RFQ - CNC Lathe parts - Aluminium - Metric - RADAR feed Horns

Hi all,

I'm working on a RADAR project that requires tiny aluminium feed horns.

The zip file enclosed gives the mechanical detail.

The info is all there but it's buried in different places. The 3D drawings are from the EM simulator that we use which is not a

mechanical drafting program and has very limited dimensioning abilities. It won't let us put dimensions on hole centres for

example.

Prototypes were hand made but they really need to be done by CNC for repeatability.


The diagrams show two types of horn and a coupler that secures either horn to a PCB.

I need a quote for 20 couplers, 10 of the large horn and 10 of the small Chaparral feed.

thanks

Tim

Hi Tim , I didn't see the zip file you attached. Where should I find it?
East

Its an attachment at the bottom of the post.

Quote:

---

Originally Posted by TSL;2313748)

I need a quote for 20 couplers, 10 of the large horn and 10 of the small Chaparral feed.

---

I need to revise the quantity
10 couplers
10 Chapperal feeds
2 large horns.

Additional detail just supplied by our designer

Material = Aluminium 6082 T6 AL6082

Interior surface roughness Ra = 0.02um or better.

Tolerance = 0.01mm


thanks

Tim

Material = Aluminium 6082 T6 AL6082
Not a problem.

Interior surface roughness Ra = 0.02um or better.
Tolerance = 0.01mm

Um - the tolerance of 10 microns is doable, but the surface finish spec of 20 nanometres is REALLY pushing it!!!
As they stand, your specs would permit a surface waviness of 10 microns, which is 500x rougher than the spec'ced roughness.

I may of course be off with the fairies, but are you SURE?

Cheers
Roger

Quote:

---

Originally Posted by RCaffin

Material = Aluminium 6082 T6 AL6082
Not a problem.

Interior surface roughness Ra = 0.02um or better.
Tolerance = 0.01mm

Um - the tolerance of 10 microns is doable, but the surface finish spec of 20 nanometres is REALLY pushing it!!!
As they stand, your specs would permit a surface waviness of 10 microns, which is 500x rougher than the spec'ced roughness.

I may of course be off with the fairies, but are you SURE?

Cheers
Roger

---

Hi Roger,

Yeah its because these are operating at 122GHz where surface roughness becomes a loss inducing concern.

At 100GHz the skin depth of the signal in Aluminium is 0.28um, so the Ra needs to be better that that to avoid too much attenuation of the signal.

I know 0.02um can be achieved with a diamond drill directly and/or interior magnetic polishing but we may have to settle for less dependent on cost of manufacture.

regards

Tim

I see. That high in frequency. OK.
All the same, I think you need some different specs than what you have given. A surface waviness of 5 um is within your specs, but I suspect it might not suit the actual needs.

Have you considered a plastic injection moulding with a heavy metal plating over the plastic? That might meet your Roughness requirements, and you could of course silver-plate the interior afterwards as well.

I would not bother with the diamond drill bit: that is not the right application for it. For machining hard glass or hard rock, yes, and I have machined up to Mohr 9 that way (Ruby, Basalt, carbide). Polished HSS would be more than enough with aluminium.

Al 6082 T6 - not so common. 6061 or 6060 would meet your requirements quite well, and they are more readily available in many sizes.

Your RFQ is not simple.
* First one would have to generate properly dimensioned drawings. They don't need to be 3D models: 2D drawings would be enough for programming.
* Then one would have to make some machining jigs to hold the parts, and they would have to be made very carefully. I think you would still need jigs with a 5-axis machine.
* Then one would get around to machining the parts. Anti-climax: automated.
* Then one might have to polish some of the surfaces to meet your Roughness specs. That is likely to be a manual job.

Or one could knock up some plastic parts and have them vacuum-coated. But that only shortens some of the steps.

Cheers
Roger
PS: if anyone tries to sell you on 3D printing - forget it. A dimensional tolerance of maybe 0.5 mm and a surface roughness not much better than that.

Sounds like the requester needs to go back to the drawing board, literally. Asking the bidders to do his designing does not seem fair.

Quote:

---

Originally Posted by RCaffin

I see. That high in frequency. OK.
All the same, I think you need some different specs than what you have given. A surface waviness of 5 um is within your specs, but I suspect it might not suit the actual needs.

Have you considered a plastic injection moulding with a heavy metal plating over the plastic? That might meet your Roughness requirements, and you could of course silver-plate the interior afterwards as well.

