Timescapes - Digital Timelapse Discussion

CNC Router Build
Page 11 of 11

Author:  DISPLACEMENT 1 [ Thu Jan 04, 2018 2:46 am ]
Post subject:  Re: CNC Router Build

Hi Kit you seem to be a the point i was a month ago and you have my sympathy. But it does get better i am quite happy with my CNC now and i know what i need to do to to improve the little niggles .
As for your torque problem could ya just keep the stepper running slow and gear up the ballscrew with some timing pulleys and belts . That way ya can tune the speed/ torque to the sweet spot . I reckon you could get 2000mm/m no problem even with your 2mm lead screw. But 1610 ballscrews would be better ,10mm pitch tend to be a bit more expensive so i would get the 5mm pitch and do a bit of gearing . So ya save a bit on money and ya have the flexibility to changing the torque band .
I will do this eventually on my X and Y as well as fitting the single Nema34 to the X axis .
Did some ali cutting the other day for day job

[flickr]Image2018-01-03 11.26.40 by D 1, on Flickr[/flickr]
Quite happy with this and noise levels although it doesn't sound it on video i know but outside shed ya cant here it at all.
Speeds and feed very leisurely 5000rpm and 450mm/m but that seemed to be the quietest combination. I know i could go a lot faster but as long as the relationship between feed and spindle is correct and the the tool isn't getting hot your top speed is irrelevant. In manufacturing i suppose speed is a issue but as i said not really a problem for me . I also think theirs a bit a Youtube machismo about how fast ya can get your CNC router to run . A case of my red button is bigger than yours :D

Author:  kruppTown [ Thu Jan 04, 2018 8:21 am ]
Post subject:  the stiffness of a mill will be a 128 times higher


maybe that helps for higher speed?


Author:  DISPLACEMENT 1 [ Thu Jan 04, 2018 2:13 pm ]
Post subject:  Re: CNC Router Build

He his a clever fella always learn something watching his videos. Afraid large diameter tools doesn't apply to me most of the stuff i do is quite small with tight radius so biggest tool i prob ever use is 6mm and i do need to go down to 1mm occasionally . Also because i am running my spindle a such low speeds i think it would struggle with larger cutters .
I could go fast if i wanted to but it does get to noisy and the neighbours wouldn't like it :(
Ps Kit i think a 3 to 1 gear ratio would get your 2mm leadscrew going at a decent speed and the stepper should still have enough torque me thinks .

Author:  edward [ Thu Jan 04, 2018 4:03 pm ]
Post subject:  Re: CNC Router Build

Glad to see it cutting ali at last!

I am no expert, so this is just my instinctive opinion from what I see:

Isn't it vibrating a a bit?

You could probably reduce the tool stickout by a lot?
Your wooden base may be vibrating too, is it resting on something solid enough? Or maybe it is the part, the way it is attached.

The chip size seems a little small to me, it is almost pleading to plunge deeper and then move onward? But of course you need the torque for that and 5000 is probably not enough. Push it to 10000 maybe?

I've been scanning through YouTube to see what others are doing and I must say the majority of DIY router videos are not too good, but you have a few fantastic ones with relatively low specs routers, so it can be done.

Here is a video of something similar to yours (far from perfect) but listen to the sound of cutting after it plunges. You can hear the chatter when plunging down to depth (too rough due to too sudden a plunge) but once at the right depth, it moves along relatively nicely. He should also reduce the stickout.

I'll probably have similar or worse problems when I build mine.

