CNC metal machinists (for Stax amp cases) unite?

[jamesmking]

I could not help but notice that the dead tree carcass machinists have a thread on wood machining. But dead trees do not make the best cases for stax amplifiers. So i'm a little surprised that there is no thread dedicated to CNC machining metal. The motivation for this thread was that I turned 50 years old a few weeks ago and for my birthday I decided to buy myself a CNC machine so that I could make my own cases for my stax builds. I decided to buy the foxalien Vasto, https://www.foxalien.com/products/cnc-router-machine-vasto?sca_ref=725103.t6qfXeHAnx based on not much more than I needed to machine 400mm by 400mm and the reviews on youtube are very positive. I'm waiting for delivery so I can't comment on if it is any good and I am a complete CNC newbie so I'm hoping this thread might become popular and eventually contain some advice for people who want to build cases for their amps. 

regards and best wishes

james

 

Edited June 12, 2022 by jamesmking

[Kerry]

Agreed

The foxalien machine looks decent with linear rails which are needed for rigidity when machining metal.  You'll definitely need a real spindle.  I noticed they include a 65mm support, but you really should look for one that supports 80mm and an ER16 collet.  

If you want to get fancy like @naamanf you could get one with an auto tool changer.

Have fun!

[jamesmking]

6 hours ago, Kerry said:

Agreed

The foxalien machine looks decent with linear rails which are needed for rigidity when machining metal.  You'll definitely need a real spindle.  I noticed they include a 65mm support, but you really should look for one that supports 80mm and an ER16 collet.  

If you want to get fancy like @naamanf you could get one with an auto tool changer.

Have fun!

Thank you Kerry. My plan, to begin with, is to modify existing hifi2000 cases and just cut holes for connectors and valve bases, do some engraving and more ventilation slots etc.. . I can't see myself scratch building a case any time soon, but I agree if I go down that route I will need more power and bigger tools. One thing that I am a little concerned about is the lack of software spindle speed control and I don't think there is a tachometer to tell you what the spindle speed actually is. But I guess I could get a hand held tach.

I notice that most high power motors are water cooled. I wonder if I could use my pcs 480mm r(4x 120mm fans and 80mm thick) radiator 🙂

 

Edited June 13, 2022 by jamesmking

[jamesmking]

A few useful basic CNC beginners videos I have found.

 

calibrating axis

setting up a spoil board.

problem solving issues

guide to bits

some speeds and feeds for aluminium and steel - starting at 20min 20secs

 

Edited June 14, 2022 by jamesmking

[jamesmking]

this stuff looks interesting  Cermark Spray for making labels on steel, unfortunately does not work on aluminium- Starts at  25:32.

marking aluminium with a lesser 

 

Edited June 23, 2022 by jamesmking

[jamesmking]

a speed and feed calculator.

https://fswizard.com/

 

[jamesmking]

Well I'm no longer a cnc virgin finished building the machine yesterday. Broke my first end mill today. 2mm carbide endmill trying to mill 27mm long ventilation slots in 3mm thick aluminium... first two slots went ok taking tiny nibbles at 0.2mm depth of cut and a feed slow increasing from between 10mm to 100mm per min.  Then I got a little too ambitious with the feeds going for 200mm feed and 0.4 and then 0.6mm depth of cut. Instant SNAP I guess this is to be expected.... I did hear the sound of the cutting change a little before hand but did not react in time...Post mortem on the end mill shows it got clogged up... guess I need cutting fluid a single flute end mill and more experimentation... I literally had to use plyers to extract the broken end from the aluminium it got welded in.

well just broke a second bit... 100mm feed 0.3 depth of cut is too much for the cheap chinesium 2mm endmill again the end mill bound and clogged up....

need to wait for some (hopefully better) single flute 2mm to arrive....

 

Edited June 25, 2022 by jamesmking

[Kerry]

Congrats  

It's been a long while since I've had any lollipops on my endmills.

I just looked at some of my typical settings.  I'm using 1/8" bits, < .039" (.99mm) depth of cut with a speed of 15" (381mm) / minute for most general cutting.  For a 2mm bit with longer cuts I might dial back the depth to .03" (.75mm) or possibly a bit less (not a typical operation for me).  I run the spindle at 14400 RPM for most milling operations.  Drilling and thread milling is slower (5K - 12K RPM).  YMMV

I have a mister with air, but I typically only use air.

