Machining Fixtures, Spacers and custom Collets for 1.3 GHz Coaxial Filters pt 3

[@tamiamibusch]

Someone else that does machining and RF stuff? There can only be one! Draw your VNA and let the best one survive. Look forward to your next video.

[@bandana_girl6507]

Gotta love when Brits are extremely proud of "not using Imperial". What's a "stone" and why do you worry about gaining or losing one?

[@MachiningandMicrowaves]

I'm a great fan of bushels. en.wikipedia.org/wiki/Bushel#/media/File:Bushel_Table_of_States.jpg and acre-feet. Working in MKS, CGS, SI, UK Imperial, Troy and Avoirdupois probably helped me to become unit-agnostic over the decades. My shoes are dual-marked using Barleycorns as a linear measure. A stone is of course the weight on Earth at sea level of six and a half UK bags of sugar.

[@cal28kim]

I understood Skrillex🤣🤣! See sonny John your not forgotten 👍.

[@dumpy4289]

"for those on monochrome sets...." hahaha cheers!

[@bobbiac]

"yadda yadda" battery died... Have you tried a Milwaukee/DeWalt DC barrel adapter? Works surprisingly well. 12 or 5 VDC

[@MachiningandMicrowaves]

If the camera could be run properly with a dummy battery, I could use that perfectly well, but you can't fit the camera on a tripod with the dummy battery installed because the battery compartment door can't close because of the cable. I don't want to throw any more money at this horrible unit by buying a spare Sony battery. I feel like I should just go buy a proper camera that is actually designed to be usable like my Nikons. This is literally the second-worst camera I've ever owned. All the reviews said how great it was. All those reviewers are now firmly in the "idiot" pile. I have a set of DC connectors for my Milwaukee 18V tools that I use with DC-DC converters to power lots of other stuff out in the field, but in the machine shop there's plenty of power everywhere. I should just go and throw a couple of months salary at a Nikon Z6 II kit and be done with it. If I get the adaptor, I can even use my existing Nikon lenses. I can probably use this thing as a second-angle camera. Or an anvil or door-stop. Actually the back is fairly flat, so maybe it could be used as a sanding block for rough timber.

[@smash5967]

Where do I get me one of those expanding twist drills like you used at 10:00? That looks super handy. Also, you do this every video and you'd think I'd know better by now, but when you zoom way in on your first cut I'm always waiting for the tool to break or the part to fly out of the lathe or something.

[@smash5967]

Also, re: the carefully set up shot at about 13:00, if your camera was on the carriage and the cutting edge of the tool was in focus you could move the shot with the lathe. Just don't try mounting your camera on the spindle, the best case there is getting really dizzy.

[@MachiningandMicrowaves]

The camera loses the plot completely if I mount it anywhere on the lathe, even the tiniest vibration upsets it to the point it just goes blurry. It is a rubbish camera. Might work OK on a gimbal though, I might give that a try. I'm going to fit overhead rails and a traversing mount to I can do fancy shots and mount the lights so I don't have to reach around the stands. Hmmm, a Go-Pro mounted on the chuck while doing slow-speed drilling with a giant drill could be amusing....

[@smash5967]

Also also, is there a reason for you order of operations on the aluminum plate? Why not drill while it's still square and in the mill? That way you don't have to reset the DRO and the vise has a lot more meat to hang onto. Also, you can often save time when making a square thing round by knocking the corners off with a saw. It's hard to beat the material removal rate of not turning your waste completely into chips.

[@MachiningandMicrowaves]

Building tension... For weeks... One of these times there really will be a huge explosion of a carbide end mill or something when everyone is least expecting it. It's such a shame I didn't catch that damn great long-series 3/4 inch 4-flute HSS mill when it let go. That was loud.

[@MachiningandMicrowaves]

@@smash5967 My bandsaw has an earth fault, which is annoying and means I can't use it. I did consider using a hand hacksaw, but then decided it might be funnier to chip the corners off on the old Colchester. I didn't really think through what I was doing about the base, and it was shot out of sequence, with a rethink of the design, so the later parts were actually shot earlier. Confused? You will be! (was "Soap" really that long ago?). As ever, these vids are "How I did XXX" and definitely not "This is how to do XXX". Probably in most cases, "How NOT to do XXX" would be a good description as I stumble about having not even a vague clue what I'm doing, but having a pile of fun doing whatever it is as incorrectly as possible

[@vladimirsvirid7705]

waiting

[@soranuareane]

Trellus? Trellus is the name of the big bad evil character person in the novel I'm writing. Don't tell me that's a real name.

[@MachiningandMicrowaves]

You're safe. The mythical Mrs Trellis is not going to infringe on your baddie's namespace.

[@smallcnclathes]

You do not appear to use machine grade aluminium any reason? Could you show us how you do the moving exploded views and moving assemblies in your cad system. I want to see if I can make mine do it.