Author:  DISPLACEMENT 1 [ Fri Jan 05, 2018 2:27 am ]
Post subject:  Re: CNC Router Build

Hi Edward yes i would say theirs some vibration you can see the chips vibrating around on the work piece . But i think this is the work piece moving its only screwed to a bit of MDF and it also close to cutting through in my vid you can hear it begging the tab cut .
So when the ali bed is fitted everything will be a lot more solid on the work piece front .As for speeds chips and everything my test is touch the tool when ya finished if ya can hold it comfortably then ya doing well and i can
But that is it am afraid it cuts to a standard i need and its not loud basically going faster makes it louder . Believe me i have tried its hard to tell from peoples videos what the actual noise levels are but when ya in a shed with a machine it quickly becomes apparent .
Well thats that i think this forum going to be even less active now :(

Author:  Kitwn [ Fri Jan 05, 2018 4:45 pm ]
Post subject:  Re: CNC Router Build

Wow! This has been an epic ride over 11 pages. So much useful information and exchange of knowledge and ideas. How will we cope without a D1 thread to read and write of an evening? We definitely need a thread just to chat about what we're all up to. I think we should call it "DISPLACEMENT 1's Front Room."

Getting back to the exchange of knowledge, I've finally found some useful stuff on stepper motors in a thread from 2008 on the MYCNCUK forum. It's a bit mathematical but actually tells you what you want to know about size and speed of motors required for a job. There's a spreadsheet to do all the maths for you which shows up some very interesting stuff. Most notable is how rapidly the torque from a motor drops away as the speed rises above the 'corner' speed, which is about 450rpm for an average NEMA23. I've been asking mine for 1500rpm and more so I think I see the problem! The spreadsheet also includes a calculation of max rotational speed for leadscrews. For a 1605 ballscrew 1000mm long this is 774rpm which equates to 3800 mm/min. I'd quite like to be able to rapid at up to 6000mm/mim, therefore I might be going back to a belt-driven router!

http://www.mycncuk.com/threads/1524-Wha ... -do-I-need


Author:  DISPLACEMENT 1 [ Sat Jan 06, 2018 2:22 am ]
Post subject:  Re: CNC Router Build

Hi Kit yes it would be good to keep swapping ideas bit i really have run out of steam on this one . I can feel summer on the way ( well not really its quite miserable at mo ) and i need to get on with my camper conversion :D
Both you and Edward are on with interesting builds and i would love to see how they turn out . Maybe its time to start a new thread for just general waffling . So if anybody would like to do that i am sure i will still put in my two penneth .
But yes 11 pages is enough i think it gets to the point when you start repeating things .The lathe situation was a close one if you go back in this thread Chardie their was a page or two on the subject .
Anyways good luck with ya builds everybody . :D
Whoops just checked this thread it must have been another thread on this forum where we talked about lathes. My brain not as as sharp as it used to be :(

Author:  Kitwn [ Sat Jan 06, 2018 5:06 am ]
Post subject:  Re: CNC Router Build

It's very definitely summer here, 46 degrees (114 in old money) this afternoon. It's 9pm and still 36 in the shed. No camping for us in this weather, all-night air-conditioning is a must for an ex-pom if you want to get a good night's kip.

Still plenty to talk about. I've discovered a whole pile of info from the MYCNCUK website including a surprisingly affordable yet recommended supplier of leadscrews on Aliexpress.com. I've never used Aliexpress before but this supplier is mentioned several times as reliable so I'm beginning to settle my plans for what to do about my recent disappointing upgrade after a day or two of wildly changing my mind!

https://www.aliexpress.com/store/produc ... 37447tzxGB


Author:  edward [ Sat Jan 06, 2018 1:25 pm ]
Post subject:  Re: CNC Router Build


I simply don't know how you can stand that kind of heat, the poor air-con compressors must be working hard to keep you alive!

The MYCNCUK forum is one I visit frequently and post to sometimes. It is very good to learn everything CNC and they tend to be very helpful.

I've ordered from Aliexpress a few times, no problem, it's like ordering from Amazon really. I recently ordered some super duper ballscrews with bespoke machined ends costing about a grand in total, and I was therefore a bit nervous, but the guys I dealt with where fantastic, polite, efficient and knew their stuff. Everything arrived as planned.

