[joeavaerage]

Hi,
lovely designs, but then you'd expect that from Mori.

Are they an especially practical design for a hobbyist to replicate?.

The most expensive part of my build was getting three identical cast iron axis beds casts, then machined and heat treated. That alone cost nearly $10,000NZD.
The axis beds are very simple, and I did this with a view to minimising the cost....and while I did get the cost down......$10,000 it not cheap in my book.

Those Mori designs are vastly more complicated, and there is no way in hell I could cast those for instance, or at least within any sort of budget. To my mind no matter what you believe
are the perceived advantages if you cannot afford to replicate them then what good is it?.

When all said and done CNC is and has to be practical.....who cares what design or fancy material or process you want to use if you can never afford to do it?. My over-arching philosophy
is that 'I can (almost) never afford what I want, so I have to do the best I can within the budget I have....OR go without.'

Craig

[pippin88]

Hi,



I am interested. You say the quote direct was very high......would you care to guess how various AliExpress vendors are selling 'TBI' products as cheap as they are?
Is it, do you suppose that the quote given you was artifically inflated? Or could it be that the AliExpress vendors are selling knock-offs?

I don't know whether you have seen that there are vast numbers of AliExpress vendors selling what look like Hiwin linear rails and cars. They look the same, identical colors, the same part numbers
as genuine Hiwin but are not in fact Hiwin at all?. I wonder whether these vendors are doing something similar with TBI products.

Craig

Yes, I also very much want to know what TBI quoted

Certain brand names have become copied / faked.
Hiwin is one, though you can get genuine Hiwin stuff. I believe that BST Automation Hiwin is genuine, the stuff I got years ago certainly appears to be. BST got me a higher precision grade and preload. The products were appropriately marked.
A lot of the knock offs / fakes have subtle differences from the real thing in terms of markings and texture.

The first tip for those not familiar with AliExpress is to buy from a store that specialises in that area. If you can buy kitchen utensils or underwear from the store, do not buy CNC gear from them.
Stick to stores selling a limited range. And ask questions. If they can give a sensible answer, you are safer.
A genuine reseller should be able to get different preloads and precision grades.

[suspension]

There are many answers here that I just glanced through (on tool surface speeds and structure) because I am trying to finalize ball screws. I have many questions on those which I will post soon. Thank you for the advice.
The debate about the structure made me rethink about what I really need to do. Simply, I want to mill lots of aluminium and soft metals fast with very nice finish. Also it would be nice if I can mill a hard metal like steel or SS once in a while with good finish. When milling hard metals, it can be slow and I can be patient.

With above requirement, I think I may have to reconsider my choice of spindle. What I have chosen is a 24000 RPM, water cooled, motorized ISO30 spindle. I believe this spindle can meet my main requirement of milling soft metals. For steel, we need to run at much slower speeds, but running below half the rated RPM can be a problem. Which means it will have to run at 12K when cutting steel also. Assuming I can buy a higher power spindle (say 5KW?) I may be able to get the torque I need. But is 12K speed too much for hard metals?

Other option is to buy a BT30 spindle cartilage + servo. This has max 12K RPM. I hope this will be ideal for hard metals but will I be able to full fill my main requirement - which is to cut aluminium fast with very nice finishes.
This is ideally I like to see for Aluminium: https://www.youtube.com/watch?v=3YzAl29Ag78&t=132s

[suspension]

Yes, I also very much want to know what TBI quoted

Here is the quote from TBI. And they also confirmed that https://www.tuli-shop.com/ is not an authorized re-seller. (but that does not mean they are fake. It could be that https://www.tuli-shop.com/ are just a re-distributor who purchase from a local authorized re-seller). The quotation include ball nut and they sau the length specified is the screw length.