[@MachiningandMicrowaves]

These spacers are only going to be used maybe ten times and I had a 25 mm diameter bar end of cheap 6082T6 in the metal store. I do use a bit of short-chipping/free-machining alloy when I need to machine something important, and sometimes use fancier 7xxx grades, but I just grabbed this bit of bar off the rack. I'm using Fusion360, just using the animation tab and creating a timeline. I use manual explodes, although auto-explode does work sometimes. It also does decent rotations and joints. Process is just create the whole assembly as individual components, then select all the components I want in the animation. Into the animation tab, I set the initial view and toggle visibility of the components, then start zooming, exploding, rotating, fiddling with the camera view and all that. It's not exactly intuitive, but once you understand how Fusion thinks, it's just about useable. I should really go all-in and script it in Blender. I used to be quite nifty with that toolset. I saw some brilliant work done with Blender last week and I think I really must get on with re-learning it for animation.

[@smallcnclathes]

@@MachiningandMicrowaves I just find animations in my package too hard to get my head around, too many constraints to get right. As for Fusion, I can’t use it, end of story. I can do additive assemblies by simply turning on layers while running screen video and editing out the on screen antics that makes it happen. I might have to have another crack at animation, should only need sliding constraints to make it work.

[@smallcnclathes]

@@MachiningandMicrowaves well you got me to try animation in my cad this morning. I am working on steam whistle and I managed to get linear constraints to assemble it from the exploded view. So well done you, inspired me to do something I have wanted to do for years

[@Xsiondu]

I'm sorry but I'm a bit lost. What have I stumbled upon?

[@MachiningandMicrowaves]

This is part of a project where I made a low-pass coaxial filter to suppress harmonic outputs from a high powered amplifier. The science behind it is interesting. I need to do a vid about that. First though, I need to do a followup showing the completion of that filter and testing it to see how close to the ideal response it gets. It's designed to take around 400 watts at 1296 MHz (23 cm wavelength) and provide around 40-50 dB suppression of the 3rd/4th/5th harmonic and a fair suppression of the 2nd harmonic, along with very low through loss and still providing a very good match to 50 ohms surge impedance. The dimensions are generated from Tchebyshev polynomials to get a very sharp roll-off about about 1450 MHz. These use N-type coaxial connectors, I usually use 7-16 DIN connectors as they can handle more power and are more robust when you are using hardline or heliax coaxial lines. It's a nice challenge to get the dimensions precise so it performs very close to the modelled specification.

[@Xsiondu]

@@MachiningandMicrowaves WAIT, WAIT, WAIT if I was buying a notch filter to block out channel 4 so my hotel could insert their own brand advertising channel and I made a mistake and the normal 1 rack unit device from blounder tongue that usually cost 700 dollars cost 3000 dollars and it wasn't one ru but instead 4 grey cylinders of varying height all interconnected with really thick coax. And because it was custom made I couldn't return it but dang it worked better than any of those rack units ever did ... Is that what your building? And if so... How did KZbin read my mind because I was thinking about that mistake 3 days ago but I didn't say anything out loud about it

[@Xsiondu]

@@MachiningandMicrowaves and is 3rd/4th/5th harmonics called composite triple beat in cable tv land?

[@MachiningandMicrowaves]

@@Xsiondu Almost certainly!

[@MachiningandMicrowaves]

@@Xsiondu Definitely not a filter! 100% about focussing microwave energy into a receiver or out of a transmitter to increase the directivity

[@Strothy2]

ahhh the "pling" of great success... 4:42 :D

[@MachiningandMicrowaves]

What WAS I thinking? I am an idiot.

[@ollysworkshop]

Good to hear Mrs Trellis is still going strong. I never realised she was into machining as well!

[@ollysworkshop]

....in other news, I had a visit from Samantha the other day. Her cavity filter needed reaming.....

[@MachiningandMicrowaves]

Ah, the joys of Radio Four. I should have a listen once in a while. Caroline listened to it a lot when she started losing her sight. I am more of a Radio Three aficionado since John Peel died and evenings on Radio One became slightly less eclectic.

[@foxbat888]

No wonder you're so skilled and creative, you have such a friendly and helpful superego

[@MachiningandMicrowaves]

For certain EXTREMELY specialized meanings of "friendly" and "helpful" perhaps

[@iiredeye]

look up optical centre punch. Axminster tools can supply...makes a Starrett auto look like a stab in the dark :)

[@MachiningandMicrowaves]

When it *really* matters, I use a centrescope on the mill to get things aligned perfectly. My father had an optical punch that was probably shop-made. I did think about the SPI one, and the Mitutoyo 985-139 but they seem to be unobtanium. I suspect buying a new one could be a bit like the centering gauge I once bought and used a massive three times. However, for the price it might be good for those few occasions where nothing else will do the job. I tend to use the Starrett by feel on the intersection of layout lines. I have a nice old vernier travelling microscope that I used to use, but now I just fit a camera on the centrescope on the mill and measure using the crosshairs and DRO.

[@Hansengineering]

Wow the plot on that Smith chart is SOMETHING!

[@MachiningandMicrowaves]

Hi Erik, that's the output from the QUCS Studio simulation, which is why it looks so good. The real one is not going to be as neat!