I'm using these carbide single flute upcut bits:

Kyocera - 1/8" Single Flute

Kyocera - 2mm Single Flute

Onsrud - 1/4" Single Flute

They say for plastics, but someone had recommended them to me years ago and I haven't had issues since.

Good luck!

[jamesmking]

19 hours ago, Kerry said:

Congrats  

It's been a long while since I've had any lollipops on my endmills.

I just looked at some of my typical settings.  I'm using 1/8" bits, < .039" (.99mm) depth of cut with a speed of 15" (381mm) / minute for most general cutting.  For a 2mm bit with longer cuts I might dial back the depth to .03" (.75mm) or possibly a bit less (not a typical operation for me).  I run the spindle at 14400 RPM for most milling operations.  Drilling and thread milling is slower (5K - 12K RPM).  YMMV

I have a mister with air, but I typically only use air.

I'm using these carbide single flute upcut bits:

Kyocera - 1/8" Single Flute

Kyocera - 2mm Single Flute

Onsrud - 1/4" Single Flute

They say for plastics, but someone had recommended them to me years ago and I haven't had issues since.

Good luck!

Thank you for the advice and support Kerry,

I suspect the chineseium 2mm end mill is the problem.... rather than single flute it has lots of little burs on it and was garbage at clearing chips and sounded like shit at 10mm min cutting speed. I'm hoping single flute will be better. I have some single flute chinesium 3mm end mills and 6mm but I have not been brave enough to try them yet. Once I have got some experience I will get some good quality mills.

I tried some engraving with a 30degree 0.2mm chinesium engraving bit and this worked much better no breakages and a lot nicer cutting sounds. I did some rate tests and got up to 160mm 0.022mm depth of cut without breaking anything. I found the line quality slightly degrades as the cutting speed goes up so I settled on 100mm min. I have not tramed the mill yet or made a spoil board or levelled it so I was not expecting perfect results but the engraving is not bad.  The lines are a little thick for my liking (I made them 0.4mm wide in the cad) but I managed to cut through the anodizing without an issue at 100mm per min and 0.022mm depth of cut. Part of the e is missing in the volume text, but the text is quite small (which I knew would happen from the cutting simulation. I need a smaller angle engraving tool for the fine detail on the e, so I have ordered some 10degree 0.1mm engraving tools which the simulation shows will fill out the entire e). (sorry its slightly blurry just used my cheap phone could not be bothered to go slr and tripod)

For software I tried fusion360 - gui and user experience is diabolical as a lecturer I can get it for free  but I hated it. I settled on paying for vcarve desktop - it does what I want, is easy to use - I just wish it had adaptive tool paths. For sender software i'm using candle although I will probably go universal gcode sender.

 

I just tried 3.175mm single flute end mill and got chatter and lots of vibration of the aluminium I was cutting when going above 0.1mm DOC at 250mm/min. I suspect this is because:

1. I have not levelled the supplied spoil board,

2 I only used 4 hold cheap down clamps (material 430mm long, 90mm deep and 10mm thick) and I think painters tape + superglue might be better,

3 my machine is missing half the bolts. (carton containing all the bolts and washers split open in shipping and about half of everything must have fallen out of the box on route) 😞

On the other hand I did not break the tool  🙂

So I'm pausing experimentation until the rest of the bolts and a thicker spoil board arrive, then I can level and make more meaningful experiments... 

 

Edited June 27, 2022 by jamesmking

[n_maher]

1. Don't mess around with work holding.  Either get it clamped right or do an adhesive method.  You do not want to experience the carnage of a workpiece coming free.  

2. Similar to #1, I fully support the pause until the machine is built right.  It's kind of insane to be running it otherwise but maybe the missing bolts aren't in key places? (fingers crossed)

3. For anodized aluminum marking you should look at a laser.  I saw the one you posted above but didn't watch the video.  There are lots of them and used in conjunction with your CNC would likely but far better than an engraving bit in terms of speed and flexibility.  The other thing to consider would be a diamond drag bit, but those are very slow.  Some cool possibilities but mega machine time required. 

4. My personal approach is to not push feeds and speeds.  Just about any mistake/error with a CNC is catastrophic to either the material, tool, or worse.  Why risk it to save minutes?  It's not like you're trying to maximize a production rate in a shop.  Attempt patience.  Caveat - I lose mine all the time, YMMV, etc. 

[jamesmking]

10 hours ago, n_maher said:

1. Don't mess around with work holding.  Either get it clamped right or do an adhesive method.  You do not want to experience the carnage of a workpiece coming free.  