Model	End Machining	Quantity	Price/Each (EXW Taiwan)	Lead Time
SFAR01605C1DFC5-550-P1-B2	Yes	1	$312.4	4 weeks
SFAR01605C1DGC5-550-P1-B2	Yes	1	$590.5	4 weeks
SFAR01610B1DFC5-550-P1-B2	Yes	1	$318.6	4 weeks
SFAR01610B1DGC5-550-P1-B2	Yes	1	$605.2	4 weeks
SFAR01605C1DFC5-450-P1-B2	Yes	1	$283.8	4 weeks
SFAR01605C1DGC5-450-P1-B2	Yes	1	$580.5	4 weeks
SFAR01610B1DFC5-450-P1-B2	Yes	1	$288.6	4 weeks
SFAR01610B1DGC5-450-P1-B2	Yes	1	$605.4	4 weeks
SFAR02005C1DFC5-450-P1-B2	Yes	1	$305.6	4 weeks
SFAR02005C1DGC5-450-P1-B2	Yes	1	$614.5	4 weeks
SFAR02010C1DFC5-450-P1-B2	Yes	1	$311.8	4 weeks
SFAR02010C1DGC5-450-P1-B2	Yes	1	$620.6	4 weeks

[joeavaerage]

Hi,
this morning I had Johno, my employee, make a steel water pump base plate. I normally would have used my servo spindle (1.8kW @ 3500rpm) but as it turns out coolant has got into the servo and screwed
the encoder, and it won't run. Very pissed off, and my own bloody fault. Looking for a new encoder or more likely a new (second hand) servo.

In order to do this job Johno used an 1/8th Raptor tool (high end Destiny Tools endmill) at 15000rpm with my regular 24000rpm spindle. Everything went well until the bloody coolant blocked up near
the end of the job and the tip of the tool went red hot, takes about three seconds. Bugger!!! The job got done but really the tool is toast now and it was about $27.00USD. It hurts when you only get an hour out
of them, I should have got four to six hours even in mild steel, but when you spin tools overfast in steel they crap out fast....real fast.

I would say go with your 24000rpm spindle. it'll be the one you use and use again and again and again!. That its a poor choice for doing steel is just too bad. If you do al lot of steel then you'l have to come up with another
spindle. Dont be a f******ing fool like me and let coolant splash onto it while not in use and let coolant get into it and wreck it.

Craig

[pippin88]

Thanks for sharing the TBI prices.
They are similar to some others I have been quoted.

600x400 working area is a large mill by the way.
Difficulty, cost, and size of components needed for rigidity go up almost exponentially with increasing size.

[suspension]

Hi,
I would say go with your 24000rpm spindle. it'll be the one you use and use again and again and again!. That its a poor choice for doing steel is just too bad.
Craig

I watched some videos of Tormach 1100 machining steel and Aluminium. The spindle they use use is 10,000 RPM. The aluminium parts also has very nice finish. If I were to use a 24,000 spindle instead of this 10,000, what additional benefits can I get when machining Aluminium?

[joeavaerage]

Hi,

If I were to use a 24,000 spindle instead of this 10,000, what additional benefits can I get when machining Aluminium?

None really. It would mean you can use smaller tools and yet spin them fast enough to be highly effective.

The Chinese manufacture a wide range of these asynchronous spindles, and they are very cheaply priced. Bang for your buck they are hard to beat. Could you do better?...most certainly
you can....but at a cost. Where do you stop?

One of the cheap Chinese spindles, say 1.5kW or 2kW, and a VFD and you're in business. You'll make all sorts of parts and I'll bet you find that you consider it the best buy you ever made.
Once you have several hundred hours under your belt you'll find that you have a clearer view of what it is that you actually need. I'd start cheap and work your way up.

Craig

[suspension]

Hi,

The Chinese manufacture a wide range of these asynchronous spindles, and they are very cheaply priced. Bang for your buck they are hard to beat. Could you do better?...most certainly
you can....but at a cost. Where do you stop?

Craig

From alibaba/express, I should be able to source a BT30 cartridge for less than 600. Lets say a servo + driver + belt + air cylinder cost another 1400. (I got a quote from alibaba supplier for all of these for 1900 USD). So total is 2000-2500 which is similar in price of a motorized 24,000 spindle. I have not evaluated this extensively, but based on the current research I feel both 24000 spindle option and BT30 option would cost similar?

[joeavaerage]

Hi,
well in that case pay your money and take your pick, there is no single best answer.

If I were you I would get a manual change 24000 rpm spindle, a cheap one, say less than $500 and run it for a while. You need to get a feel for what these spindles can do.
If you really like it THEN go for a really good 24000 rpm toolchange model, but you may find that your work is best done at lower speeds, in which case buy that kind of spindle.