[@jangoofy]

Happy to stumble upon the UK version of This Old Tony :-)

[@MachiningandMicrowaves]

I am wholly unworthy and don't deserve to be spoken of in the same breath as The Master....

[@omsingharjit]

so This is the technology in Our smartphone RF technology inside .

[@soranuareane]

My parents have a fridge magnet that says "measure once, curse twice". Perhaps something for your shop?

[@MachiningandMicrowaves]

I usually measure three times, don't believe it, measure again and then once more for luck. If three of the readings match, they are probably ALL wrong.

[@AndyFletcherX31]

Sounds like Mrs Trellis (of North Wales) was more on topic than usual, she must be taking her medication these days.

[@kevina.4036]

Reminds me of a sliding load. Think you could machine one of those?

[@MachiningandMicrowaves]

There are some amusing challenges in making things like sliding loads and phase trimmers, but sliding loads are perhaps a little bit easier. I've never seeing one dismantled, just seen a Maury 2517H with APC-7 from a distance and some ancient HP thing that looked like it was from a 1940s radar lab or something. I don't own a "proper" VNA, best I have is a two-port bidirectional PocketVNA, nothing that is a two-person lift into a rack. I did make a slotted line with help from my dad back in the 70s, that was fun. I could do wiht some phase trimmers, I have some nice ones that can do about 15mm range on SMAs, but I'd like some for a combiner on 1296 MHz, those are a little short. now I can do gold and silver plating, it might be time to have another think. 187 other projects to get done first though.

[@apexmcboob5161]

I totally get your enthusiasm for high quality tools. It's almost like the precision in the tool's manufacturing gets transferred to the user. I find I do better, more careful work with a finely made tool and they help keep me calm when frustration sets in.

[@roseroserose588]

I think the fact that you're not also trying to fight your tools really helps, a good tool should get out of your way entirely and allow you to perform the task at hand. I've recently upgraded from a cheap chinese clone 3d printer to a shiny new Creality machine and I've found i'm using it more than i ever have, because it Just Works when I need it to

[@MachiningandMicrowaves]

I'm also a woodworker and I get a lot of pleasure from using a REALLY good chisel on some green oak when I'm framing up a building, or using my dad's old boxwood plane. Same feeling with my Hewlett-Packard, Agilent, Racal, Marconi, Agilent an Fluke electronic test equipment

[@apexmcboob5161]

@@MachiningandMicrowaves 1. I'd have to add Keithley, General Radio, Stanford Research and most of all Tektronix to your list. 2. We only have red oak & white oak in Canada, is green a different UK species or just not completely dried and aged?

[@MachiningandMicrowaves]

@@apexmcboob5161 Green in that context means freshly felled and sawn. It cuts like butter if your chisel is razor-sharp. Two or three years later it's like iron. I built my sun-room using traditional oak-framed carpentry, with wooden "nails", or trenails and self-locking joints. We are on the border between Quercus robur (English Oak) to the south and and Quercus petraea, Sessile oak to the north. Hilariously, here we are in a nice temperate climate, but my latitude is level with the top end of James Bay, way in the frozen north of Ontario. Just about level with the bottom of Polar Bear Provincial Park in fact. Sounds chilly. I've never been further north than Haileybury and North Bay and now my remaining relatives live in the melon belt near Niagara-on-the-Lake

[@johnydl]

Okay just caught the ISIHAC reference made me laugh out loud XD

[@xenoxaos1]

Am I an asshole for thinking this could have been done with threaded rods?

[@MachiningandMicrowaves]

It's certainly possible if you add in a calculation adjustment for the effective diameter and any length modification needed for all those wiggles up and down in the 12-15 micrometres of skin depth. At lower frequencies it's pretty much the norm to use threaded rod. If the disks were threaded and held in place with a stainless steel locknut on the low side, you could still solder them, but you'd have to make them from the top down so you could get the nuts off. Easy enough to use calipers to set the spacing. You'd need to do something about the ends to ensure you had a constant impedance, perhaps filing a spigot on the ends or using a different type of coaxial socket and changing the inside diameter of the holes in the end walls. I don't honestly know what the effect on the electrical length and inductance would be, as although the total path length in the skin-effect layer would be roughly doubled, the electric field might not be affected that much. You wouldn't be able to get copper threaded rod, but brass is available. That's rather more lossy than copper, but you could silver-plate it easily enough. It takes much longer to show the process than it takes to do the soldering once you've made the end pins and spacers. I need to make at least ten, so the extra faffing about making the spacers is perhaps worth it. Perhaps.

[@xenoxaos1]

@@MachiningandMicrowaves I was thinking only tapping like 1-2mm of the end of the rods. You could then solder inside the threads where it meets the discs

[@theafro]

i can't watch Keith Rucker any more, at first, you forgive his odd mistake as nobody's perfect, but eventually it dawns that in reality, he's just a bloke with loads of machines in his shed, and far less of a clue about how to use them than he thinks he does. shame, because he does have some rather cool old machines.