2. Similar to #1, I fully support the pause until the machine is built right.  It's kind of insane to be running it otherwise but maybe the missing bolts aren't in key places? (fingers crossed)

3. For anodized aluminum marking you should look at a laser.  I saw the one you posted above but didn't watch the video.  There are lots of them and used in conjunction with your CNC would likely but far better than an engraving bit in terms of speed and flexibility.  The other thing to consider would be a diamond drag bit, but those are very slow.  Some cool possibilities but mega machine time required. 

4. My personal approach is to not push feeds and speeds.  Just about any mistake/error with a CNC is catastrophic to either the material, tool, or worse.  Why risk it to save minutes?  It's not like you're trying to maximize a production rate in a shop.  Attempt patience.  Caveat - I lose mine all the time, YMMV, etc. 

 on further inspection the clamps held the work down ok, but the table is bowed downwards in the middle, hence the vibrations in mid way. I will experiment with the tape and ca glue. I also have purchased a thick spoil board which I will attach very firmly to the bed and level that. But I don't see much point in doing the levelling until I have put in the missing bolts which I cant do without some disassembly anyway. I also need to square the Y axis (dual motor) and square the spindle...

all the axis come pre assembled - no bolts missing there and I did check all the bolts were tight. almost all the missing bolts are to things like limit switches, drag chain rails, thin guards to stop chips jumping onto the axis threads, bolts for the additional spindle holders etc. I would not run the machine if it did not feel the structural parts were not well bolted together. I could stop all the vibrations by putting weight into the middle of the work piece. The machine (other than the standard spoil board) feels really solid.

the laser should arrive in 2 weeks or so. I hope 20W will be powerful enough, but its a free gift from foxalien for the missing bolts and scratches to the controller box....

I don't mind taking things slowly including the speeds and feeds i'm in no hurry - I have a lot of learning to do and a lot of experience to get. I know i'm going to break bits, ruin some work etc.. and im ok with that, hence the cheap bits and working a scrap materials.

I was fascinated by wood work and metal work when at secondary school but I ended up studying computer science and university. I like the idea of making things, (I guess it was having lots of lego when I was a child) - and like the idea of making better looking front, top and back panels for my stax amps, so i'm viewing the entire thing as a learning experience a to keep me occupied now I have built all the stax amplifiers I wanted to.

 

 

 

Edited June 28, 2022 by jamesmking

[jamesmking]

I got bored waiting for the replacement bolts to arrive and so purchased my own and finished installing the missing bolts.

Here is my first CNC project:

Making my own Stax 5 pin sockets using ptfe

design in vectric vcarve desktop:

 

making some (ptfe) chips

and no router bits broken...

final result after deburring and running the drill holes all the way through:

overall I am happy with the result. The tabs holding the piece into the material could have been smaller and after test fitting the pins the holes should have been 4.6mm rather than 4.5mm but I don't have a 4.6mm drill yet. other than that I have a working Stax socket 🙂

This first version uses 3 tools: 3.175mm end mill, 3mm drill and 4.5mm drill.... its slow and boring changing tools, so I remade the tool paths only using a 3.175mm end mill, ran the code and found the drill holes were smaller than expected. I measured the 3.175mm end mill... 3.13mm... hmm... so I updated the tool data base with the correct diameter and re-ran the cut again.... I realise now I have to measure all my tools and not rely on the size they claim.

I could not find any metal m3 nuts that fit the threads but are small enough not to touch each other so I used nylon nuts instead. I tested the insulation between the closest pins pairs to each other and got 300Gohm at 5.5Kv.

Edited July 5, 2022 by jamesmking

[jamesmking]

well the missing bolts arrived and the free 20W laser... so I had some fun trying to burn front panel markings.

After a few experiments I realised the default settings in light burn are insane...6000mm/min movement speed no way! and 0.1mm between steps does not provide good resolution and gives a streak effect on text and slightly jagged lines... I improved by changing the settings to two passes at 90% power, 0.05mm stepover and 1000mm/min. This  gives a crisp result (crisper than my camera phone can show).

The result is not a shiny as engraving with an actual 10degree 0.1mm engraving tool, so I will experiment a bit more. I guess I am still not quite getting all the way through the anodizing.

Edited July 5, 2022 by jamesmking

[JoaMat]

Very nice indeed! 

[Kerry]

For laser etching anodized aluminum, you typically only burn out the dye from the anodizing. Once you burn into the aluminum, it typically turns the aluminum brown. 