The link you sent is a popular one with CNC guys, order with confidence. 10mm pitch is the norm for wood. I have 5mm because I will be doing ali mainly, my ballscrew revs will be around 1000, about the limit according to your link. I don't think I have any issues with torque on my Nema 3.1 Nm motors using the EM806 drivers (low inductance, parallel wiring) and 70V supply. It's what all the guys use at that forum with far bigger machines than yours or mine with no problems.

D1, maybe you feel you have reached the end of the line, but you still have improvements to make, like the sound enclosure, your ali bed and so on, so please keep posting, if you have the time. I think you just have the winter blues, that's all:)

I will be posting my progress as from February.


Author:  Kitwn [ Sat Jan 06, 2018 4:19 pm ]
Post subject:  Re: CNC Router Build

January for Christie and I here (Exmouth, Western Australia) is much the same as it always was in England: We hide indoors burning lots of electricity to adjust the temperature :D
Fortunately the shed (aka 'studio') has air-con but I only went out there for a few minutes so didn't switch it on. You do get a bit fed up with this heat after the 'poms-in-the-tropics' excitement wears off. In colder climes you can add clothes when it's too cold. Here you can't stay cool even in your birthday suit, you just get burnt instead.

Thanks for that information. I'm now confident that I'm on the right track to get my router working to an adequate standard for now, if only to cut out some of the more complicated wood shapes for it's own next upgrade and some prototype parts for a couple of other projects that are on hold. I would have posted some pictures of my oscilloscope tests the other day but couldn't get 'screen grab' working for some reason. It was enough to make me want to look at an alternative to LinuxCNC software and go for an external pulse generator card so there's another project to add to my list! Something will get finished eventually, though half of my current projects seem to be dependent on each other.


Author:  geraldft [ Sun Jan 07, 2018 6:07 pm ]
Post subject:  Re: CNC Router Build

Hi Kit.

If I may weigh in late. You said at 50mm / sec power is down. And at a motor speed of 1,500 rpm. In my experience that's about the upper speed limit for average nema 23 motors.

With any given motor, the only way to reduce the torque roll-off at higher speed is to increase the supply voltage. For example - If you are currently running ~ 36v then increasing to ~ 70 vdc to give a useful increase. (Might need a change of drive of course)

Otherwise as D1 suggests - re-gearing by using a small timing belt is a good idea, so you get a virtual pitch of 4mm - 6mm. But as you also comment - there is a max speed for ballscrews - above which you probably start having vibration problems I expect...

I would agree that for very high speed a timing belt will be better. If you use something like an AT belt profile it will be low backlash. I've seen this used on German made factory industrial equipment and it can work very well.

FYI - As a general rule, larger frame sizes motors have lower max rpm. I believe related to the inertia of the rotor.... though the maximum useful speed of any motor still depends on the load and how aggressive the acceleration is...

Author:  Kitwn [ Mon Jan 08, 2018 10:20 pm ]
Post subject:  Re: CNC Router Build

I've learned a lot over the last few days. That MYCNCUK post from 2008 and it's spreadsheet actually put some numbers onto the general information given on the GekoDrives website and show how unacceptably fast I was trying to run my motors with these 2mm pitch leadscrews. I don't know the provenance of the information in that post but nobody's debunked it in the last 10 years so it can't be a work of fiction.

At present I'm awaiting a final quote for 10mm lead ballscrews from a recommended supplier on AliExpress. These will give me the kind of feed speeds I want for cutting wood without exceeding the corner speed of the motors or the recommended max speed of the ballscrews themselves as specified in the same spreadsheet mentioned above. I'm hoping the present motor drivers and 36v supply will cope for now. An upgrade of all the electronics is on my ToDo list and that will definitely include an increase of supply voltage to around 70v.


Author:  geraldft [ Tue Jan 09, 2018 9:58 pm ]
Post subject:  Re: CNC Router Build

Nice Excel calculator - so long as you can find and convert the motor specs correctly... I doubt it could be "debunked" given the 3 x safety margin it uses!

To calculate the approximate effect of increasing your motor gearing by 2, you could just double the leadscrew pitch.

