Author Topic: The CNC experiment build.  (Read 50500 times)

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #25 on: February 14, 2015, 05:01:29 PM »
This is actually great reading. damn i want that book and learn german..

Here is some design iterations...
Head is redesigned, the frame is bulked up, the hangout from Z is shortened, the table is made smaller for.. well reasons..

Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

RobWilson

  • Guest
Re: The CNC experiment build.
« Reply #26 on: February 19, 2015, 11:21:04 AM »
Hi Neo ,

 I can see this being a canny project to keep an eye on mate  :thumbup:  ,interesting design  :med:


Rob


Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #27 on: February 19, 2015, 02:11:17 PM »
So i thought i should dumpsom pictures.. of stuff sourced so far.

Start with an image of the last design iteration - can be useful i guess.
And after that first out, 16 pin connectors for encoders, aviation graded, 1.5kW servo motors for driving stuff around.
Granitedevices Argon Servo drivers, these i got for a hefty discount, thank you Tero over at granite.
And lastly a 6i25 + 7i77 + 7i84 MESA card setup, i first bought a parallell kit of this and couldnt find any good info - so thats a waste of money. But this should work.

Then cabinette., Hiwin 30mm rails with heavy load carriages and heavy preload, second 32mm ballscrews with double nuts and a 5mm pitch, and lastly claw connectors as im planning to directdrive this in the first iteration..

I need to get my hands on a 5.5kW Spindle servo motor that runs 3000rpm minimum - if anyone can scrounge up one for a good price point please gimme a shout.
And im still haggling over the BT40 spindle housing and pulleys with the chinese lady.. that drive a hard bargain.
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline awemawson

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 8966
  • Country: gb
  • East Sussex, UK
Re: The CNC experiment build.
« Reply #28 on: February 19, 2015, 02:21:16 PM »
Who's been spending his hard earned cash then?  :ddb:
Andrew Mawson
East Sussex

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #29 on: February 19, 2015, 02:36:12 PM »
Hard earned is a stretch.. most of this is chinese actually.. not the cabinette - cuz that thing weighs a ton..
The most expensive is the Argon drivers.  =)
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #30 on: February 19, 2015, 05:54:52 PM »
Can anyone gimme some input on this;

Roller plunger limit switch or rotary roller limit switch.?
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline awemawson

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 8966
  • Country: gb
  • East Sussex, UK
Re: The CNC experiment build.
« Reply #31 on: February 19, 2015, 06:44:20 PM »
Magnetic proximity sensor - long term reliability - what virtually all recent professional machines use
Andrew Mawson
East Sussex

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
The CNC experiment build.
« Reply #32 on: February 19, 2015, 06:46:20 PM »
You have a link to what im looking for would be much appreciated. :)
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline RotarySMP

  • Sr. Member
  • ****
  • Posts: 463
  • Country: at
Re: The CNC experiment build.
« Reply #33 on: February 21, 2015, 01:41:33 PM »
Hi Neotech,

After I received a copy of Thomas Zietz GradingCurve.XLS, I had a look at what my local hardware store had which would be useful aggregate. From the garden section I got 4x 25 Kg bags as follows:

5-8mm pebbles - http://www.quarzwerke.at/datenblaetter/Quarzkies_ME_5,0_-_8,0.pdf
2-4mm crushed rock - (No data sheet - this stuff is sharp edged crushed rock for spreading on the icy footpath)
0.5-2mm quarz sand for improving your grass - http://www.quarzwerke.at/datenblaetter/Quarzsand_ME_0,5_-_2,0.pdf
0.1-0.3mm quarz sand for grout - http://www.quarzwerke.at/datenblaetter/ME23-Beschichtung.pdf

Cool that the local quarry which supplies them publishes the datasheets with the screen sizes. Total cost of that 100kg was about €13.

I still have some very fine alumina from making a furnace, so that will be my finest filler. It cost €16,50 for 5kg from a pottery supplier.

So without even driving the 50km to the quarry, I could pick up a ton of aggreate for about €130 five minutes from home, although the alumina would get pricy for that amount.

