Best way to mill a pocket in Al on a manual mill

[JonParticipant @jon]

Posted by Ian Phillips on 14/11/2014 19:57:50:

Posted by Neil Wyatt on 14/11/2014 19:17:49:

> Careful if using ER collet cutter may drop unknowingly or cut a taper, used to cop us out at work.

I've been caught by this a couple of times. Really pays to lubricate the collet (outside) and tighten as much as you can.

But I'm starting to think of getting a clarkson type screw-in cutter holder.

Neil

Maybe I have been lucky but I have never had a cutter move, trouble with the 'tighten as much as you can technique' is that it depends on the users interpretation. Ian P

No this was after excessively tightening up full force with supplied spanners. By that 6ft 4 beefy build blokes couldn't undo by normal means.

The bearings for the ER have been available for years from Gloster Tooling.

[Steve WithnellParticipant @stevewithnell34426]

I'm well on with the pocket, I bored the centre out, which worked well. I then moved the workpiece over to the mill and set the stops so I had about 0.5mm shy of each finished dimension. Then moving back and forward between the stops to get rid of the rest. Once I'de done half, I then tried chain drilling out the waste (with a slot drill), again working between the stops and this did seem to be more effective than milling it away, mainly less problem with chips and faster progress (remember I'd bored out the majority on the lathe already). I used a 10mm slot drill, but could only get 18.5mm deep, so I finished up the bottom of the pocket with a 12mm 3 flute end mill.

I then changed to a 5mm long series carbide "high helix" 3 flute end mill and have taken two of the sides (between the stops) to the finished dimensions. The finish has turned out really well. The last pass was 0.05mm at 20mm deep, climb milling with some tension on the table locks. Biggest cut was 1.25mm x 1.25mm – 25% the diameter of the cutter. I've got the mill running flat out at just over 2000rpm.

Shouldn't take too long now to finish up the remaining two sides.

After that there is just a load of holes to drill and a few to tap. Two of the holes will be 3/8 x 32 tpi UNEF – all the tapping charts I can find quote 8.85mm tapping drill, but don't quote any figures for thread engagement. How does 8.8mm sound in context of a through tapped hole in this casing? I need the male thread to "bite" into the casing to ensure a good electrical connection.

On the other comment – I've never had a cutter come loose from my collet chuck (ER32) in an MT3 arbor. BUT I've never done anything particularly heavy duty either.

[Neil WyattModerator @neilwyatt]

Perhaps it's my home-made chuck? But things are better with my new one using a ball bearing nut.

Neil

[Ian S CParticipant @iansc]

Steve, the 8.8 mm drill should be ok, are you using a taper tap? If so it will start no trouble.

Ian S C

[MuzzerParticipant @muzzer]

Of course, to get in the pocket properly, you need a set of drums and a degree of skill. And who better than Dennis Chambers to show how!

Merry

[Steve WithnellParticipant @stevewithnell34426]

The pocket turned out really well, hit all the dimensions and the 8.8mm tapping drill worked well, then managed to drill a simple straight through hole 1.5mm out – D'oh! Otherwise the whole thing is coming together really well. I'll tap the duff hole and plug it with a threaded Al rod and loctite, then try again.

So thanks for all the advice!    Must take more care over where I point that drill

 

 

Steve

Edited By Steve Withnell on 20/11/2014 21:40:36

[Steve WithnellParticipant @stevewithnell34426]

Here you go –

This is what the pocket was for – this is a butterworth response bandpass filter, the length of the copper rods is related to centre frequency (in this case 1090MHz ). Still a couple of holes left to drill and tap for the lid and each rod should have a brass 2BA screw to provide a little bit of tuning.

The BNC connectors are the typical "single hole" mounting type which happen to be threaded 3/8 x 32TPI UNEF. Plan was to turn the threaded part of the connector down so it was the same length as the wall thickness. Which worked except that the innards are effectively crimped in place so fall apart as you turn the crimp off. I've then used PTFE as a push fit to replace the original nylon dielectric. This eliminates the need for the crimp. PTFE is a) a far better dielectric at UHF and b) will standup to the heat from soldering the inner conductors to the copper bars. I'll be using soft solder and a small tip on my propane torch to make sure I've plenty of controllable heat. The copper bar should allow me to keep the torch well away from the joint..

Steve

[Neil WyattModerator @neilwyatt]

> a) a far better dielectric at UHF

So it will change the impedance of your connectors

 

very nice job – I thought you were making a gunn oscillator to fir microwave beams at the neighbour's cat.

Neil

Edited By Neil Wyatt on 23/11/2014 13:42:11

[Steve WithnellParticipant @stevewithnell34426]

Just working my way up…

Swapping solid nylon for solid PTFE should cause the CI to move from 50 to 70 ohms, but part of the reason the connector falls in bits when you turn the end off, is that the nylon is in part airspaced (and crudely so), so hopefully the difference isn't that marked…

To be honest I hadn't spotted that – so have been doing some back tracking

Steve

[Steve WithnellParticipant @stevewithnell34426]

I made the first filter using Jason's method, as the pocket was almost square – roughly 5:4. I've just roughed out a second filter where the pocket is 2.2:1, so I would have needed to bore two holes, resetting the work between the two holes, so just milled out the pocket conventionally. It would have easier to bore two holes on the lathe!

Thanks for the tip Jason, definitely the way to go!

Steve

[Author]

Posts