I did some adjustments so it can also be used for calculating speeds and motor sizes for a simple slider or dolly. I will probably use 2 x as my margin to start with and see what it tells me... :)


Author:  Kitwn [ Wed Jan 10, 2018 10:17 am ]
Post subject:  Re: CNC Router Build

Well I've gone ahead and ordered the 1610 ballscrews. All future iterations of my CNC router will have no choice but be built around these particular parts.

In almost every use of an electric motor the gearing is arranged to have the load running slower than the motor itself. With 10mm pitch screws if I do need to use belts to change the speeds it will be to run the motor faster rather than speed up the screws and whip them into a frenzy. This is definitely the better option.

The motors I have are rated at 1.9Nm. I think for my tabletop machine which will be mostly used for cutting wheels for wooden clocks and similar sized jobs I should be OK.


PS For the Z axis I'm still going be using a NEMA17 motor and an 8mm x 8mm pitch ACME screw. To parody what someone wrote in the MYCNCUK forum: there's no point building a Z axis that could raise the Titanic, it just adds unnecessary weight to the gantry. Also, backlash isn't an issue in a vertical screw carrying the weight of the spindle.

Author:  geraldft [ Thu Jan 11, 2018 4:53 am ]
Post subject:  Re: CNC Router Build

Hi Kit

"In almost every use of an electric motor the gearing is arranged to have the load running slower than the motor itself."

That may be an accurate observation, but not a reason for not to gearing up when necessary. You know it was fairly common for the industrial shafts driven from steam engines to be geared up - due to the slow moving nature of the piston and flywheel... your situation could be seen as analogous... :)

Author:  Kitwn [ Thu Jan 11, 2018 9:13 pm ]
Post subject:  Re: CNC Router Build

Have you been watching those old Fred Dibnah documentaries again? This talk of steam engines reminds me of the comment I used to make regarding my first motorbike, a 1958 (same age as me) Matchless 250cc single..."All torque and no action".

You are right of course, though in this instance I'm more concerned with the vibrations that would result from the whip on two 1m long 10mm diameter leadscrews hurtling round at up to 3,000rpm.


Author:  DISPLACEMENT 1 [ Fri Jan 12, 2018 1:20 am ]
Post subject:  Re: CNC Router Build

I see this thread is still chundering along :D
I see no problem with up gearing i find my steppers impossible to stop at slow speeds so whats the problem of using all that torque . Their seems to be a challenge of getting steppers to run fast but thats not what they like doing . Edwards vid was very impressive very fast but no load .
All i am saying is rather than using a flexible coupler use a timing belt and pulleys cost is prob about the same . I have already discovered the spring type of connectors are just not man enough for this type of application so ya up to Oldham lovejoy type which can be expensive for decent ones . So timing pulleys and a decent belt is a cost effective solution . Plus you can tweek things more to hit the power range of your motor driver power supply set up .
I dont think maths calculations can ever tell you what is the best combination of things suck and see is the best approach . Theirs just to many variables at play . Quality of motor and driver , how high a voltage applied , the amount of micro stepping . even the length and type of wire to motor can effect things . Not to mention the amount of mechanical inertia in the drive train . You would spend so much time trying to quantify these things so you can calculate the end result .Its just quicker to put it all together and fiddle for a bit plus i am no good at maths :D
Kit i would not worry to much about a 1meter ballscrew wibbling about . If its straight and mounted right you should be fine . My gantry ballscrew is about that length and it runs sweet . Its when i get to a meter and a half i seem to have problems it sags enough i middle to start wibbling ( my technical term)

Author:  geraldft [ Fri Jan 12, 2018 4:08 pm ]
Post subject:  Re: CNC Router Build

D1. If you look at the excel spreadsheet in question, you'll find it doesn't require any math at all. It makes calculations on your behalf for inertia and friction in the system. Input the screw length, diameter and pitch tand it reports maximum safe rpm. Then input the motor specs and it estimates if it is going to run at your desired speed, and with how much margin. While there is a fairly large safety margin in the calculations, it does serve as a useful reference for analysis of a new or current system, and interesting to compare with your suck it and see approach...... :)

