Here we go again . . . MK4

[driftspin]

What do you mean with 'No West though'?I was unable to get 105 icw 209 so i looked for a alternative epoxy.

With similar end strength properties an slow hardning like west 105 / 209.



Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Chaz]

I was unable to get 105 icw 209 so i looked for a alternative epoxy.

With similar end strength properties an slow hardning like west 105 / 209.



Verstuurd vanaf mijn SM-A320FL met Tapatalk

Ok, however different application here.

[routercnc]

Episode 16 is now ready. Having another go at the epoxy leveling . . .


https://youtu.be/FDrcpnVUcEE

[mekanik]

Much more better,Looks promising.
Mike

[routercnc]

Much more better,Looks promising.
Mike
Thanks Mike. Yes much better this time thanks to help on this forum. Something to note - I edited it out because it didn’t work but I had an electric sander in the vice (no paper) inverted which I rested the Epoxymatic container on but it just fizzed the water bath and the epoxy did nothing. I think the sander was too strong and the epoxy was isolated by the rubber washer. Mixing slowly, draining from the bottom (I mixed a bit more than I needed), and using a stick to pop the bubbles all helped.
Getting up close there is still the occasional small bubble but it is much better than before.
Need to leave it a week or two and not be tempted . . .

[Clive S]

Very nice and what patience, I have done a similar thing with a plastic jug with a hole in the side with tape over it then removed the tape for the pour. Nice work with the broken tap as well:applause:

[routercnc]

Very nice and what patience, I have done a similar thing with a plastic jug with a hole in the side with tape over it then removed the tape for the pour. Nice work with the broken tap as well:applause:

Cheers Clive. Yes I thought about your tape method but in the end I wanted to also have a warm water bath around it to help with the bubble removal so went with the water butt tap. As it turned out I only put a bit of hot water in and then chickened out as it felt like the plastic was relaxing and I had visions of the rubber washer failing letting the water into the epoxy. In the end it was pretty hot in the workshop ( heaters on again) and it flowed ok. Not sure I would go with the water bath again . . .

On the other tap I could have ordered a carbide drill but I had it fixed within the hour and it was on with the build !

[Zeeflyboy]

Meticulous as always! Looks more promising this time.

Broken taps... what a pain!

[routercnc]

Cheers Zeeflyboy !

Looks much better this time. Couldn't resist a quick look tonight (I only poured it yesterday) and it looks so much better than before . . .

I used a ground flat edge and there was no rocking.
24281

24282

Then I noticed something at the far end !!!
24283

Checked it with the flat edge and it is very slightly raised:
24284

This is where I popped and removed the last bubbles with the stick so it looks like it was starting to set. I'm confident I can lap this back down when dry.

Onwards . . . !

[Desertboy]

Episode 16 is now ready. Having another go at the epoxy leveling . . .


https://youtu.be/FDrcpnVUcEE

Good job, I bookmarked this vid for when I have a go.

[routercnc]

Episode 17 - time to find out if the epoxy worked and how flat it is . . .


https://youtu.be/ebc7tiAAEZ8

[Nickhofen]

I am prepared this time!:joyous:
24490

[routercnc]

Haha ! Nice one.:eagerness:

[Nickhofen]

Haha ! Nice one.:eagerness:

The timing was amazing, I was preparing this pop corn for the kids when I saw at my email that there is an update to your thread,imagine what happened next!haha!!!

Edit:I didn't ate all the pop corn, just took the nominal taxes!:tyrannosaurus:

[routercnc]

Episode 18. More metrology measurements and flatness checks . . . . then a bit more building to finish the bed.

https://youtu.be/0X1BItweoSY

[Nr1madman]

Nice! You are a true inspiration :)

Skickat från min SM-G955F via Tapatalk

[routercnc]

Nice! You are a true inspiration :)

Skickat från min SM-G955F via Tapatalk

Thank you !

[diycnc]

Awesome looking build!

Is that a piece of granite worktop you are using to measure flatness from?

[routercnc]

Awesome looking build!

Is that a piece of granite worktop you are using to measure flatness from?

It’s a marble plate from our old fire surround. I checked it against other ground surfaces and it is pretty flat at least down to 0.05 mm. Also a ground block placed on it starts to wring together like gauge blocks so it must be close to flat.
Obviously nothing like a grade A plate or AA plate but it was free.

I keep looking at the proper granite blocks for metrology but they are pricey . . . One day maybe

[routercnc]

Episode 19 is ready. This time we are mounting the gantry and getting it moving.


https://youtu.be/BRpSN7UU1cs

[routercnc]

Episode 20. Significant progress !


https://youtu.be/aw0pM_uV1kw

[Nickhofen]

Cool!!! I am a great fan of your work and now that I see that you know karate,I understand it from the way you brake the mdf, I am even greater!!!:witless:
Nice work!

[routercnc]

Cool!!! I am a great fan of your work and now that I see that you know karate,I understand it from the way you brake the mdf, I am even greater!!!:witless:
Nice work!