Try using less power to see if you can get it a little whiter, but it looks very good. 

[audiostar]

Kerry, can you recommend a laser head for etching anodized aluminium? 

James, which laser are you using to engrave the above? Couldn't find but maybe I missed it..

PS: soon to join the CNC machinist club with a DIY CNC and a 3hp ATC motor.

[jamesmking]

56 minutes ago, audiostar said:

Kerry, can you recommend a laser head for etching anodized aluminium? 

James, which laser are you using to engrave the above? Couldn't find but maybe I missed it..

PS: soon to join the CNC machinist club with a DIY CNC and a 3hp ATC motor.

i have the foxalien 20W which is a single diode with 5W actual optical output.

https://www.foxalien.com/en-gb/collections/cnc-accessories/products/foxalien-20w-fixed-focus-blue-laser-module-kit

fox alien also sell a 40W which is dual diode and about 10W optical output. You need a cnc controller which has a laser output. The fox alien is 3 pin - 0V dc power ( I think 12 or 24V) and a pwm for laser power control. Im using lightburn software which works well and has some laser test and calibration templates built in. General consensus on the internet seems to be that if you run a diode laser at full output you wear out the diode very quickly. So im tending to go slower movement and lowish power. The cnc controller board takes the spindle speed control commands and converts them to pwm signals for the laser output power control. On the fox alien vasto this requires setting a slide switch on the back of the controller to laser mode AND sending a G code command to go into laser mode. The slide switch disables the 48VDC motor output via a relay, disables rpm control via the speed potentiometer and enables  the 3 pin laser output. The g code command tells the cnc controller to convert g code rpm commands to a pwm laser power control signal. 

I have also changed the stock 400W 48VDC air cooled spindle for a 65mm diameter er11 1.5Kw water cooled and a vfd. I would have preferred 80mm diameter er16 as suggested in this thread but I don't have an 80mm mounting bracket and the Z axis of my cnc does not look like it will take a bracket for 80mm spindles. Initial results are promising compared to the 400W. I still need to tweak the vfd settings for accel, breaking and torque. The stock spindle is cheap and nasty. I took it apart and it has two small bearings and the case is thin and the entire thing is held together with two long bolts. Its not very rigid at all and certainly seems to be one of the weaknesses of the machine along with the spoil board.. The 1.5Kw feels far more rigid and is 3 phase 220VAC.

 

   

Edited August 14, 2022 by jamesmking

[Kerry]

1 hour ago, audiostar said:

Kerry, can you recommend a laser head for etching anodized aluminium? 

PS: soon to join the CNC machinist club with a DIY CNC and a 3hp ATC motor.

I send my stuff out for the moment. They use a fiber laser. Very nice stuff. 

Congrats and good luck. 

[audiostar]

Thanks, James and Kerry. I can do that additional control for the laser, no problem. Haven't researched much about mounting a laser etching head and what it would take. 

Looks like I am going to use that ATC spindle with a Hitachi sensorless vector VFD.
And this would be a cheaper alternative.

Edited August 14, 2022 by audiostar

[jamesmking]

7 hours ago, audiostar said:

Thanks, James and Kerry. I can do that additional control for the laser, no problem. Haven't researched much about mounting a laser etching head and what it would take. 

Looks like I am going to use that ATC spindle with a Hitachi sensorless vector VFD.
And this would be a cheaper alternative.

the diode lasers tend to come in rectangular aluminium blocks which have fins and act as heatsinks with a fan blowing downwards through the heatsink and a small controller on the top.

The standard 400W spindle mount on my machine has 4 small notches in it so it can also mount the laser just by holding on to the corners. You don't need a high clamping force. So you could probably 3d print something if necessary. The hole is 52mm diameter.

 

 

If you are planning to engrave or cut wood then additional air flow is recommended to clear the smoke (which would otherwise scatter the laser light) for aluminium engraving there seems to be no smoke.

wow that spindle with quick change costs more than my entire cnc machine with the upgrades I have done so far....  My vfd and spindle is cheap chinesium, the only issues going that route is that there is a general lack of documentation and the seller could not help so i had to guess, experiment and look at you tube videos of similar setups for the vfd settings etc to get things to work. The vfd did come with a manual which was mostly understandable but although I purchased the spindle and vfd as a bundle the spindle had zero documentation and no information on how to set it up the vfd for it... the seller could not even tell me the diameter of the tubing required to fit the compression fittings on the spindle for the water cooling... so I had to guess and fortunately got it right first time. I also had to guess how many poles the motor had and use a tachometer to verify.... So if you are going chinesium I would ask the seller some questions about the tube diameters, motor poles, vfd settings etc and see if they have a clue before purchasing from them...