[@MachiningandMicrowaves]

His comments section certainly makes for interesting reading.

[@Thefreakyfreek]

I have the exact same micrometer I love it simple and exact and a lot cheaper than the electronic ones

[@MachiningandMicrowaves]

I have some excellent Tesa, Brown and Sharp, Starrett, Moore and Wright and Shardlow mics as well as some Mitutoyos. I love the feel of the Tesas in my hand, but those sloping scale lines make my brain squeak. I have some Insize groove mics that are very good for the price, but my Mitutoyo 6mm diameter groove mic is next-level. I have a range of Tesa three-point bore mics plus some Mitutoyos and a Bowers that are all a joy to use. I gave up on the didgycals after the second one died on me. They are excellent for incrementals and error measurements, so I might treat myself to a new IP67 Mitu one day. Maybe.

[@mrwidget42]

I should stop watching Adam Booths channel, then. He just showed off his groove micrometer. That just gives me a case of metrology envy. Next one up for me to waste/invest some bux on should be a back button DTI. Makes tramming up the mill to the nearest nanobanana a lot less tedious.

[@MachiningandMicrowaves]

@@mrwidget42 I'd better have a look. I found a few decent ones on ebay at sensible prices, but the supply of them in the UK has dropped off. I was hoping to bid on a Harig Grind-all last week, but the thing sold for £500+ and it wasn't in showroom condition. Tom Lipton's fault about the demand for those.

[@saxunvladimir892]

💎💎💎💎💎💎💎💎💎💎💎💎💎👍👍👍👍👍👍👍👍👍👍👍👍🐼🐼🐼😘😘😘😘😘😘😘😘🔟🔟🔟🔟🔟🔟🔟🍧🍓✔ 🇸🇰

[@mrwidget42]

My cat was named right after the passing of Humphrey Lyttleton. His name was Chairman Humph. I swear I could read the thought balloons over his head..._Mornington Crescent_!

[@MachiningandMicrowaves]

Excellent name indeed. One of my friends had a cat named "Spokesman Said".

[@mrwidget42]

@@MachiningandMicrowaves Yup. Humph, who was a shorthair tuxedo, was the most vocally needy cat I have known. We took to calling him "Mister Personality". I had to let him go to kittykat heaven last year he was so poorly, and I still miss him awful. I would post a photo of him in his full Humph-ness, but I suppose KZbin would balk at the cheeky quotient.

[@DotiksFlores123]

Nice presentation

[@MachiningandMicrowaves]

Thanks!

[@sommersetcoker5455]

holy crap, i just realised i wasnt subscribed to this!!! >.< how many have i missed!@#!

[@MachiningandMicrowaves]

It's a Stealth channel, I'm running a cloaking device that hides the channel very effectively. Welcome aboard...

[@LordOfNihil]

its funny when people assume americans don't use metric units. every measuring device ive ever used is a dual scale gadget with mm along side inches, i prefer to do math in metric. unnecessary conversions lead to (sometimes fatal/expensive) error. the only real blowback you get is with hardware stores, and people who live in their cars more than their homes. military uses metric. nasa uses imperial units for some damn reason. i dont like having to keep 2 sets of tools either.

[@MachiningandMicrowaves]

I love my Imperial metrology gear, I still have my father's Starrett micrometers and a Mitutoyo height gauge calibrated in bananas, sorry, inches. I cut a lot of Whitworth and UNC/UNF threads and deal in Angstrom units, micro-inches and nanometres, foot-pounds, acre-feet, quarts and both types of gallon. I'm form that mixed-up post-Boomer lost generation that had to deal with Imperial, Avoirdupois, Troy, CGS, MKS and SI, but still managed to make a few things, despite unit confusion. Thou, mil, mill, tenth, all mean something to me, and I think it's helped to give use a certain flexibiity of mind. I can still handle money in pounds, shillings and pence, from back in the times when a US cent and a British penny were about the same value.

[@daretodreamtofly3288]

Not but a minute into it and your crippling dependence on collects is obvious. You go any smaller and you'll need a jewelers lathe.

[@MachiningandMicrowaves]

Just buying a baby ER16 collet chuck and D1-4 backplate to fix the problem. Should be fine down to 0.5mm so long as I make the backplate extremely finely adjustable. "If you can't make it accurate, make it adjustable", as a great engineer once said. Allegedly

[@irishwristwatch2487]

Im gonna continue watching these videos and pretending the maths isnt soaring so high about my head it'll hit the moon That copper does look nice to work with though, not like that crap we use at work....

[@MachiningandMicrowaves]

Well you are spot on there! This filter is part is a system that's going to be used to bounce 1.3 GHz radio signals off the surface of the Moon. That feat is close to the limits of Physics so every part of the transmit and receive system and antenna has to be right at the top of its game. I use a lot of C111 Sulphur Copper as well. Nice short chips, not very gummy, doesn't cold-weld to tools much and the surface doesn't tear too badly. I do have a bit of C101 left, but that is only suitable for making vice jaws or earthing bars!

