Which bits to use for PCB milling?

[evalon]

Hi again,

@Craig: Thanks once more for your feedback & input. However, I realize that I need to take some days off from my CNC mill considerations - so as to fully give attention to my other work.

Cheers,

Jesper

[skrubol]

If you aren't going belts, from what I've read, Oldham couplings are generally the best way to go. Cheaper than bellows, multi-part (makes assembly/disassembly easier,) and low to no backlash. Lovejoy couplings are similar, but seem to have a bit more backlash than Oldham. Oldham couplings also fail open (overtorque failure will decouple the motor,) whereas lovejoy do not.

Automation overstock has some great deals on ground screws if something they have in stock fits your needs. Sometimes gems show up on Ebay as well (I found a new NSK ground 36mm 950mm travel screw for $100..)

[joeavaerage]

Hi skrubol,
I use a Lovejoy coupler on my spindle motor with good succsess. You are right I don't think I'd use them as an axis coupler. Despite the torsional flex
I still use the aluminum spiral-cut-slot version, they work well enough and cheap. Better, if I have a real 'moment' they twist off and shear, like a fuse.

Automation Overstock, surplus automation equipment, overstock automation devices, manufacturer overruns, linear bearings, variable speed drives, circuit breakers, contactors, starters, relays
I often look, I haven't seen any at that price but it is a good site and is all new.

Craig

[evalon]

Hi All,

Hope life is well where you are ...

It has been a while so I have been thinking that I would like to give an update on how things are progressing.

Probably the most significant thing to happen was that I sold my other CNC so that the field of options now became more open.

Also, giving considerations some time to "align themselves" I arrived at a somewhat different need specification than I described the last time I posted here. The main difference is that instead of making a Bed Mill type CNC I have instead decided to make a "fixed gantry" type CNC ... It can be seen on the attached screendump. To me the advantage of this design is that it can be very rigid and also allow for a quite large milling cube: It will be 470mm*330mm*150.

In this process I would say that life smiled at me ... As it is I decided to drive 450 kms to visit an importer of cast iron materials, and while being there I found/was offered some 32*64 mm G40 quality cast iron bars - that were pre-milled - at a very attractive price. So while camping in the evening after the first visit to this company I re-drew the CNC and decided to buy these cast iron rods. Additionally, this importer was so very generous to give me appr. 40 kgs of left-over G40 quality cast iron bars and "square solids" so that I now will have a 270*270x20mm spoilboard and motor mount plate - also in G40 quality.

The down-side to this is that there will be a lot of drilling and thread making - and also that the CNC's weight will be just about 150 kgs. Which is much more than I had aimed for ... but on the other hand I reckon this CNC mill will be very rigid and probably even low-noise due to the damping of the G40 material (which would be great).

Additional to the cast iron I have also bought some THK HSR15 & THK SR15 linear rails. This offer came up and although I assume that HSR15s would have been the preferred choice for all rails I decided to compromise a bit here. The HSR15 will be for the Z-axis while the SR15s will be for the x & y axes.

Another compromise I decided to make was to buy some (genuine) C7 1605 ball screws. Thinking about it and searching the internet I found out that - although not cheap - C5 or C3 quality ball screws can be bought at a specified length. Thus, if I find out that I need this higher precision I can replace the ball screws at a later point in time.

While on the subject of ball-screws: If someone here is interested in C2->C3 2005 ball screws at a very reasonable price I made contact with a seller having such used ball screws available. However, although I reckon the ball screws are fine, my experience was that the communication with this seller can be a bit "challenging" ... but, well would like to mention this option if someone is interested ...

For the steppers I have decided to stay with the DM542s combined with some 425 oz/in step motors.

One decision still a bit "up in the air" is which Milling motor to go with. I have found this one on ebay:

https://www.ebay.de/itm/CNC-Square1-...m/263123919037

which combines low TIR (2um) with high weight (6 kgs) and an ER20 collet. One (?) potential down side is that it likely hasn't got much torque at low rpms ... but I will go with this one unless someone here has a better suggestion. I consider this motor to be a versatile precision motor where the main shortcoming is that it won't be comfortable using wide diameter end mills in hard materials ...

So, things have turned out a bit different than I aimed for. But I reckon that I will eventually be more satisfied with having a CNC that can work with almost all materials (and given its rigidity I would expect it to also be capable of milling steels with low-diameter end mills) instead of making a CNC that will do "most-but-not-all" materials. And then, in the meantime a lot of work to be done to actually make it ..