Using the grading spreadsheet, it gives me the following recipe to make up a liter of Polymer concrete:

Grain 5    6,35 mm   100,00   26 %   259 ml   399 gr
                  
Grain 4    3,00 mm   74,09   26 %   263 ml   403 gr
                  
Grain 3   1,00 mm   47,74   20 %   203 ml   350 gr
                  
Grain 2   0,25 mm   27,42   16 %   157 ml   243 gr
                  
Grain 1   0,03 mm   11,74   12 %   117 ml   197 gr
                  
                                100 %1000 ml   1592 gr

The two largest aggregates are wet, so I am drying some in the oven. I read that moisture has a very detrimental effect on E/G strength.

Best regards, Meilleures salutations, Mit freundlichen Grüßen, Cu salutari
Mark
https://www.youtube.com/c/RotarySMP

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #34 on: February 21, 2015, 01:46:09 PM »
 :headbang:  Very cool - i have found most of the grades as well but not the larger ones in a pebble state.. I can find coarse granite what is commonly used in concrete here in Sweden. Its not the rounded floodbed pebbles i had hoped for. I could find most of the aggregate i needed at a local flower and plants store in their outdoor section. When i said i needed about a ton in total of aggregate, they started to look funny at me.  :lol:

You have my email in a message regarding the xls file. =)
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #35 on: February 21, 2015, 02:27:28 PM »
Recipe dont say how much epoxy. is it 10% on weight or 8% ??
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline RotarySMP

  • Sr. Member
  • ****
  • Posts: 463
  • Country: at
Re: The CNC experiment build.
« Reply #36 on: February 21, 2015, 03:25:27 PM »
Because I am using super fine alumina, I'll do a test peice with 10% first.

By the way, nice motion hardware collection you have there.

Mark
Best regards, Meilleures salutations, Mit freundlichen Grüßen, Cu salutari
Mark
https://www.youtube.com/c/RotarySMP

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #37 on: February 21, 2015, 03:26:49 PM »
Thanks, need to start phase to, get steel parts, do casting experiments.. Get a spindle and spindle motor.. the last one is elusive. =)
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline awemawson

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 8966
  • Country: gb
  • East Sussex, UK
Re: The CNC experiment build.
« Reply #38 on: February 21, 2015, 04:03:37 PM »
You have a link to what im looking for would be much appreciated. :)

This is the top(ish) end of the market:

http://www.sick.com/group/EN/home/products/product_portfolio/industrial_sensors/Pages/proximity_sensors.aspx

There are cheaper ones, ebay abounds
Andrew Mawson
East Sussex

Offline Will_D

  • Hero Member
  • *****
  • Posts: 668
  • Country: ie
    • National Homebrew Club of Ireland
Re: The CNC experiment build.
« Reply #39 on: February 21, 2015, 04:42:36 PM »
Now all this talk about graded agregates sparked an unused neuron:

Concrete Springs:
Remember seeing a video that if tou mix GRADED glass spheres (i.e.Start with 2 mm dia spheres, the gap between them is Xmm so you add enough X dia. spheres. Now the gap is Y so you add Y dia spheres. Eventually you stop and add portland cement to lock it all together. Result is a Coil Spring.

Now back up to date:
Quick Google to check if my (only one) functioning neuron was Ok, I came up with thi spaper:

http://www.concrete.ie/downloads/ultra%20high%20strength%20concrete.pdf

NJoy!  :smart:
Engineer and Chemist to the NHC.ie
http://www.nationalhomebrewclub.ie/forum/

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #40 on: February 21, 2015, 04:56:26 PM »
That was so weird reading.  :thumbup:

I wonder how a polymerconcrete spring would act as a milling machine frame.  :poke:
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline Will_D

  • Hero Member
  • *****
  • Posts: 668
  • Country: ie
    • National Homebrew Club of Ireland
Re: The CNC experiment build.
« Reply #41 on: February 21, 2015, 05:08:39 PM »
I wonder how a polymerconcrete spring would act as a milling machine frame.  :poke:
Well it would certainly put a spring in your cut  :Doh:
Engineer and Chemist to the NHC.ie
http://www.nationalhomebrewclub.ie/forum/

Offline dsquire

  • In Memoriam
  • Hero Member
  • *****
  • Posts: 2275
  • Country: ca
  • Kitchener, Ontario, Canada
Re: The CNC experiment build.
« Reply #42 on: February 21, 2015, 05:35:42 PM »
Will_D

That was a very good read. If they keep going they are going to figure out the right proportions to mix that dust to build what we now think is impossible. It is just about scary when you think about it.  :lol: :lol:

Cheers  :beer:

Don
Good, better, best.
Never let it rest,
'til your good is better,
and your better best

Offline RotarySMP

  • Sr. Member
  • ****
  • Posts: 463
  • Country: at
Re: The CNC experiment build.
« Reply #43 on: February 22, 2015, 03:09:46 AM »
In engineering, everything is a spring! Nothing is absolutely rigid. You design for acceptable stiffness for your application.