Thanks Nick. I’m only messing around to add a bit of interest. Less karate and more video editing.

[Blazing Black Beard]

I've been following on YouTube. I have just one question, why the epoxy leveling? Great machine by the way. I was thinking of combining ideas from several projects and picking the best bits from about 4 DIY CNC router machine builds I've seen on YouTube. Yours has been an inspiration. Hopefully being a complete beginner I'll find some very helpful people here.

I plan on starting my design this week and hopefully some time in the next fortnight I will have a solid idea and start working on a bill of materials after finishing the design. It will be very much like yours at about 1.5m square work area.

Again thanks for the great videos on YouTube. You have added an unholy amount of ideas and knowledge to this community and inspired everyone who's watched your YouTube series of 20 videos so far on this project :-)

[routercnc]

Sorry I missed this post. I don’t check the forum too often.

I used epoxy to get the surfaces level and parallel to each other. I don’t have a milling machine and if I did it would need to have 1000mm of travel.

Thanks for the kind words. If it is inspiring others then it makes the video creation worthwhile.

[Zeeflyboy]

Wow... Man that's looking amazing.

Great stuff. I need to do some catching up on this thread!

[Blazing Black Beard]

Sorry I missed this post. I don’t check the forum too often.

I used epoxy to get the surfaces level and parallel to each other. I don’t have a milling machine and if I did it would need to have 1000mm of travel.

Thanks for the kind words. If it is inspiring others then it makes the video creation worthwhile.I may have access to a CNC mill but I have been looking at extruded aluminium as well. I can't yet decide what I want to use because I need to consider milling steel and aluminium on the CNC router build I am designing pretty much from scratch.

I may also produce gantry ends, end stops and motor plates with a CNC mill as the person with access to the mill may accept 3D plans from me. I have yet to design it all but hope to make a start on the plate designs some time during this or next weekend. I am familiar with a few 3D design packages but I have not decided what I wish to use yet. I am hoping Fusion 360 will suffice for all my CNC build needs. I do not wish to build a single item of my CNC mill with hand tools and would much prefer the precision and finish of a professional CNC mill as it is likely that a mill will be available via a friend who is a mechanical engineering apprentice.

I'll open a build log as soon as the 3D design is done :-)

[routercnc]

Things are always a compromise. Cutting steel needs rigidity and a mill style layout is better, but then the travels are low. Router style with moving gantry has a much larger cutting area but then rigidity is less.

[routercnc]

Episode 21 now complete . . . .


https://youtu.be/YLJWRtgedpI

[driftspin]

Hi Routercnc,

If i ever have a promotion video to be produced... Great job again.

Grtz Bert

Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Clive S]

Awesome as always.:beer:

[Zeeflyboy]

Looks awesome! Getting really close to being finished now I guess?

You must be extremely proud - it's going to be fantastic.

[routercnc]

Sorry been away from the forum again
Thank you Clive, Bert, Zeeflyboy for the kind words. I’m very pleased to get this far and have tried to make the best machine I can.

Yes I have motion at last but there is a list of tiny jobs to sort out including:
Proximity switch mounting and wiring in for both home and limit functions
Proximity switch targets
Z home proximity bracket
Small connecting brackets to finish the gantry installation
Rewiring the spindle back to a removable plug with a longer cable which goes through the energy chains this time
Water cooling pipes, tank and pump fabrication and installation
Lighting
Mist spray
Belt guards
Mag base mounting plate on Z axis for DTI
Mounting the spindle
. . .
Skimming the bed
Squaring the gantry and getting the home switches on X axis aligned
Tramming the spindle using the adjustment screws

I think that is it?
Then let’s see what she can do !

[Vulcano]

Sorry for my language - i think is not a perfect (Nederland)
Im very interesting of your router CNC - is it possible to send plans for me via email ??:sentimental:

[routercnc]

Sorry for my language - i think is not a perfect (Nederland)
Im very interesting of your router CNC - is it possible to send plans for me via email ??:sentimental:

I don’t have detailed plans that someone could follow. I’ve made changes along the way and it is quite a complicated machine.
However I plan to put overall dimensions in a future video as a start point for other people to start their own design from.

[routercnc]

Episode 22 is now ready . . .


https://www.youtube.com/watch?v=E6preVxRQJo&feature=youtu.be

[Zeeflyboy]

Looking awesome as ever - looking forward to seeing it at work, it should be a beast!

[routercnc]

Looking awesome as ever - looking forward to seeing it at work, it should be a beast!
Thanks Zeeflyboy. Good thinking, maybe I should call it El Beast ? 😋

Enjoying your beautiful build also, great work.

[1nbcnc]

I have been following along your build, it is really inspiring and also your gantry rigidity spreadsheet helped out a ton in my design, thank you! I do have a question, I noticed in episode 22, your spindle connector is not the original connector that came with, it's got a strain relief, can you point me to where I can find those? I've searched on McMaster and no luck. I know it is metric, do you happen to have the dimensions? I measured it, its something like M19x1 thread and 22mm OD. Any guidance would be appreciated, thank you!