Author:  DISPLACEMENT 1 [ Sat Jan 13, 2018 2:59 am ]
Post subject:  Re: CNC Router Build

I thought it would be handy to list my conclusions from building my first cnc router .
1) 10mm pitch ballscrews prob better than 5mm
2) 1 ballscrew centrally mounted below bed would be better than two as long as your gantry is no wider than 1 meter
3) Dont use flexible couplers use belt and timing pulleys instead
4) If you running a stepper at reasonable speeds attained by 10mm pitch ballscrew and timing pulleys then cheap power supply is fine .
5) Because of high spindle speeds feeds depth etc are critical and getting a decent cut can be frustrating . On my vertical mill with a spindle speed of 2500 it was easy to get a good cut as settings dont have to be spot on . Higher spindle speeds extenuate all problems like gumming up tool waste clearance etc .
6) for cutting ali a constant fluid flood would be advisable as makes setting speeds feeds less critical .
7) In ali if ya machining in one small area a lot heat builds up in work piece and this will cause gumming up . If ya machining a large sheet and moving tool rapidly heat doesnt build up . Thats why their are examples out their showing ali plate been machined at ridiculously high speeds . The tool leaves the machining area so quickly no heat is generated . If you were to work a 60mm area at same speeds for a long period a lot of heat would be generated in work piece .

Prob other things and people will disagree but just my real life conclusions .

Author:  Kitwn [ Sat Jan 13, 2018 5:05 pm ]
Post subject:  Re: CNC Router Build

Some very useful conclusions there. There's nothing like experience to tell you what works best.
One of the problems I have with the suck it and see approach is the lack of access to nearby suppliers. Deciding to try something I don't already have the parts for introduces a delay of weeks while I wait for parts to arrive in my remote corner of WA from the other side of Australia ( that's the equivalent of Moscow to John 'O' Groats, all by truck) or the other side of the world (SDP/SI in the USA is my preferred supplier for belts and pulleys).
Having that spreadsheet really helped in getting into the right ballpark and has given me the confidence to lash out the $320 on those ballscrews. I've ordered some Oldham couplings as well and will be direct driving the screws, though one member of MYCNCUK suggests using two screws but with one centrally mounted motor and belt drive to prevent any possibility of racking.

There's never going to be the one ultimate best design and, even if there is, none of us can afford to do the research to try every combination of power supply, controller, drivers, motors belt ratio, leadscrew type and pitch. The best we can do is publicise our mistakes to give the next person a leg up and then work out how best to use as many of the bits we've already bought as possible in a better machine.


Author:  DISPLACEMENT 1 [ Sun Jan 14, 2018 2:44 am ]
Post subject:  Re: CNC Router Build

Hi Kit yes many ways to crack a nut and we all have different priorities some times money no option or availability of parts . I only suggested the timing pulley approach rather oldham connectors because it does give you flexibility to tweak your design as you develop a oldham doesn't .And lets face it once the machine is functioning it can make its own timing pulleys so you not at the mercy of prolonged delivery's .Then as long as you have on timing belt you can tweak to your hearts content. This is the wonder of CNC you should eventually get to the point when you just need to have a good supply of stock material .
Just looking back on post is 3000rpm correct for ballscrew speed you know me and maths :D If a pitch of 5 equals 5mm of travel per rev would that not give you a feed speed of 15000mm/m ?

Author:  edward [ Sun Jan 14, 2018 9:45 am ]
Post subject:  Re: CNC Router Build


is 3000rpm correct for ballscrew speed you know me and maths :D If a pitch of 5 equals 5mm of travel per rev would that not give you a feed speed of 15000mm/m ?

Hi, D1, yes, that's correct. But a 5mm screw rotating at 3000rpm may be interesting to see...