Mark
Best regards, Meilleures salutations, Mit freundlichen Grüßen, Cu salutari
Mark
https://www.youtube.com/c/RotarySMP

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #44 on: February 22, 2015, 01:57:16 PM »
Just more design iterations..
Added anchors, changed the table, by adding 5mm to the platten of the saddle, fitted the ballnuts and figure out travels and such.
Will need to start making the way covers now to see how many i will need to make.. if they are gonna be thin and short there will be a shitload of them. *grmlbl*

Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline RotarySMP

  • Sr. Member
  • ****
  • Posts: 463
  • Country: at
Re: The CNC experiment build.
« Reply #45 on: February 22, 2015, 03:59:12 PM »
Your x bearing blocks look far too close together.

I just cast a polymer concrete test bar. Added 5g of carbon black from an old printer toner cartridge. Made 1.2kg of E/G nice and black. 10% epoxy was stiff but still sagged why a little vibration. I just tapped the mould on the floor for a few minutes. With a proper vibrator 8% epoxy should be fine.

I mixed in the aggregate from fine to coarse. The first and last were hardest to mix.

Mark
« Last Edit: February 22, 2015, 04:35:11 PM by RotarySMP »
Best regards, Meilleures salutations, Mit freundlichen Grüßen, Cu salutari
Mark
https://www.youtube.com/c/RotarySMP

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #46 on: February 22, 2015, 04:12:03 PM »
Ah well, i cant fit them any other way, its wide heavy load bearing blocks. on a 30mm rail. So it starts to interferce with other stuff when moving around sadly. So its a bit of a compromise.

So the question comes to then i guess, can i conjour up a vibrator for a 300kg casting.. i have a 0.75kW electric motor that maybe could do the trick.. 1450rpm, a slightly offseted weight out of some aluminium roundstock maybe works?!

Got any picks of the brick? =)
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline philf

  • Hero Member
  • *****
  • Posts: 1108
  • Country: gb
Re: The CNC experiment build.
« Reply #47 on: February 22, 2015, 04:20:12 PM »
Your x bearing blocks look far too close together.

Mark

Neotech,

I noticed that too. When working at the ends of the table it's cantilevered. I think I would have the rails attached to the table and the blocks attached to the saddle. To do this you'd probably have to fix the screw to the table and the nut to the saddle. This way the spindle is always centred between the bearing blocks.

Just my 2p worth.

Cheers.

Phil.

Phil Fern
Location: Marple, Cheshire

Offline NeoTech

  • Hero Member
  • *****
  • Posts: 826
  • Country: se
    • Roughedge Hobbyworks
Re: The CNC experiment build.
« Reply #48 on: February 22, 2015, 04:26:26 PM »
Have actually considered that approach.. should prob. take me the time to cad that out as well. Just redone the table like 10 times already so. =)

Would also allow for a smaller saddle with no overhang im guessing.

One of the issues i have is getting thick plate cut and drilled or even get hold of it. Its a bit why i approached the table like that. But i should reconsider that - and see what a slab of castiron cost actually.
Machinery: Optimum D320x920, Optimum BF20L, Aciera F3. -- I have not failed. I've just found 10,000 ways that won't work. http://www.roughedge.se/blogg/

Offline RotarySMP

  • Sr. Member
  • ****
  • Posts: 463
  • Country: at
Re: The CNC experiment build.
« Reply #49 on: February 22, 2015, 04:38:03 PM »
Normally, on the rigid industrial CNC's they put the blocks on the ends of the table and the rails on a fully supporting saddle. That gives the most rigid configuration. The way you have at now, you have the disadvantage of a full length saddle, with the low rigidity of bridgeport saddle.

Mark
Best regards, Meilleures salutations, Mit freundlichen Grüßen, Cu salutari
Mark
https://www.youtube.com/c/RotarySMP