[routercnc]

Thanks 1nbcnc. Glad the old spreadsheet is still helping people out to get a feel for it all.

The original connector plug that came with the spindle was either GX16 or GX20. Certainly that style anyway. It was many years ago so can't remember if it was 3 or 4 pin. I think it was 4 and only 3 pins were used in the socket on the spindle side (going to the 3 windings in the motor). I added an earth cable from the spare 4th pin on the spindle side to one of the screws behind the end cap and used 4 core CY shielded cable to wire it up to the VFD. Note there is not really any strain relief with the GX16/20 connector.

This all went well for many years until one day there was a large blue flash from the connector and it damaged the VFD beyond repair. After years of use the repeated movement of the cable pulling at the connector plug it had loosened the pin connections and it arced. This is fatal for VFDs as they must not see a spindle disconnect when running.
Adding a strain relief above the connector had been on my to-do list, but you know how it is, just one more job to cut and then maybe I will get around to it. Well, it served me right.


So after that I wired the cable DIRECT to the windings, including an earth, and added a cable relief grommet which passes through the hole in the cap and has a nut on the back. 22mm hole in the end cap rings a bell but I really can't be sure as it was many years ago. This is what you saw in the video, the old strain relief grommet. What you then saw me do in the video was replace this with a longer cable, but it is still DIRECT wired to the spindle with the strain relief to manage the stress. This means it cannot be quickly disconnected.

It is similar to this one, but I think it might be a bit larger. Next time I am out in the workshop (sorry, not this weekend and at least the following weekend) I will see if I can measure anything although the spindle is buried in the Z axis.
https://www.ebay.co.uk/itm/Spiral-Cable-Gland-Strain-Relief-4-8mm-M16-Thread/291127273408?hash=item43c889a7c0:g:a2YAAOSwQPlV8UF t:rk:18:pf:0

Many on this site would recommend fitting a removable connector AND strain relief after it, and I would not argue against it. However, I've had no cause to remove the spindle, and I know with the direct wiring and long strain relief grommet it should run OK for a long time. Hope this helps.

[1nbcnc]

Thanks 1nbcnc. Glad the old spreadsheet is still helping people out to get a feel for it all.

The original connector plug that came with the spindle was either GX16 or GX20. Certainly that style anyway. It was many years ago so can't remember if it was 3 or 4 pin. I think it was 4 and only 3 pins were used in the socket on the spindle side (going to the 3 windings in the motor). I added an earth cable from the spare 4th pin on the spindle side to one of the screws behind the end cap and used 4 core CY shielded cable to wire it up to the VFD. Note there is not really any strain relief with the GX16/20 connector.

This all went well for many years until one day there was a large blue flash from the connector and it damaged the VFD beyond repair. After years of use the repeated movement of the cable pulling at the connector plug it had loosened the pin connections and it arced. This is fatal for VFDs as they must not see a spindle disconnect when running.
Adding a strain relief above the connector had been on my to-do list, but you know how it is, just one more job to cut and then maybe I will get around to it. Well, it served me right.


So after that I wired the cable DIRECT to the windings, including an earth, and added a cable relief grommet which passes through the hole in the cap and has a nut on the back. 22mm hole in the end cap rings a bell but I really can't be sure as it was many years ago. This is what you saw in the video, the old strain relief grommet. What you then saw me do in the video was replace this with a longer cable, but it is still DIRECT wired to the spindle with the strain relief to manage the stress. This means it cannot be quickly disconnected.

It is similar to this one, but I think it might be a bit larger. Next time I am out in the workshop (sorry, not this weekend and at least the following weekend) I will see if I can measure anything although the spindle is buried in the Z axis.
https://www.ebay.co.uk/itm/Spiral-Cable-Gland-Strain-Relief-4-8mm-M16-Thread/291127273408?hash=item43c889a7c0:g:a2YAAOSwQPlV8UF t:rk:18:pf:0

Many on this site would recommend fitting a removable connector AND strain relief after it, and I would not argue against it. However, I've had no cause to remove the spindle, and I know with the direct wiring and long strain relief grommet it should run OK for a long time. Hope this helps.

Thank you! That's helpful, I will certainly give your suggestion a try.

On a different note, have you had any challenges with the ball screw lock nut in the past? I noticed on my router that the lock nuts will back out when the gantry is moving in one direction, I know that the lock nut is supposed to preload the inner race of the AC bearings, and the set screw on them are so tiny and I can't really torque them hard or they will strip, to be honest the set screws don't seem to work that well. I've seen people order ball screws with a longer thread for two lock nuts back to back, but it's not an option for me any more. Any suggestions or solutions would be appreciated.

Thank you!

[routercnc]

I have had them come loose occasionally over the years. I've always managed to tighten the nut and carefully tighten the grub screw but yes there is the fear of stripping the thread or more likely rounding out the allen key hole.

Other options to think about:
As you say, longer screws and double lock or jam nuts
Maybe blue loctite on the nut but don't get it near the bearing, or maybe PTFE tape??
Dropping a short length / disc of brass rod in the grub screw hole first, then the grub screw will mold it onto the thread
Making a brand new nut from steel (and then heat treating it) with a larger grub screw?