[@juyfjgfjhgfjugf4702]

im here for the jokes and machining,I have no idea what else youre talking about ,thank you!!

[@MachiningandMicrowaves]

Me neither! Thanks for watchin', as a great man once said.

[@hardwareful]

This RF engineering gave me severe PTFE

[@g0fvt]

Fascinating, I am not sure I will be making any SHF filters this week but it is interesting to see your approach. Something I often see "Clickspring" doing when machining on a fixture is he uses superglue to attach the part to the arbor or whatever and releases it with heat when he has finished. FWIW when I was building my 2m amplifier many years ago a friend in the print trade got me some saturated silver nitrate (I think) it was a case of having the copper parts absolutely clean and then dipping them for literally under a second in the solution and then straight under the tap. It might be feasible to do similar to the internal section of your filter, just throwing the idea out, it might benefit someone.

[@MachiningandMicrowaves]

My dad showed me how to use beeswax mixed with Carnauba wax to stick watch parts to a blank bit of bar to spin them on the lathe, but these days I do use high-viscosity cyanoacrylate as in part 1 of this series. Works like magic and the best bit is that it leaves no residue on the workpiece, it all stays on the chuck and you can just do a face skim and get a totally fresh surface. kzbin.info/www/bejne/g4KkiJiBjtiDbLM I've got all the kit to do the surface prep, electro-nickel plate then silver electroplate. Whether it's worth it is very questionable as the Q is very low and the circulating currents are not many times the input current, (about 3.2 amps at 500 watts). Skin depth at 23cm is about 2 micrometres, so I need at least 12 micrometres of silver over the very lossy nickel to ensure that almost all of the current flows in the silver. I'm not sure how thick a layer you can get using just chemical deposition. I guess it's possible to find out by masking an area off and doing a dip, then measuring the thickness of the silvered area compared with the non-silvered. Loads of fun to be had anyway!

[@g0fvt]

@@MachiningandMicrowaves thanks for the reply, no idea how thick this silver was, I only used it on the PCB for my active bias circuit where of course there is no RF, made the board look a bit more professional and it was lacquered as soon as it was populated. Sorry I did not see that detail in Part 1.

[@MachiningandMicrowaves]

@@g0fvt I bet some smart Surface Chemist would be able to explain how the dip plating works and how thick the layer is. It probably stops once no more copper ions can react with the silver in solution. Definitely great for the bling factor! Cheers, Neil G4DBN

[@g0fvt]

@@MachiningandMicrowaves it is certainly a flaky process, the silver "flashes" on very fast but strips back off if you don't move very fast. Took a few tries to get it right. Bare copper on a PCB does look bad, of course no-one except me got to see it....

[@cmdrdamaskus3308]

banannas lol

[@MachiningandMicrowaves]

kzbin.info/www/bejne/ZmTSdWONoJuAsK8

[@soranuareane]

But how will the solder affect the frequency response? Solder is conductive, does it have a substantially different frequency response (assuming perfect joint with no bubbles or fissures)?

[@MachiningandMicrowaves]

the fillet is very small, and I'll try hard to keep the meniscus level with the face of the disk. The main effect is resistive loss, but also the skin effect is different in solder. Ideally I would machine off the excess solder afterwards, but that's too fiddly. I'll probably use a file for any excess, then clean and polish and silver-plate the whole central assembly. I don't expect that the nickel underlayer will plate over solder! At this relatively low frequency it's not much of a problem, but the losses get significant above 6 GHz and you have to think very carefully before using solder in a filter cavity

[@WillieAndBilly]

I'm curious what software you used for the Smith chart, and defining the sizes needed. having been used to playing with a network analyser on 900MHz and 3GHz systems, I look forward to your testing of the filter.

[@MachiningandMicrowaves]

That was in QUCS studio, running a model of the filter. I'll test the finished thing on one of my cheapo VNAs and on the signal generator/power meter and on the spectrum analyser just to make totally sure the instruments agree and to see how close the result is to reality. The design process is almost exactly the same as for a lumped-element filter. You generate the values of the lumped components, then decide on the diameter of the core and disks and the ID of the tube. Then it's just a case of working out the lengths of the disks and the caps to get the correct values of capacitance and inductance, using the expressions for fringing fields from Whinery, Jamieson and Robbins (1944) and S.B.Cohn (1960s). Dominique Delerablee made an excellent Excel design sheet that does all of the calculations including the basic inductance/capacitance values to match the requirements of the filter. f1frv.free.fr/main3e_Filtres_LP.html I'll run through the process in the next video if I get a chance

[@williefleete]

Love the added sound effects

[@petrsuchomel9639]

One question, would not be it easier to make filter in rectangular shape - at least one can save self from the pain of turning/boring, easier to do it all on CNC mill? Or there is a problem of build up on the edges? Otherwise damn, cool man!