For ali, I don't think I will need more than 5000mm/m and that's for rapids. For normal cutting, much less. But for wood, you need to move much faster, so the 10mm pitch makes sense, you can always reduce the rpm with pulleys when you are cutting ali. All in all, it makes sense to go for 10mm pitch for versatility. Having said that, mine are 1605, as my main interest is cutting metal. Things may evolve with usage.

Author:  DISPLACEMENT 1 [ Sun Jan 14, 2018 10:41 am ]
Post subject:  Re: CNC Router Build

Yeah me done good on maths for once :D

Author:  geraldft [ Sun Jan 14, 2018 2:50 pm ]
Post subject:  Re: CNC Router Build

Hi Kit.

If you are after an economic source of pulleys and belts then have a look at Ebay. " silvers-123" also known as Cupcake World - sell bits and bobs for CNC ex Melbourne. In fact they have just about all the mechanical bits you'll ever need.

I have also used SDP-SI but they are very slow and very expensive - but the only resort when I can't get what I want elsewhere.
BeltingOnline in the UK is faster and will customise shaft sizes and fit grub screws, but shipping still tends to up the price a bit...
The other option is PIES in Melbourne, you can call them and make an order - or go online, though prices are a little higher than the UK.

For Oldham couplings I go to RS Components surprisingly... for some things their prices are actually competitive... and they deliver really fast.

Hi Edward. I have a short 10mm pitch Japanese ballscrew which I've had running at 5,000 rpm. Runs smooth. :)

According to the calculator for a 5mm screw pitch of 10mm diameter and 50cm length, 3,000rpm is the maximum recommended speed. I'm pretty sure the calculation is simply based on the ratio between the diameter and the length of the screw.

Author:  Kitwn [ Sun Jan 14, 2018 3:35 pm ]
Post subject:  Re: CNC Router Build

I have used the unlikely named 'CNC and Cupcake World' site before, in fact that's where I got my 2mm pitch leadscrews and the 16mm linear rails I use in the Z axis. I obvioulsy hadn't paid enough attention to their choice of pulleys and belts as it is more comprehensive than I realised.

3000rpm is what I'd need to get 6000mm/min rapids out of the 2mm pitch screws. I was setting this as a desirable speed on the basis that if I can get that out of the machine I'll be very unlikely to complain of it being too slow in practice.
One interesting thing to come out of the spreadsheet is that the cutting forces (it recons 5N for wood, 20 for aluminium) are only a small part of the motor load. It's overcoming friction and accelerating the mass of the gantry that need all the force.

I'm sticking with direct drive for the time being as it minimises the mods I need to make to the existing design in order to fit the new screws, but I certainly see the point of being able to experiment with speed v torque. If the 1.9Nm motors have trouble driving the 10mm pitch screws then a reduction drive may be necessary. I just need the machine to be working to some extent so I can use it to make the parts to make it work better!

On another note, I need to get on with fetling my lathe a bit better. Santa brought me not one but two kits of materials for making model Stirling engines but I need the tailstock to be much better aligned than it is now to get those made.


Author:  DISPLACEMENT 1 [ Mon Jan 15, 2018 2:43 am ]
Post subject:  Re: CNC Router Build

My problem with my ballscrews is not the mass of the gantry its those dammed oil seals they act like little brakes . Its a quite considerable factor you would have to build into any speed calculation for ballscrews . Without then much higher speeds and acceleration could be achieved. Not sure how you would measure the resistance caused by the oil seals and build that into your excel
By way when the gantry is running with no chance of ballscrews stalling its dangerous. Ya cant stop it i have had a 10mm tool just drag the work piece across the bed and it was very tightly bolted down . It could cause serious damage to any thing trapped in it thats why i am mounting a stop button on it :( .

Author:  Kitwn [ Fri Jan 19, 2018 6:51 am ]
Post subject:  Re: CNC Router Build

4 days!!! 4 days!!!!!!!!!!!!!! For my new ballscrews to get from Shanghai to Exmouth, Western Australia. It often takes longer than that to get stuff from Perth.


Page 11 of 11 All times are UTC - 8 hours
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group