Anything from anyone else?

The other thing to think about is how to hold the ball screw when tightening the nut (once it is all installed with coupler on etc.). I fold up plenty of shop towels into a narrow strip about 30 mm wide then wrap it around the ball screw. Then apply as much pressure as you dare using a mole grips . . . . . pause while I find the US name for such as tool . . . . vice grips, to hold the ballscrew and tighten the nut using a spanner/wrench.

[1nbcnc]

I have had them come loose occasionally over the years. I've always managed to tighten the nut and carefully tighten the grub screw but yes there is the fear of stripping the thread or more likely rounding out the allen key hole.

Other options to think about:
As you say, longer screws and double lock or jam nuts
Maybe blue loctite on the nut but don't get it near the bearing, or maybe PTFE tape??
Dropping a short length / disc of brass rod in the grub screw hole first, then the grub screw will mold it onto the thread
Making a brand new nut from steel (and then heat treating it) with a larger grub screw?

Anything from anyone else?

The other thing to think about is how to hold the ball screw when tightening the nut (once it is all installed with coupler on etc.). I fold up plenty of shop towels into a narrow strip about 30 mm wide then wrap it around the ball screw. Then apply as much pressure as you dare using a mole grips . . . . . pause while I find the US name for such as tool . . . . vice grips, to hold the ballscrew and tighten the nut using a spanner/wrench.

Thank you for all the suggestions, I will give them a try for sure.

[jkkmobile]

I use oneway bearing wrench to hold the ballscrew when tightening the nuts.

Also, my SYK bearing holders have nuts with setsrews, but the thread hole does not go all the way tru so you don't destroy the ballscrew threads.

This is my wrench:

https://www.amainhobbies.com/align-feathering-shaft-wrench-8-10mm-shafts-agnhot00005/p213519

[1nbcnc]

I use oneway bearing wrench to hold the ballscrew when tightening the nuts.

Also, my SYK bearing holders have nuts with setsrews, but the thread hole does not go all the way tru so you don't destroy the ballscrew threads.

This is my wrench:

https://www.amainhobbies.com/align-feathering-shaft-wrench-8-10mm-shafts-agnhot00005/p213519

Thank you for your suggestion, I will give that a try as well.

[silopolis]

Anything from anyone else?


Think that's This Old Tony that uses a hardened steel ball under the screw to create a "single point" of contact and increase the pressure on the shaft.
Could even have a video just about that IIRC



Envoyé de mon ONEPLUS A5010 en utilisant Tapatalk

[routercnc]

Yes, for a plain shaft that is OK but remember this is pressing down onto the thread. A steel ball would damage the thread making it difficult to remove the nut so the brass disc is a better option as it is soft and will mold into place.

[jkkmobile]

SYK nuts have brass inserts:

https://www.syk.tw/en/product/201705040003

You can buy them from resellers like cnc4you:

https://www.cnc4you.co.uk/BK12-12mm-Nut?search=Nut

[Vulcano]

Episode 22 is now ready . . .


https://www.youtube.com/watch?v=E6preVxRQJo&feature=youtu.be

Episode 22 is now ready ok hmmm when you planing episode 23 with some mesurment\s ??

[routercnc]

Episode 22 is now ready ok hmmm when you planing episode 23 with some mesurment\s ??
It could be a couple of weeks away based on the little free time I have at the weekends and the long list of things to finish off.
I have been doing a bit here and there and filming short sections. See if you can spot the joins !

[paulus.v]

It could be a couple of weeks away based on the little free time I have at the weekends and the long list of things to finish off.
I have been doing a bit here and there and filming short sections. See if you can spot the joins !

I'm eagerly waiting for your actual ali cutting video and hopefully some stiffness measurements.

I have a questions as I'm starting a new build trying to use some of your design solutions.

Have you researched or tested how the double beam gantry with the rails on top (80-100 mm more leverage) performs in the Y direction compared with the twisting of the popular single beam gantry?
You have mentioned at some point that mori seiki used this gantry design but I couldn't find any information. Could you point me where to find anything about it?

[routercnc]

I'm eagerly waiting for your actual ali cutting video and hopefully some stiffness measurements.

I have a questions as I'm starting a new build trying to use some of your design solutions.

Have you researched or tested how the double beam gantry with the rails on top (80-100 mm more leverage) performs in the Y direction compared with the twisting of the popular single beam gantry?
You have mentioned at some point that mori seiki used this gantry design but I couldn't find any information. Could you point me where to find anything about it?

Here is the pdf link for the Moro Seki 5000
https://uk.dmgmori.com/resource/blob/308434/a555d7494ad79fd57b5a92f6609b338d/pm0uk-nmv5000-8000-pdf-data.pdf

This shows the box in box idea which was part of my inspiration.

I was worried about rails on top as they are further away but when they were underneath they were difficult to level plus would get chips and coolant on them or risk being hit.
On the sides made the gantry even wider which I didn’t like. When I saw the Mori rails on top and the beefy construction I went with that.