[@MachiningandMicrowaves]

The power-handling capacity of a square version would be slightly lower, or the gap would need to be slightly wider, because of the field concentrations at the corners. The capacitance and inductance calculations are very much simpler when you have a pure radial field from circular elements. In theory, you can use any shape that results in the correct Chebyshev polynomial solution. It works fine in a slab line with flat elements, or even with etched metal elements on a substrate. Also, I don't have a CNC mill, so simple round versions are easy for me. There are some commercial versions from Huber & Suhner and others that use the coaxial topology.

[@petrsuchomel9639]

@@MachiningandMicrowaves yes, the field concentration in the corners is what concerns me. I am thinking that the best approach for me would be to take 1/2 of profile, do a rough milling (fast), then bolt them together and ream/bore them on lathe for a final diameter - I can also use dow pins to make sure the alignment is perfect between two pieces. This would make it much easier to solder, adjust etc., as you can always take it apart.

[@trollforge]

Sixtyfourths, what ever they are? They're about a third of one of your millipedes... ;) You know, it strikes me as odd, that you chasers of the micron show such distain for one thousandth of one twelfth of a set standard, while you prefer One millionth of one ten-millionth of the distance from the equator to the North Pole along a great circle, assuming the Earth's circumference is estimated to be approximately 40000 km... Sounds to me like some Frenchman's best guess! ;)

[@MachiningandMicrowaves]

Yes, the metre isn't exactly human-centric, is it... I need one of those Caesium rulers, but I'd keep losing count once I got to a few billion wavelengths. My right thumb, when pressed just hard enough, is exactly one inch wide. My thumbnail is 25 thou. Useful for spark plug gaps. Despite my protestations, I am totally fine with non-metric measurement frameworks. I have a good set of inch calipers and micrometers, plug gauges and jo-blocks, taps and dies and drills and even an inch height gauge.

[@trollforge]

@@MachiningandMicrowaves my height gauge is dual... I didn't pay enough attention when I brought it, on sale, mm and 128ths of an inch... Oh well, set it with the vernier calipers at the edge of the layout block...

[@MachiningandMicrowaves]

@@trollforge I picked up a pair of Mitutoyo dual-format vernier height gauges in an auction last year for very little money. I find I rarely use the scales though, I almost always make a stack of gauge blocks as a reference, except for a bit of Dykem and scriber layout when I'm feeling old-school

[@trollforge]

@@MachiningandMicrowaves someday I hope to be able to justify the expense of a set of gauge blocks...

[@Ashley.0000]

Another CEE fan gotta love Kurtus's banana ruler.

[@MachiningandMicrowaves]

He's a fine chap, that Kurtis. Makes a good part.

[@2mrRB]

Really looking forward to your next video. It looks like you are are drilling at a rather high rpm. Do you have a reason for that?

[@MachiningandMicrowaves]

A lot of the machining segments are running at between 2x and 8x, the actual speed for the 3mm drills is around 1300 rpm and the 6mm is about 600 rpm. These drills are cobalt HSS and the manufacturer's cutting data says they should run about 130 metres/minute in aluminium for drills over 9.5 mm diameter. At 1300 rpm with a 3mm drill, that's only 12 metres/minute, so I'm running them very slow. That speed is based on maximising metal removal though, and results in a limited lifetime. Important in a production environment and also to ensure good chip formation. My poor old Bridgeport can manage about 3600 rpm max and the lathe can only do 1800 rpm, so I'm very limited on small drills. i do use a toolpost spindle for very small drills, that can do 12000 rpm, but I need to go MUCH faster than that for milling 0.4 mm face grooves while turning the lathe at 4 to 8 rpm. I'm making a new toolpost spindle which should be able to go more than 40,000 rpm, but that's serious engineering. A 25 mm drill should run at almost 2000 rpm according the the published data. I run them at more like 300 rpm in aluminium. I should be able to run a drill that size at 750 rpm in most steels, but I choose not to. It just feels mean to the poor drill. I should really label the sped-up video segments or show the cutting data in an on-screen table, but it would take AGES unless I can find a way to automate the process.

[@2mrRB]

@Machining and Microwaves thanks a lot for your reply. That makes sense indeed, sometimes im unable to see the video speed. In that case it's probably best to show the video speed instead of showing the machine set up and machining speeds

[@flikflak24]

Well everything we turn is only hold by friction

[@MachiningandMicrowaves]

Yep, I guess this is somewhere on a scale between "Six-jaw Bison with Abom Torque" and "A wing and a prayer", but closer to the latter.

[@someoneelse7629]

Ahh, the constant hints to BlondieHacks and now a kick at Curtis from CEE too, I love how the machining gang on KZbin all seems to follow each other. Curtis is on the other end of the scale of worksize though.

[@MachiningandMicrowaves]

Kurtis is a proper star, so chilled. Brian Block is under-appreciated, I could watch him do stuff for hours on those giant machines. There are lots of channels doing interesting stuff, I must update my little hall of fame

[@chrisstephens6673]

The new guy on the block is "inheritance machining". Very well produced videos light hearted but good content.

[@MachiningandMicrowaves]

@@chrisstephens6673 Thanks for the pointer, really enjoyed watching him.

[@nottelling6598]

26:27 Oh look, a spring pass on most of the threading.