It is usually the case that a spindle is the single biggest expense of any machine. In this instance you are uncertain about the exact style of spindle. Maybe try cheap to start with
and allow that to inform your real purchase down the track aways.

As you have no doubt discovered building a CNC machine is not a cheap exercise, quality components are eye wateringly expensive. It makes sense to me to spend big, or at least
adequately on items like linear rails, ballscrews and servos, all things that you absolutely need to be the best you can get and/or afford. Likewise the frame however you design and build it needs top be
the most rigid you can possibly make it. Both of these endeavors will absorb all the cash, time, and energy you have as it is. Defer the final piece of the puzzle, a spindle, for a time yet.

Craig

[peteeng]

Hi Craig - I feel that the machine configuration is really important to get right. Otherwise why go to the trouble of building your own machine? Any config is achievable by a hobbyist just takes time, thought and persistence. The best config for the person is up to them. Often people are not aware of options they are just familiar with the common configs. I think Sus is casting in EG or grout? Not sure but both materials can realise the M1 for instance. Peter

[ardenum2]

Hi Craig - I feel that the machine configuration is really important to get right. Otherwise why go to the trouble of building your own machine? Any config is achievable by a hobbyist just takes time, thought and persistence. The best config for the person is up to them. Often people are not aware of options they are just familiar with the common configs. I think Sus is casting in EG or grout? Not sure but both materials can realise the M1 for instance. Peter

that's how it happens Pete. Most common diy builds are a c frame and a gantry. The reasons for that are simple. People who started off with a pm or a grizzly etc. make a c frame, they go with what they know, then new people see that and also make c frames and round and round it goes. Those that started with cheap routers go for a high walled gantry or a fixed gantry with a moving table. It's really sad to watch.

There's a guy over here that sells ready made frames, I'm sure they will look familiar. If I didn't know any better I'd be all over them.

https://www.ebay.com/itm/134247839234
https://www.ebay.com/itm/133667598594

https://www.youtube.com/watch?v=hfSkHNGPKWs

[peteeng]

Hi Ard - Thats a well executed design and he has made provision for tool changing or a rotary on the side. To achieve 600x400 I'd reduce the 400 to 300 (or make it more square) and have a long bridge, some call it a double column (gantries move, bridges are static in my world), move the 300mm through the bridge. Steel is still the issue. But you can/t have everything. Peter

[ardenum2]

Hi Ard - Thats a well executed design and he has made provision for tool changing or a rotary on the side. To achieve 600x400 I'd reduce the 400 to 300 (or make it more square) and have a long bridge, some call it a double column (gantries move, bridges are static in my world), move the 300mm through the bridge. Steel is still the issue. But you can/t have everything. Peter

that's the beauty about going high speed, dental mills are tiny they can fit on a table but they can mill titanium for implants. Sure its slow as hell but possible.

Here's a few other bases he sells, included with longer bed and extra travel for atc.

[NordicCnc]

Those FS3MG machines are really impressive for their price. They seem very rigid especially with the stiffer Z-axis for belt drive spindles. Also very precise machine, small footprint and large work area.

Skickat från min SM-A528B via Tapatalk

[peteeng]

Hi Sus - Here's my current router build, its Called Frankie. Its footprint is 1200x1200mm its envelope is X740mm Y890mm and Z310mm. It has a 24000rpm 800W water cooled spindle. Its made from plywood and aluminium. Z is all steel. Its a precurser to a UHPC cast mill. I'm happy with the build and most of its aspects. The base and gantry will be cold cast, the saddle maybe. I have sent some material samples to a lab for modulus/strength testing and that will determine the final materials. It cuts aluminium easily so I imagine yours will too. My mill is required because the things I want to make (cnc machine parts, gantries, saddles and bases) are composite; carbon fibre and fibreglass and other materials. The machinists I use for various bits won't let these materials near their mills so I have to build a wet mill suitable. So all the plywood you see in this image will become engineering grout (a type of concrete) or will be epoxy mineral cast. Results available soon... So I don't want to machine steel but I have spent a lot of time looking at the spindles for both and its a hard thing to sort. The 100m/min surface speed is the limiting factor for steel.... Peter

[suspension]

I am still struggling with ball screws!