I would not bother with the diamond drill bit: that is not the right application for it. For machining hard glass or hard rock, yes, and I have machined up to Mohr 9 that way (Ruby, Basalt, carbide). Polished HSS would be more than enough with aluminium.

Al 6082 T6 - not so common. 6061 or 6060 would meet your requirements quite well, and they are more readily available in many sizes.

Your RFQ is not simple.
* First one would have to generate properly dimensioned drawings. They don't need to be 3D models: 2D drawings would be enough for programming.
* Then one would have to make some machining jigs to hold the parts, and they would have to be made very carefully. I think you would still need jigs with a 5-axis machine.
* Then one would get around to machining the parts. Anti-climax: automated.
* Then one might have to polish some of the surfaces to meet your Roughness specs. That is likely to be a manual job.

Or one could knock up some plastic parts and have them vacuum-coated. But that only shortens some of the steps.

Cheers
Roger
PS: if anyone tries to sell you on 3D printing - forget it. A dimensional tolerance of maybe 0.5 mm and a surface roughness not much better than that.

---

Hi Roger,
Jigs not required - we've already machined these items on a small tool lathe where our EE even made a tiny tool to machine the 1mm grooves in the top of the feed, so we know these items can be turned on a lathe since we've already done that and done by our EE who is does machining as a hobby.

We're just looking for a plant that can manufacture these things in small quantities but with better specifications than what we've achieved manually and make them look nice too.

Yes the 2D drawings suck, but that is the output of our EM simulator program that microwave horn and wave guide design is done in.

We know that a carbide tool will produce an Ra of about 0.15um in aluminium and we know that a diamond tool bit will produce an Ra of about 0.02um in aluminium regardless of whether that's a "suitable" tool for machining aluminium. This is data we have from actual tests done in Japan machining wave guides for 70-116GHz back in 2012.

This is the first time we've had requirements to manufacture wave guide, PCB's etc we've done for years so this is a learning experience for us.

regards

Tim

Quote:

---

Originally Posted by wmgeorge

Sounds like the requester needs to go back to the drawing board, literally. Asking the bidders to do his designing does not seem fair.

---

That's not fair.

We're not asking anyone to do the designing, that's been done, prototypes made and proven.

We're hoping to meet somewhere in the middle between hand made parts and machined perfection.

We know what finish we need for optimal performance to minimise losses and given the actual transmitter power is only 500 micro watts, every tiny bit of loss makes a difference.

I will concede our 2D diagrams suck and to that end I've spent last night in CAD training learning a new software package so I can redraw the 2D diagrams in a suitable format.

To quote the meme on my coffee mug ... "Perfection is our goal, excellence will be tolerated" :)

When you talk about a diamond tool bit, just what do you mean? We may be talking about different things.

You see, to me a diamond tool bit is a shaft or a disk encrusted with very small diamonds plated on. They have a surface a bit like a disposable nail file. You only use these diamond tools for materials which are orders of magnitude harder than aluminium alloy. As I mentioned, I have used these tools on some projects, on very hard materials up to Mohr 9.

Or do you mean PCD tools: cutters coated with polycrystalline diamond? They are a totally different beast. They will not be as sharp as good aluminium-oriented HSS for a number of technical reasons. Perhaps I am missing something?

We know that a carbide tool will produce an Ra of about 0.15um in aluminium
What surface roughness you get depends as much on the machining parameters as the cutter. And I am not sure how one measures surface roughness inside a bore of 2.0 mm ID and 30 mm long. Or how you go about trying to improve it.

The grooves at the end are not a problem: you can buy grooving tools for that off the shelf. You can get them custom-made to exactly your dimensions as well. But drilling those 4 holes on a number of pieces will require a jig of some sort if you want reproducibility.

Your 2D drawings are quite adequate. You should somehow specify tolerances, but that can be done in a text list.

I am not knocking what you are doing at all. I am just trying to bridge the gap between your position and that of a CNC machine. I can see both sides.

Where are you located?

Cheers
Roger
(BSc (Hons) Physics; MSc Physics, PhD, 27 yrs with CSIRO as a group leader, and 15 yrs as an independant research consultant)

hi Tim, we are cnc machining manufacturer in China, provide custom cnc parts service, 3d printing, injection molding parts. welcome to contact us( [email protected] ) if you have any more need.