[@MachiningandMicrowaves]

Hahaha, yes indeed.

[@parttime9070]

Why is there a photo of some woman always on the opening frame.? I don't get it, seems creepy..

[@gigigigiontis8]

It's a computer generated image

[@MachiningandMicrowaves]

Have you actually watched any of the videos? AIMEE is a regular star of the show. Her image was generated using a Generative Adversarial Algorithm, where one program generates what it thinks might be a human-looking image, which is then judged by a second program to decide if it can tell whether the image is real or not. The images are very much in the Uncanny Valley rnage, and all of them are slightly disturbing in one way or another. The image is just a collection of byte arrays and has no connection with any human, but elicits a wide range of responses, from proposals of marriage to OMIGODS CRINGE. Interesting. Her (well, I say "her", it's strange assigning a gender to some luminance tables) script is passed using a JSON body to Google Cloud TTS. The request body is like this: "audioConfig": { "audioEncoding": "LINEAR16", "pitch": -1.2, "speakingRate": 1.15 }, "input": { "text": "Google Cloud Text-to-Speech enables developers to synthesize natural-sounding speech with 100+ voices, available in multiple languages and variants. It applies DeepMind’s groundbreaking research in WaveNet and Google’s powerful neural networks to deliver the highest fidelity possible. As an easy-to-use API, you can create lifelike interactions with your users, across many applications and devices." }, "voice": { "languageCode": "en-GB", "name": "en-GB-Wavenet-C" } but it can also interpret SSML markup for emphasis, pauses, rate, pitch and lot of other features. It should also be able to interpret the international phonetic alphabet for custom pronunciations, but I haven't have the energy to play with that. A fascinating psychological experiment anyway.

[@dandeeteeyem2170]

... For those listening on monochrome sets, that's about a thou' 😂 love the shout out to Curtis from Cutting Edge 😁

[@MachiningandMicrowaves]

The total competence Kurtis shows on the channel is inspiring, and Karen's camera work really adds another level that isn't easy when you work alone.

[@dandeeteeyem2170]

@@MachiningandMicrowaves the outtakes at the end of the videos are hilarious 😂

[@jimsvideos7201]

3:57 Pizzi-cut-o?

[@PeregrineBF]

I love wrens! Please don't disturb the baby birds.

[@MachiningandMicrowaves]

I have three wren nests this year, one in the old pighouse, one in the barn where the machine shop is, and another in a different outbuilding. They are really fierce about defending their territories. There are two blackbirds nesting in the barn and I'm expecting two or three swallows to nest up in the roof. I have five Swift boxes, but two of them have large broods of starlings. I saw swifts yeasterday so it's almost time to start playing the recorded swift calls. I had buzzards nesting in one of the big trees, I hope they will return this year. I have 38 nestboxes. There is a great tit nesting inside one of my antenna mast bases and I have 200 metres of hedges with many blackbird, thrush and chaffinch nests. Up in the pine trees I have goldcrests and somewhere there's long-tailed tits, but I never see their nests. No owls this year, I have a barn owl box up high on the end wall of my barn and a tawny owl box up high on a tree, although it was used by kestrels. About ten years ago I had more bird feeders and in 12 months they ate over 1 ton of sunflower seeds and peanuts. I have woodpeckers and a flock of tree sparrows. I'm watching five goldfinches and some greenfinches on the feeders right now, with three blackbirds a dunnock and two collared doves on the grass beneath. I also have a resident sparrowhawk who takes a few of the other birds, mostly the doves. As well as the birds, I have bats and stoats, as well as a population of field mice and brown rats, voles, moles, frogs, newts, toads, grass snakes, shrews and at least four hedgehogs. The wrens are safe here.

[@PeregrineBF]

@@MachiningandMicrowaves Very nice!

[@jontylewis7301]

That tailstock DRO is brilliant haha

[@MachiningandMicrowaves]

I made a vid about it, the idea came from ChrisB257 and it seemed like a really useful addition. Very simple, no critical parts - other than the collar needs to be a good fit. You can even have a remote display on most of those cheapo scales if you don't mind trailing wires. kzbin.info/www/bejne/jWPLnItsoZaCgJI

[@vladimirsvirid7705]

ty

[@HexenzirkelZuluhed]

So you did set up a chamfering tool for use perpendicular to the axis. Also nice work on the rest, but I understand little about the "Microwaves" Part of your channel.

[@MachiningandMicrowaves]

I have three chamfering tools on the lathe. One mounts at 90 degrees to the lathe axis on the side of the QC toolpost, the other two mount on the back of the toolpost parallel to the lathe axis. One is for internal chamfers, HSS with an undercut, the other is the same as the 90 degree one, but mounted the other way round in the toolholder. I'd really like one that is double-ended. I hope things might become a little clearer as I explain how the thing is designed and then show it in use. I'm just sorry I don't have enough time to do more videos at the moment, there are some BIG projects going on that I'm not permitted to discuss yet, but also I'm working through an 18 month backlog of projects that were held up during Caroline's final illness. June 22nd is going to be tough, and would have been her birthday on 4 weeks time. It's not getting any easier yet, but keeping flat-out busy is my go-to coping mechanism.