I have not measured the stiffness yet still other jobs to finish (build coolant system, skim bed, tram head, belt guards. . .) and other commitments are limiting workshop time but we will get there and get a video out when it’s done 😀

[paulus.v]

Thank you!

That dmg mori machine is a real beast. The blocks spacing on the Y rails is greater than the travel distance... And everything looks like cast iron. It is great that they put a lot of technical details and images in that presentation.

I never thought to put the rails hanging under the gantry with all that ram weight trying to snatch off the rails...

One more question, what is your gantry (rails) total length and travel distance?

[Zeeflyboy]

re tightening the retaining nuts, you can also dremel a little flat head screwdriver slot into the end of the ballscrew shaft and use that to hold the shaft still while tightening the nut.

Well I say "you can", I've never seen anyone else do it but it soother's probably a really great reason why not to, but it works for me! No more worries about damaging the important part of the screw vs not tightening it well enough.


https://i.imgur.com/JpyoZXO.jpg

[routercnc]

re tightening the retaining nuts, you can also dremel a little flat head screwdriver slot into the end of the ballscrew shaft and use that to hold the shaft still while tightening the nut.

Well I say "you can", I've never seen anyone else do it but it soother's probably a really great reason why not to, but it works for me! No more worries about damaging the important part of the screw vs not tightening it well enough.


https://i.imgur.com/JpyoZXO.jpg

Nice idea. I was thinking of flats machined on the end or on the last part of the ballscrew behind the housing but a slot looks good as it still supports all of the pulley. My other thought was that with the right equipment you can wobble broach a hex pocket into the end.

Paulus, rail is about 700 long and travel is currently 360. I could go a bit more but when the spindle is quite low nearer the bed it would hit the sides so have limited it to that.

[routercnc]

Episode 23 is now ready . . . . . finishing off ready for first real chips !


https://youtu.be/5KOWKN2rFi4

[Chaz]

Well done mate, certainly one of the best DIY builds I have ever seen. The attention to detail is immense.

In the cutting videos, the spindle bogs down. What power is it? Are you running full speed?

[routercnc]

Hi Chaz,

Thank you. I can't tell you how relieved I am to have finally 'finished' the machine (they are never truly finished are they?) and I can spend some time with the family again. I hope it gives some ideas and inspiration to others.

Yes, the spindle is a 1.5 kW Chinese spindle and I knew it would be the next limiting factor. I have other designs in CAD for various spindle options but they will have to wait for now. It works, and I can certainly make a few hobby parts with it.

It was running at around 12,000 rpm in the film (full speed is 24,000 rpm), with a 6 mm 2 flute 45 deg cutter for aluminium, with the approximate WOC and DOC shown on screen. Feed rate varied - the heavy cuts were around 700 mm/min and the finish cuts were around 1400 mm/min. I finished the tram and final build yesterday morning, and I basically had about 20 minutes yesterday afternoon to get some cutting footage before my 'time was up' so basically went for it. Although I didn't have much time to experiment I soon found the limits. I actually broke the tips off both flutes by being so aggressive.

Much to think about and reflect on over Christmas, but first a break !

[Clive S]

Nice to see it finally finished and making chips. It certainly is a big credit to you and I have enjoyed all the vids in the making.

[Chaz]

Hi Chaz,

Thank you. I can't tell you how relieved I am to have finally 'finished' the machine (they are never truly finished are they?) and I can spend some time with the family again. I hope it gives some ideas and inspiration to others.

Yes, the spindle is a 1.5 kW Chinese spindle and I knew it would be the next limiting factor. I have other designs in CAD for various spindle options but they will have to wait for now. It works, and I can certainly make a few hobby parts with it.

It was running at around 12,000 rpm in the film (full speed is 24,000 rpm), with a 6 mm 2 flute 45 deg cutter for aluminium, with the approximate WOC and DOC shown on screen. Feed rate varied - the heavy cuts were around 700 mm/min and the finish cuts were around 1400 mm/min. I finished the tram and final build yesterday morning, and I basically had about 20 minutes yesterday afternoon to get some cutting footage before my 'time was up' so basically went for it. Although I didn't have much time to experiment I soon found the limits. I actually broke the tips off both flutes by being so aggressive.

Much to think about and reflect on over Christmas, but first a break !

Ok, that makes sense. Go 24K, anything less has even less torque. I'd fit the largest spindle that you can in there, clearly limiting the machine. Congrats once again.

[routercnc]

Nice to see it finally finished and making chips. It certainly is a big credit to you and I have enjoyed all the vids in the making.

Thanks Clive. It’s been a long time coming !

[routercnc]

OK I’ll try some other speeds / feeds next time

[Nickhofen]

I enjoy so much the build and the detailed video you upload.
Thank you so much about all the effort and the documentary!
Great machine!

[routercnc]

Thank you. There should be more videos in future but time for a break !