Earlier I wanted to build a smaller mill than Stef's as I was thinking of keeping it inside my house's 2nd floor (concrete structure). But later decided not to do that due to safety reasons so the machine will stay on the ground floor. So weight is no longer a big issue. I am thinking that the perfect travel sizes would be similar to Syil X7 or Tormach 1100. Lets say its going to be 550 to 600 X 275 to 350 X 400 to 600. This is very similar to Stef's mill. He has used 20mm for X/Y and 25mm ballscrews. To understand the sizes for my machine I tried to use various calculators available (for example from misumi, NSK, etc). These tools are asking following information and I like to get some recommendations.

1. Friction co-efficient of the slide. I am using HIWIN so .0003?
2. Distance (mm) for bucking load and safety factor
3. Distance (mm) for critical speed and safety factor
4. Threading speed (mm/s)
5. Acceleration constant (s)
6. Cycle times per 1 sec (1/min)
7. Work factor

I used some of the values I gathered from datasheets and my existing CNC router and NSK software recommended a 40mm diameter screw for X which should not be the case!

Thanks
Sus

[NordicCnc]

While it is good to calculate and simulate things; I think you are overthinking it. A 20mm ballscrew will be perfect for your build (given the work area) dont stress anymore about the size. What you should decide is the ballscrew accuracy and what kind of travel speed you want. A 5mm pitch screw with 1:1 direct drive, or belt drive. A 10mm pitch screw with 1:1 direct drive or 1:1 or 2:1 belt drive. 400w servos will be enough to move 100kg weight in X/Y with 1:1 drive and 5mm pitch, or 2:1 with 2:1 drive, no problem (personal experience), maybe even more.

A larger diameter ballscrew will heavily affect the servo size you need. The bigger the servo, the more space, higher coat and more power consumption (you need to ensure you have enough to power your whole mill without tripping the fuse in your shop also..). What if you run the mill + a compressor, will it trip your fuse?

20-30mm linear rails + 20mm ballscrews are perfect for your application. Anything bigger will require a huge frame and motors, which becomes increasingly difficult to build and costly.

Skickat från min SM-A528B via Tapatalk

[suspension]

While it is good to calculate and simulate things; I think you are overthinking it. A 20mm ballscrew will be perfect for your build (given the work area) dont stress anymore about the size. What you should decide is the ballscrew accuracy and what kind of travel speed you want. A 5mm pitch screw with 1:1 direct drive, or belt drive. A 10mm pitch screw with 1:1 direct drive or 1:1 or 2:1 belt drive. 400w servos will be enough to move 100kg weight in X/Y with 1:1 drive and 5mm pitch, or 2:1 with 2:1 drive, no problem (personal experience), maybe even more.

A larger diameter ballscrew will heavily affect the servo size you need. The bigger the servo, the more space, higher coat and more power consumption (you need to ensure you have enough to power your whole mill without tripping the fuse in your shop also..). What if you run the mill + a compressor, will it trip your fuse?

20-30mm linear rails + 20mm ballscrews are perfect for your application. Anything bigger will require a huge frame and motors, which becomes increasingly difficult to build and costly.

Skickat från min SM-A528B via Tapatalk

The reason I wanted to do a calculation is to see if I can use a 15mm one for X and possibly Z. The ebay seller has some nice new-other ballscrews that match my travels, but diameters are 15mm. He also has a 20mm one which I am going to use for Y.

[NordicCnc]

I have a 16mm ballscrew on my Z-axis and it works fine. I have a BT30 JGL-100 spindle on it so its about 20-25kg in total (moving part). So far, no issues at all with 5000mm/min rapids (I dont recall the acceleration). My machine is a moving gantry made out of aluminium profiles. 20mm rails on all Y, X and Z. 20mm screws on Y and X. I can do aluminium no problem with an 8mm 1-flute end mill, 25mm depth of cut, 1mm width of cut and 1500mm/min feedrate.

For steel, my machine is not suitable but Ive done some tests and it works as well, them rpm is the problem.

I'd say take the 20mm ballscrews at least for X and Y. For Z, you could do 15mm but I'd do 20mm if you are shooting for a "Stefs" build. Its simply more rigid and rigidity is your friend. Dont cheap out on the motion parts and the frame, those are your core components. Its a struggle to save money in any area when you want to built a mill for steel or a mill which can hog away aluminium with decent precision and repetability.

Skickat från min SM-A528B via Tapatalk