[@_cheesestraws_]

Glad to see Mrs. Trellis is still keeping interested in what's going on in the world...

[@MachiningandMicrowaves]

Oh, I do miss Humph....

[@Chriss120]

looking at the graphs at the beginning it seems to be a filter of quite the high order.

[@MachiningandMicrowaves]

It's ninth order, Cheby type 1. It's set for 0.01dB ripple and about 1450 MHz corner frequency. I only need about 35dB attenuation at the second harmonic and 45dB at the third with my amplifier, but it is running well below the P1dB level. Folks running digital modes and CW for moonbounce will be pushing well into gain compression, so need a bit better suppression, it's well over 55dB at the third harmonic but of course it does the usual weird thing above 7-8GHz where the through loss drops again, but that's the price you pay for simplicity of mechanical design. 1.3 GHz amps aren't going to be making much output above the 5th harmonic, and feeder loss and antenna mismatch will all help to soak up anything the filter doesn't remove. I'll go through the design process in another vid to show my working...

[@9h1gb]

Hi Neil. Sorry to bother here but looks like you have an issue with your mailing address. Can you check you email mailbox please

[@MachiningandMicrowaves]

Hi Mans, I am sitting amongst a pile of components and boxes of F6BVA transverters. I've agreed with my day job employer that I can drop to one day a week for the rest of May so I can get caught up with everything. I've been trying to put the kits together at weekends, but with visitors and gardening and house repairs, it's hard to concentrate. Your call is one of the first on my alphabetic list so I'll drop you an email soon as I'm ready to ship. I have to finish a project for a TV programme that has a deadline of the 23rd, but now I don't have to be at work so much, I'm making progress with boxing up the kits. I seem to be missing a lot of emails from PA5Y and some UK folks. Not sure why, I don't get an entry in my spam list, the email service just drops them. Mostly BT Internet accounts, but also some other EU providers. I have an alternate email neilg4dbn AT gmail DOT com that works well. There are 800 component sheets and about 7,000 strips of adhesive tape, 90 cardboard boxes and six crates of parts on Caroline's old sewing table. It's been a huge mental strain because of how involved Caroline was with the project, but I've got two photos of her on the wall looking at me to make sure I get the job finished.

[@9h1gb]

Hi Neil. Thank you for the reply and my apologies for my Late reply. I am having an issue with KZbin reply notifications. I can now read the replies on the KZbin channel but still I am not receiving any notifications via email and that is why I did not answer before. I will contact KZbin again for assistance on this mater. All noted and yes one does seem to not have enough time to do all. I know I have been there too. So I fully understand. I will await your email re kit and if I can I help in any way please let me know. Now that I know I have this issue I will regularly check the KZbin reply on your channel. Regards and stay safe.

[@AdamEarl2]

Inserts aren’t designed for spring cuts. They usually need 0.125mm DOC in order to use the edge prep, otherwise youre just rubbing material

[@FixItStupid]

TY Good Tech

[@NuttyforNissan]

Conductive heat loss! how much power are you passing through this?

[@MachiningandMicrowaves]

I've tested one of these at 500 watts for 30 minutes using a watercooled amplifier and a huge 50 ohm load and there was no significant rise in the case temperature. Checking the inside temp isn't easy, but aiming an IR thermometer through the hole did show a rise of 6 to 10m degrees. I suspect there's a bit of convective cooling going on inside. Production models will be silver plated to reduce loss a little more, but at those power levels the loss is tiny, the main "loss" is actually a reflection mismatch rather than a resistive dissipation. The coaxial cables did get a little warm after 30 minutes, but only ten degrees over ambient as seen on a FLIR. I think that with 7-16 DIN connectors, it would happily pass a kilowatt without any stress.

[@JainZar1]

Is the frequency change from thermal expansion/contraction negligible in this application? I imagine the cut-off frequency must shift quite a bit in the summer, compared to winter.

[@MachiningandMicrowaves]

The expansion of the disks and aluminium are similar, so there's little effect on the capacitor values as a result, the main change is probably to the spacing, but this is a very low-Q system, unlike a resonant cavity, where it cam matter a great deal. Some cavities are indeed tuned by varying the temperature of the body of the resonant chamber. A change of 16 parts per million per Kelvin, even by 40 degrees, would only change the inductance of the copper core by a factor of 0.00064, so potentially could drop the corner frequency by about 1MHz, but other compensating factors might reduce that a little. As the corner frequency is around 1450 MHz and the filter is working at 1296 GHz, the change would be almost impossible to detect. However, that level of temperature change might cause diurnal pumping and the cavity might get some condensation, so it's safest to keep the body of the filter well above the dew point or perhaps feed it with a whiff of dry air if it's going to be exposed to outdoor temperature swings.

[@JainZar1]

@@MachiningandMicrowaves Thanks for the detailed answer! Liked the new animation btw.