[Fancy]

Congratulations on a phenomenal build. What ever i write i feel i won't do you or your build justice. Love the videos and would love to see more of your projects.

[routercnc]

Congratulations on a phenomenal build. What ever i write i feel i won't do you or your build justice. Love the videos and would love to see more of your projects.

Thank you ! Feels good to have finished it - or at least get it working (already thought of a few tweaks here and there).

Something I asked about along time ago on this forum and received plenty of help on (thank you!) was getting the 24V prox signal into the 5V breakout board. In ep22 you saw the circuit that was recommended to me but I have found it a bit unreliable and actually crashed into the sensor on one occasion. So I did a bit of searching and tried a diode in the black wire from the sensor (bar pointing to sensor I think?). This works very well without any issues and gives either a solid 5V or 1V when switched. The diode I had to hand was a generic 1Nxxx type which is meant for relays (?) and such but have since bought a proper switching diode (not fitted yet) which I guess is much faster or at least more suitable. When I’m back at the PC I’ll post a diagram for the electrical guys to comment on.

[paulus.v]

..and finally making chips after a 3 year journey! Congratulations!

I would be grateful if you could measure the stiffness in the x and y directions.

What breakout board/ controller are you using? Does it have optocouplers on the inputs? I had the same problem feeding 24v signal to 5v bob. In my case the signal was fed to the optocoupler through a current limiting resistor with a value calculated for 5v input. I just had to replace the resistor with one calculated for 24v.

[routercnc]

..and finally making chips after a 3 year journey! Congratulations!

I would be grateful if you could measure the stiffness in the x and y directions.

What breakout board/ controller are you using? Does it have optocouplers on the inputs? I had the same problem feeding 24v signal to 5v bob. In my case the signal was fed to the optocoupler through a current limiting resistor with a value calculated for 5v input. I just had to replace the resistor with one calculated for 24v.

Thank you Paulus. Great to have it working like it should.

I was planning to measure the stiffness and compare to my old machine, plus a typical commercial stiffness. Depends on when I can get another slot in the workshop. Probably a bit rude to do it when the in-laws are here - although . . .

The breakout board was free with the drivers and is very basic. I think it is called a DB-2505. ?
From memory it has black chips near the inputs which I always assumed were optocouplers. Replacing the resistor sounds interesting.

[routercnc]

Time for one last video in the series ! Some finishing touches . . .


https://youtu.be/vd-Ly1454rw


Still plan to measure the stiffness - I may get a chance to get in the workshop today and report back tomorrow on this thread . . . sorry for the delay

[AndyGuid]

Many thanks routercnc for an absolutely awesome build!

Apart from the build aspects that relate to your specific CNC design (some of which is frankly way too complex for me to contemplate), I have learnt so much from your videos about what I can only call "general machining principles", which I would never have learnt otherwise!

Thanks again,
Andy

[Nickhofen]

Most excellent machine build ,I enjoy very much the videos and I hope one day I will make good use of all thiaw tips you shared!

[routercnc]

Thanks Andy, Nick. I'm just a hobby machinist and just passing on what I have learnt along the way. Glad it was all helpful.

[routercnc]

OK so yesterday I managed to take some measurements for the X and Y stiffness of the new Atlas Mk4 machine.

The machine was positioned so the gantry was half way along the X axis, and the Y axis was positioned at the mid point of the gantry. So the X and Y positions represented worst case. The Z axis was lowered 50 mm from home in a typical machining position.

I used a digital luggage scale with a hook on the end to pull on the spindle just slightly above the collet nut (6mm end mill was inserted), while a DTI was pointed to the round section on the collet nut just above the flats.

5kg load was applied (~50 N) and the reading in for both X and Y was somewhere around 0.01 - 0.015 mm. I'm wondering how reliable this reading was given the needle barely moved and 0.01 to 0.015 makes quite a difference on the stiffness values. I guess I really need to buy a 0.001 mm DTI !

Anyway, for Atlas mk4 this works out at:
X between 3333 and 5000 N/mm (0.015 and 0.01 mm respectively)
Y between 3333 and 5000 N/mm (0.015 and 0.01 mm respectively)

My previous mk3 machine was:
X 1000 N/mm
Y 1250 N/mm

So up to 5 times stiffer. It certainly sounds different when machining and as you might have seen in the videos you can certainly take a good cut so long as there is some lubrication. I didn't measure Z this time as it was a bit awkward Vs the previous machine.

I did not have long to explore further but just before I finished I quickly pulled on the ballscrews when the DTI was still on the spindle and could make the needle move a bit. So maybe the limit is the stiffness of the small AC bearings in the ballscrew housings, or come to think of it they might be the deep groove ball bearings? Anyway I think as it is all working well I should stop there otherwise I will never use it to make something !

[paulus.v]

Thanks a lot for the stiffness measurements!

I had to put aside the research and design for my new cnc, that's why I didn't bother you.

If you accept challenges, here is one for you, as I see you are an expert at sheet metal bending: dust covers :joyous: 25475

Re homing/limits, I have set in Mach3 the sensors for both home and limits as well as setting soft limits with the min. max. travel. It works nice, stops before "hitting" the sensors in jog mode and I get limit warnings when g-code is out of the machine travel. With this setup I have never hit the hard limit switches so far.

