I can do that for you in my anechoic chamber. Just let me know. /Glenn
@@BAR_220 I'll definitely bear that in mind for a future project, many thanks Glenn
Nah, it's magic!!
I recognize the chap spinning that reflector, I use to work there. 😊 It was at Elite Antennas in Leominster. Interesting video.
Why do you build antennas
It's a habit. I started in 1972 and can't kick the addiction. Also I sell them so I can buy materials and machines so I can make more antennas. Mostly it's because amateur experimenters often need parts which are not commercially available or parts to repair systems which were discontinued decades ago. I like helping people to bounce radio signals off the moon or to build their own radio telescope systems. Mostly though because I enjoy doing it.
You can glue polyethylene if you briefly run a flame over the surface (flame treatment) not enough to burn or melt it, just enough to functionalize the surface and make the glue stick. It is workable, but limited to relatively simple geometries that the flame can touch
it should be relatively straightforward because typically those edges will be radial slices and will have some thickness but in addition I could glue another supporting sheet behind the main lens and possibly incorporate ribs to stiffen it without too much impact on the electric performance.
Using dovetail joints or dowels of the same material that the bodies made from may be just as easy but there are issues with moisture ingress and if I could use glue instead it would certainly improve the long-term performance
@@MachiningandMicrowavesit's relatively tolerant, each part of the surface needs to be exposed to the flame somewhere between 0.3-2 seconds. Glue up needs to be relatively soon after treatment, it tends to revert to the non sticky state within some tens of minutes
Right, I'm on it!
@@MachiningandMicrowavesplasma, UV, and oxidizers like ozone also work and have advantages, but fire is the most easily accessible :)
What does AIMEE think of all this? Did she get hurt in the accident?
@@ke9tv oh, I've been rumbled! The original purpose of this video was the basis of a lecture which I was due to give at Microwave Update 2024 in Vancouver. I wasn't sure that the distinguished audience was quite ready for AIMEE, who was temporarily consigned to the bit-bucket for the duration.
I feel like you are trying to simplify the information to be more approachable. I do wonder though if you use terms like "person who discovered it" effect would end up leading me down researching those effects. At least till I ran out of tea, coffee or energy drinks for the new age back yard mad scientist
@@daretodreamtofly3288 my excuse is that the target audience was a lot of non-UA-camrs at microwave update in Vancouver recently. I was supposed to be there presenting in person but things got messed up so I didn't go. There will be much bigger and denser pieces about some of the things that I discussed. The second version of this is because I've been away for so long I thought I needed to give an introduction to the sorts of things I'm messing about with so that people can unsubscribe!
And that's why 10s of MHz are better. Pass the tape measure and hacksaw. 😂
@@kimbledunster my other interest is in radio frequencies around 136 kilohertz, where the required precision is measured in tens of feet or thereabouts. also 8270 Hertz, where even getting out of the near field is tough!