Also I have found a general lack of attention to detail going chinesium, for example the vfd has three terminals marked earth. Two are totally  isolated from everything including each other. Only one actually connects to the earth pin of the mains plug and the spindle takes a 4 pin plug, 3 for the power phases and you would assume the 4th pin is to ground the spindle body, nope no continuity to the body. Also on the vfd it has a temperature readout for the driver mostfets... always says 0C googling says this is normal temp never implemented....

 

 

Edited August 15, 2022 by jamesmking

[Craig Sawyers]

Anyone using laser engraving with a 20W laser needs to be exceptionally careful. That is 2000 times eye safe, and is a Class IV laser https://www.lasersafetyfacts.com/laserclasses.html . The raw beam will drill a hole in your retina, causing explosive boiling of whatever humor is next to the retina.

Assuming that you aren't going to get the raw beam in your eye, scattered light from the focus is also far above eye safe.

So - you absolutely need to wear safety goggles. Not Amazon hokey ones - but serious ones from say Thorlabs https://www.thorlabs.com/ , and expect to pay $200.

Then interlock the door to prevent anyone in your household inadvertently walking in and getting eye damage.

I spent many years working with lasers in this class, and wearing googles was mandatory as was interlocked doors. During laser safety training (also mandatory) an ex-Nam veteran said that nothing in active service remotely compared with the horror of looking at the world through his blood-filled eyeball.

Nuff said.

[jamesmking]

4 hours ago, Craig Sawyers said:

Anyone using laser engraving with a 20W laser needs to be exceptionally careful. That is 2000 times eye safe, and is a Class IV laser https://www.lasersafetyfacts.com/laserclasses.html . The raw beam will drill a hole in your retina, causing explosive boiling of whatever humor is next to the retina.

Assuming that you aren't going to get the raw beam in your eye, scattered light from the focus is also far above eye safe.

So - you absolutely need to wear safety goggles. Not Amazon hokey ones - but serious ones from say Thorlabs https://www.thorlabs.com/ , and expect to pay $200.

Then interlock the door to prevent anyone in your household inadvertently walking in and getting eye damage.

I spent many years working with lasers in this class, and wearing googles was mandatory as was interlocked doors. During laser safety training (also mandatory) an ex-Nam veteran said that nothing in active service remotely compared with the horror of looking at the world through his blood-filled eyeball.

Nuff said.

my solution to have solid brick between me and the laser. I run the laser in one room with the room door closed and control/monitor it from the adjoining room using a cheap web camera. I don't put any faith in it or the goggles that came with the laser....

 

 

[Craig Sawyers]

5 hours ago, jamesmking said:

my solution to have solid brick between me and the laser. I run the laser in one room with the room door closed and control/monitor it from the adjoining room using a cheap web camera. I don't put any faith in it or the goggles that came with the laser....

 

 

That is a good solution! But put an interlock on the door anyway. Costs nothing, and is a belt and braces precaution.

My experience is with pulsed IR lasers, which of course you cannot see. Had to be careful to keep fingernails out of the beam, because that particular laser would blow a fingernail off. Frequency doubled into the green it used to attract flies and wasps, which would explode spectacularly.

An interesting calculation was that a single pulse from that laser, if divided up evenly and without losses, was enough to cause eye damage to the population of London.

It ran at 30 pulses per second.

[jamesmking]

I'm starting to get the feeds and speeds dialled in now I have a 1.5Kw water-cooled spindle and some good spoil board bed rigidity, properly levelled and trammed the machine. I'm slowly gaining confidence and beginning to believe the reviews that said the Vasto can cut aluminium...

I managed to cut ventilation slots in aluminium using a 2mm diameter single flute DLC end mill with 300mm/min feed rate, depth of cut 1.6mm, 50% overlap, 18000rpm.  The test slots were 160mm long, 3mm deep, I tried 2mm, 3mm and 4mm widths.... surface finish is like a mirror, no burs and no chips welding themselves to anything. No cutting fluid was used and there was no finishing pass:

Thats so better than my first attempts with the 400W motor and a corn cob end mill which produced garbage surface finish at much lower surface speeds, much lower depth of cut and destroyed the end mill very quickly:

 

 

Edited August 24, 2022 by jamesmking

[audiostar]

Now that looks like the real thing already! Congrats, James!