[Voicecoil]

Hi Routercnc I've just been going back through one or two of your videos when the dire internet here allows, and I'm curious about one of the things shown in episode 23, where you square up the gantry.It looks like you're achieving this by having independent limit/homing switches on each of the horizontal axis ballscrews, but doesn't this leave the whole assembly under quite a bit of tension?? or do you home it with the screws slackened off a tad then tighten?And does it mean you need to have separate drives from the control software for each of the steppers?
Although it's not assembled yet I'm trying to figure out how I will square up the gantry on my little machine, bearing in mind that my gantry is quite short (~600mm) and a very rigid bit of box section.
Thanks, Trevor

[routercnc]

Thanks a lot for the stiffness measurements!

I had to put aside the research and design for my new cnc, that's why I didn't bother you.

If you accept challenges, here is one for you, as I see you are an expert at sheet metal bending: dust covers :joyous: 25475

Re homing/limits, I have set in Mach3 the sensors for both home and limits as well as setting soft limits with the min. max. travel. It works nice, stops before "hitting" the sensors in jog mode and I get limit warnings when g-code is out of the machine travel. With this setup I have never hit the hard limit switches so far.

That sheet metal looks like a challenge. I guess they are concertina type? I think Chris Deprico (?) on YouTube made some and it looked very fiddly making the X bracing supports underneath which guide it.

Yes I’ve since added soft limits and it works well

[routercnc]

Hi Trevor
Yes you can only use this to square up a mm or so otherwise there is too much tension. You need to make and assemble it as square as possible first. Depending on the design you might be able to slacken the gantry bolts and square it up using separate home switches then retighten as you suggest.

I had to be careful with mine as to do independent homing I needed to switch off slaved axes temporarily. This meant my Z motors were also no longer slaved (I have 2 on Z) so after homing X I switched slave back on and now usually just home off the left sensor with the motors slaved. I check each side with callipers to ensure they are still the same distance from the end plates.

Yes separate drives are required plus a spare axis on your break out board (usually A). You can’t do this trick from a single motor with belt coupling to the other ballscrew and would need a way of making tiny adjustments on one ballscrew relative to the other mechanically. Probably the grub screws holding the pulley onto one of the shafts

[Voicecoil]

Thanks very much for the clarification. I'm not so worried about squaring the gantry on the horizontal sense (having it parallel to the bed) - I have a big digital height gauge so can get that initially set pretty good, besides mine's single motor on the Z.- It was more the X,Y squareness for which the only "old fashioned" tool would seem to be a fairly huge and vastly heavy 3 way angle plate. I'm running twin motors/screws on the long axis (which I call Y since I look at it from the front), but was going to simply parallel up the inputs to the drivers, in the light of this maybe I need to have a bigger BOB or a bit of switching.

[Lee Roberts]

Would a 3.1Nm motor really find it that much of a challenge to drive 4 ball nuts all engaged at the same time on your Y ?

Super cool of you to do what your about to do btw :wink:

P.S Don't forget we have an open source section/forum.

[routercnc]

I'm not sure if I'm honest. To drive 4 ballnuts without pre-load would be fine I think, but when the ballnuts are clamped together there is an extra load to overcome, and the Y/Z is pretty heavy. Is it worth the risk to have missed steps on a critical job? Not sure.
Decided to leave it like that and put some pre-load between the front and back ballscrews instead.

Sending pm shortly . . .

[routercnc]

Hi Lee, pm sent . . .

[routercnc]

For anyone interested here is episode 26 of the CNC series giving an overview of my humble but functional CNC controller.

Also, for anyone who likes oscilloscopes there is a review of the Hantek DSO5102P oscilloscope supplied by Banggood.

Then right at the end of the video is a short montage of the controller build.


https://youtu.be/KFPSKisqVVw

[routercnc]

If you see this Lee PM sent . . .

[routercnc]

I've had many requests for the CAD of this machine. I've also had requests for narration so I decided to put together a 3 part mini series to show a shorter version of the build for new viewers and also explain how the CAD can be downloaded for free (explained at the end of part 3). If you have seen the original 27 episodes of this build there is no new material, but the narration may help with explaining how I approached it.

Here is part 1 of 3:


https://youtu.be/FQDzNQ-sZA0

[routercnc]

Like buses you wait for ages then 2 come along together. Here is part 2 of 3 of this miniseries. Again, if you have seen the original 27 episodes of this build there is no new material, but the narration may help with explaining how I approached it.


https://youtu.be/3XREfXwKxKw

[routercnc]

Again, if you have seen the original 27 episodes of this build there is no new material, but the narration may help with explaining how I approached it.
Part 3 of 3 . . . .

https://youtu.be/v1YXjxckv4U

[gojones]

The major benefit of this design is the ability to machine a workpiece that is way larger than you could fit in a conventional fixed gantry mill. You could place one end of a 20' beam on the bed and machine one end, or even machine the entire length in stages.