Cheers from the other side of the world,

Jesper

[joeavaerage]

Hi Jesper,
that looks good. I see on the advertising that accompanies the spindle that the torque of the motor is about 0.8Nm
down around 6000-9000 rpm.

With coolant and light cuts that will do steel at 6000 rpm and a 6mm coated carbide tool. At 9000 rpm you will want to be down to
3mm diameter or the surface speed will be so high that the tool will burn up. Flood coolant will still be required.

No doubt you will have noted how expensive high torque spindle motors are. I encountered the exact same thing. I ended up with two
spindles, a high speed low torque unit much like yours but only 750W and then I made another spindle 3500 rpm @6.2Nm for steel.
Of course it is possible to buy one which would do both but cost a fortune. I recommend the approach you are taking....you will end up
with the MOST functionality for a given sum of money.

Craig

[evalon]

Hi Craig,

- thanks very much for once more replying to a question of mine ...

Well, I am pretty pleased with the CNC concept I have ended up with. As it is I did hours after hours of 3D modeling with a combination of aluminum and cast iron only to find out that if I wanted to save quite a bit of weight, yet still wanted rigidity, it would cost a small fortune in aluminum - and not really be good. So eventually, when the offer for the G40 cast iron bars appeared, this seemed like just what I had been looking for. I hope it will turn out fine or even great.

About the spindle motor I just happened to see that the same seller offers a 2.2 kW version which has 1.17 Nm at the same speeds (~+50% torque) and not much of a price difference ... I am not yet familiar with the calculations for this - but do you think it would allow me to e.g. use an 8mm carbide tool at 4000 rpm? Or doesn't it really make a difference in practice?

And then there's another topic I have been pondering ... When choosing bolts & the threaded rods to go through the bars on the right and the left bottom sides of the CNC I wonder what is a sensible choice ... ? Apparently there are different steel qualities (4.6, 8.8, 10.9, 12.9) and just briefly considering I would choose the harder/stiffer versions BUT if the CNC vibrations cause some parts to undergo permanent deformation - even on a small scale - I'd rather have this happen to the bolts than the cast iron bars. Possibly meaning a less hard/stiff steel quality (all bolts will have washers between the bolt and the cast iron bars) ... Might you have any experience with this?

Cheers & have a good day (ouchhh - we are heading towards winter and you are heading towards spring/summer ... sounds attractive ;-))

Jesper

[RCaffin]

8.8 would be about optimal.
And affordable.

Cheers
Roger

[evalon]

Thanks Roger - will keep it there ...

Cheers to you as well

Jesper

[joeavaerage]

Hi,

but do you think it would allow me to e.g. use an 8mm carbide tool at 4000 rpm?

I suspect not. Firstly a 8mm tool will require nearly all the torque available for all but the lightest cuts in steel.
That will mean running your spindle motor at or near rated current and at that low speed it will overheat if not water cooled.

If you really want torque at slow speed then you are up for a multipole motor designed for the job. If you then
wish to maintain high rpm for PCB's say then you will require a VFD that can get to 800Hz or more. All of a sudden you
have gone well outside of hobby parts and you can expect to pay five or ten times what you are looking at currently.

This is an example of what I mean. I have a air-cooled 750W spindle from this same company and it is good. I wish I could
afford something like this:
https://www.mechatron-gmbh.de/filead...HFP-M-8022.pdf
It has two pole pair per phase and note how the torque is available right down to 3000rpm but note also that to get to 24000rpm
you'll need an 800Hz VFD.
The next step up is this one:
https://www.mechatron-gmbh.de/filead...FP-M-12040.pdf
But this unit requires 400VAC, it has 3 pole pair per phase and can do over 4Nm at 3000 rpm but you'll need a 900Hz
VFD just to get to 18000 rpm......and who knows how much its worth!

As far as bolts go they need to be the best you can get. The rigidity of your machine is dependent on bolted joints, the
cast iron really offers you nothing in terms of rigidity. As Roger has suggestes grade 8.8 is a good compromise, availability
and cost. I personally would be going for grade 12.9, I call them Allen Head Cap Screws and they are TOUGH!
I suspect that the ultimate strength and rigidity that you can achieve will depend on bolted connections so the tensile
strength of the bolt (8.8 or 12.9) is likely to be stronger than the cast iron threads ie you'll pull the thread out BEFORE you reach the
tensile strength of the bolt.

Craig

[RCaffin]

Cast iron eh?
I suspect I would probably go for matching HiTens nuts rather than rely on the threads in cast iron.
My 2c
Cheers
Roger

[joeavaerage]

Hi Roger,
looking at the drawing Jesper supplied in a lot of cases through bolting with nuts is not possible.
If he uses sharp taps and is prepared to tap to say 3X-4X diameter deep then the threads should hold against any reasonable
bolt tension.