[routercnc]

Since I have an air compressor . . . I decided it was about time !


https://youtu.be/tSb5bdfHPyg

[rhay80]

As per the YouTube comment, here are some pics of my cool mist set up, but for my manual milling machine.

31119

31120

[Clive S]

Those water filters will take 8 bar it will probably say that on the box and the little screw on RH side is for letting the air out (priming) when being used as a water filter .

Nice vid.:thumsup:

[routercnc]

OK good to know thank you

[routercnc]

I decided to do some of my own testing . . . here are the results:


https://youtu.be/tc2L6QT_G6I

[routercnc]

About time I built this thing !


https://youtu.be/xZEE-I_cbtc

[routercnc]

I'm not looking to start a new project any time soon, but I'm planning ahead for a new or repaired WC spindle. In my last video (above) you probably heard the noise as the spindle ramped down. It makes a rough noise when it goes through a resonance at a few 1000 rpm or so.
When running at 12,000 rpm it sounds ~OK but it does not bode well for the life of the motor. When stationary if I grab the end of the spindle I can feel a very slight amount of play so this is probably the cause. It has been doing this for a while now.

I think these spindles have 2 angular contact bearings at the nose end, and one ball bearing at the far end. It might be a case of sorting out the preload as I believe there is a LH threaded bolt at the top end which may set this. Or they might just be deep groove ball bearings at the nose end, not sure on this model. Maybe I can just install new bearings and that would sort it out.

This is a 1.5 kW Chinese WC spindle which has been doing great service since about 2008/9 (?) and has cut a lot of aluminium, up to 20mm thick. So it has had a fairly hard life considering, but I do need to start looking into the options in case it goes downhill quickly.

I'm right at the start of the planning, so it is early days, and replacing the bearings would be an option. However, the spindle has always been the lowest performing part of this mk4 machine and I think it could handle much more power, especially with the new mist system.

After a few hours research I've got this table so far (obviously still looking into the options and specs):

31178


I'm not looking for an auto tool changer, I'm OK using ER collets and manually swapping tools over. I want to cut aluminium mainly, and some steel parts (using small cutters probably) such as keyways. Any thoughts and experience welcome, thanks.

[routercnc]

I’ve looked into lots of options and one of them is to go for a 3kW motor (Chinese wc spindle) via a belt drive onto a DIY cartridge with an ER32 collet. With a ratio of 2:1 I get x4 the torque of my 1.5 kW direct drive spindle and 9,000 rpm. Or 1.5:1 I get x3 the torque and 12,000 rpm. I’ll need to check the max rpm of the bearings I use for the cartridge.
Anyway only planning at this point so my question is can I run the 3kW spindle off a standard UK circuit (13A). The spindle claims it runs up to 10A which is OK on paper but I wanted to check any real world usage, thanks.

[routercnc]

Finished !
I got the mixer block modified to take the blast system - and then I broke the locline fitting. Still, managed to get a very basic bit of machining done and looks promising so far. Not a very in-depth machining session at the end, just a quick check and I think this is going to work OK.



https://youtu.be/LVnqZSLWL3M

[Voicecoil]

I’ve looked into lots of options and one of them is to go for a 3kW motor (Chinese wc spindle) via a belt drive onto a DIY cartridge with an ER32 collet. With a ratio of 2:1 I get x4 the torque of my 1.5 kW direct drive spindle and 9,000 rpm. Or 1.5:1 I get x3 the torque and 12,000 rpm. I’ll need to check the max rpm of the bearings I use for the cartridge.
Anyway only planning at this point so my question is can I run the 3kW spindle off a standard UK circuit (13A). The spindle claims it runs up to 10A which is OK on paper but I wanted to check any real world usage, thanks.

Another option might be to go for a constant power spindle, i.e. one where the torque increases as the speed decreases/ I have one from Jian-Ken and it chomps through ali nicely.

[routercnc]

Another option might be to go for a constant power spindle, i.e. one where the torque increases as the speed decreases/ I have one from Jian-Ken and it chomps through ali nicely.

Do you have a link?

[mekanik]

was having a look the other day
https://www.jian-ken.com/constant-torque-motor-spindles/
Regards
Mike

[Voicecoil]

Do you have a link?

www.jian-ken.com/high-speed-motor-spindle-with-constant-power/jgh-85-1-8kw-24000rpm-er20-motor-spindle.html

[routercnc]

www.jian-ken.com/high-speed-motor-spindle-with-constant-power/jgh-85-1-8kw-24000rpm-er20-motor-spindle.html

Ok great thanks both, more to consider . . .

[routercnc]

I had someone ask about the ProfiCAD electrical diagram I drew up for the AXBB-E electrical re-work on my mk4 CNC machine. It's been a while since I drew but it made sense at the time and seemed to work. By all means use it for ideas and as a start point but ultimately you need to be happy and confident with what you are doing to ensure it meets your requirements.

I've uploaded this as a zip so you will need to unpack it.

31774