Craig

[RCaffin]

a lot of cases through bolting with nuts is not possible
Looking closer - yeah, I can see that.
Oh well ...

Cheers
Roger
PS: tapping cast iron is MESSY!
But so is drilling it.

[joeavaerage]

Hi Roger,
when I made my mill went to the scrap yard and bought some cast iron elevator weights and milled the
axis beds out of them. As it turns out the quality of the cast iron was pretty poor, I mean in practice all they have to
be is heavy, are they certainly were that! The casting, particularly the surface had lots of crusty, hard inclusions.
It took me a while to work out the correct strategy for peeling the crud off. Thereafter it was good.

Machining cast iron is a bit messy with all the flakes of graphite through it all, but otherwise entirely possible, fun even.
Just as well, I ended up with about 200kg of cast iron swarf by the time I was done.

I suspect Jesper's material will be much better and sounds like it has been dimensionally milled when it was cast.

There are one or two manufacturers of continuous cast iron sections. One division of Tata Steel were offering sections
up to 300mm wide and 100 through 300mm deep in 3m lengths. I would have had to ship it in from Europe and would therefore
have been expensive.

Turns out that if I were to make my own patterns I could get them cast locally for similar money to importing blank sections.
I will consider the matter a lot more carefully should I decide to build another or substantially rebuild what I've got.

Craig

[Zorbit]

Worth remembering - bolts are for clamping, dowels are for locating. Reamed holes and ground dowels make good joints.

[RCaffin]

Hi Joe

You at least (ie USA) still have a manufacturing industry, or the remains of one. Here in Oz we simply don't. It all got sold off ages ago leaving us with little but mining. We read about people in the USA finding bulk ingots of CI and other goodies at local merchants etc: we have none of that. You might think that means that second hand machine tools should be cheap, but no. The second hand vendors continue to expect to sell old 1940s machinery at high prices. I suspect some of them must have big dusty warehouses. (I have been to a few.)

A mate of mine did his apprenticeship in the naval dockyards here (LONG ago). They had a lathe big enough to machine the drive shaft for a destroyer. Unique in Australia. But the gov't prefers to buy from overseas now (with appropriate kickbacks), so the lathe (and the workshop) were scrapped. This HUGE lathe went for $0: the buyer had to chop it up to get the bits out the door for scrap. That was the only way the dockyard could get rid of it.

Sigh.

Cheers (or otherwise)
Roger

[joeavaerage]

Hi Roger,
I'm in NZ and we've never had a huge manufacturing sector, however what we do have has weathered the ups and downs and is still
with us, within the exception of those techniques/skills/materials that have fallen out of favour.

Remember we had a major earthquake here and quite a few (hundreds) of elevator weights went in the scrap. I got three, about 1000mm x 220mm x110mm
for $90NZD each, roughly scrap value.

When I made my comment about getting mill beds cast I was thinking particularly of a craftsman foundaryman operating a small
business in Christchurch. I came into contact with him when he bought his plasma machine in for repair. Hell of a nice bloke and been in the foundary
business since leaving school 40 years ago. His specialty is non-ferrous but can do up to 65 kg in iron....grey, white, SG whatever you want.
Getting something cast is never cheap but he is exactly the sort of bloke that I would be happy to see profit from my money.

Craig

[RCaffin]

Hi Joe

OOPS! My apologies! NZ.

Elevator weights ... cute! And a craftsman foundaryman - nice.

Cheers
Roger

[evalon]

Hi All ... this is just briefly as I am about to go out milling the cast iron bars ... An alternative to bolting the CI bars could be to glue them with Araldite 2014 which is a known good epoxy for iron, steels etc. ... It would give less strength per area but a much larger connecting area as the whole contact surface on each bar would be used ... Any thoughts here?

Cheers,

Jesper

[RCaffin]

I doubt the bonds will last. The surface of CI has a lot of loose carbon dust at the micro level, and there just won't be enough strength to withstand constant vibration.
On the other hand, Araldite PLUS bolts might be fine.
My 2c
Cheers
Roger

[evalon]

Hi Roger ... Thanks for your prompt feedback ... Now having milled the CI I am also getting a bit cautions regarding using epoxy on CI ... And regarding combining the Araldite and bolts it apparently doesn't really work in practice. The epoxy needs a 0.05 to 0.1 mm layer for maximum strength and it isn't really supposed to be cured under pressure. So although an attractive idea it apparently won't work in practice.

Cheerio

Jesper