[wado1971]

Time for an update on my build…

First, I’ve completed installing the limit and home switches and have posted a blog post explaining the design – nothing major but glad that one is complete. Take a look here if you have interest -> Limit & Home Switches

Second, the machine enclosure design is complete and installed. See this post -> "Enclosure" I’ll eventually post a video of this implementation and the conveyor moving on the end of the blog post (coming very soon - subscribe to channel to get instant notification)

BTW – I did also post a video some time back showing an overview of the machine and my first attempt at machining aluminum -> https://youtu.be/herMY_2WowM

I won’t post the new v3 cad files with the above changes to my website until I live with the enclosure parts for a week or two. Will post here when new files are ready for download.

And just for the curious minded... In the download section of my website I ask a survey question which allows me to track # of downloads... The results are below.
"What size router work area do you need?" Out of 52 replies. 18"x18"=49%, 24"x22"=26%, 'bigger'=25%
"What is your budget for CNC Router?" Out of 60 replies 0-$1000=22%, $1-2k=18%, $2-3k=27%, >$3000=33%
"What is your interest in this design?" Out of 32 replies 67%=Would like to see complete plans and try build, 33%=Just Curious
I'm surprised at the small size of 18"x18" and the highest cost winning out here.

Cheers.

[wizard]

You could probably forget using a dust shoe if it's Aussie aluminium, most is quite gummy as has been pointed out, and I can tell you it aint fun cleaning oily swarf off the dust shoe!

The other thing to consider is the high risk of explosion if your machining aluminium dry, and using some form of dust extraction, EXTREMELY good earthing and anti-static ducting is mandatory.

I must have missed this post but wanted to point out that the finer the dust the greater the potential for an explosion. Sanding and grinding operations with aluminum can be very dangerous. Worst some dust explosions happen in the very duct work that is supposedly removing the dangerous stuff.

Look up thermite!

Thermite is a reaction that takes place with iron and steel often with added stuff when used professionally. It is equally dangerous and the fires it produces are incredibly hot but it is slightly different than a dust explosion where only aluminum is the active ingredient.

At one time the railroads used Thermite to weld rails together. An interesting process but I'm not sure it is used anymore for that.

I'll simply be building a 100mm or so high barrier around the outside of the frame, with drainage holes going to the pump reservoir, doesn't need to be anything fancy, just high enough to catch all the spray, and a decent filtration system so you catch all the swarf but the liquid drains out. If it was a mill, I'd build a screw into a sloping base to actually move all the swarf out to a bucket, with the liquid draining off prior.

Either way, the novelty of cleaning up after alu wears off quick. So in my book, an easier more workable solution is the way to go.

cheers, Ian

Aluminum does produce a lot of big flaky chips.

[wizard]

Time for an update on my build…

First, I’ve completed installing the limit and home switches and have posted a blog post explaining the design – nothing major but glad that one is complete. Take a look here if you have interest -> Limit & Home Switches

Second, the machine enclosure design is complete and installed. See this post -> "Enclosure" I’ll eventually post a video of this implementation and the conveyor moving on the end of the blog post (coming very soon - subscribe to channel to get instant notification)

BTW – I did also post a video some time back showing an overview of the machine and my first attempt at machining aluminum -> https://youtu.be/herMY_2WowM

I won’t post the new v3 cad files with the above changes to my website until I live with the enclosure parts for a week or two. Will post here when new files are ready for download.

Just remember that enclosures are both good and bad things to have. No enclosure is perfect and the restricted access always creates issues.

And just for the curious minded... In the download section of my website I ask a survey question which allows me to track # of downloads... The results are below.
"What size router work area do you need?" Out of 52 replies. 18"x18"=49%, 24"x22"=26%, 'bigger'=25%
"What is your budget for CNC Router?" Out of 60 replies 0-$1000=22%, $1-2k=18%, $2-3k=27%, >$3000=33%
"What is your interest in this design?" Out of 32 replies 67%=Would like to see complete plans and try build, 33%=Just Curious
I'm surprised at the small size of 18"x18" and the highest cost winning out here.

I wouldn't be that surprised. First you are building a high quality machine here so people with suitable budgets will,gravitate to it. Second an 18" square work area is pretty huge when you think about it. To buy a conventional milling machine with a workspace that big would be costly and heavy. A lot of useful work can be done in a work area that large especially if the focuse is one aluminum.

Cheers.

Keep up the good work!

[wado1971]

Just remember that enclosures are both good and bad things to have. No enclosure is perfect and the restricted access always creates issues.

I wouldn't be that surprised. First you are building a high quality machine here so people with suitable budgets will,gravitate to it. Second an 18" square work area is pretty huge when you think about it. To buy a conventional milling machine with a workspace that big would be costly and heavy. A lot of useful work can be done in a work area that large especially if the focuse is one aluminum.


Keep up the good work!

@Wizard - thanks for the additional info on aluminum dust - important info to know!

Now that I read your reply, I realize your correct on the size 18x18 being enough. My mill is only 18x8.... I guess I get caught up in comparison to the long strokes of the moving gantry designs. If in the future if I release a 'standard' design of this build I may choose a base size of 18x18 and simply allow variations on that to 'stretch' the model.

As for the enclosure being good/bad - understand and I tried to keep that in mind for the design by making all parts very quick removal but the main feature of the conveyor protecting the rails/screw was a must for this design and the main item which will be permanently mounted (I've included a pic below because I'm being lazy in making the video). I really like how the window shade works with the PVC rollers - didn't have enough room for standard bellows style protection. I'm just now researching what material I can buy from McMaster which will be more robust than the window shade but function the same as it slides over the PVC perfectly. Anyone have a suggestion as to what poly-whatever matches up well?? - I'm all ears. The current shade is 0.005" but too thin.. would like to double that at a minimum.

[wizard]

As for your window shade upgrade you might try suppliers to injection molding. There is a device much like a window shade that is used on injection molding machines to contain material when the mold opens. These are basically heavy duty window shades, here is an example: http://www.ppe.com/15cat/0073.pdf. Probably not heavy enough for what you need.

You may need to consider alternative approaches.

[jonathlee]

For your wish for an enclosure without all of the access problems, and also your need for thicker plastic for the "window shade" protector, there is another thread with what might have answers to both issues.

Check out the picture of the enclosure here: http://www.cnczone.com/forums/tormac...st1834290.html.
Post # 7 on that same thread, by pickled, has a link to a site that sells heavy duty, heat resistant plastic. His suggested use was for the enclosure but it seemed thick enough for use in your protector as well. If not that exact product, the site he linked to probably has other types that may fit better.




Sent from my SM-P900 using Tapatalk

[wizard]

I knew I saw something purpose built for this before: Roll Up Way Covers and Rollaway Covers and another example: Roll Up Covers.

I could imagine some of these solutions are pricey. At least they offer ideas that might be suitable for what you want.

- - - Updated - - -

I knew I saw something purpose built for this before: http://skirting-and-bellows.com/roll-up-way-covers-and-rollaway-covers/ and another example: http://www.dynatect.com/protective-covers/roll-up-covers.

I could imagine some of these solutions are pricey. At least they offer ideas that might be suitable for what you want.

[UA_Iron]

Great video and update, Wado! Your machine is awesome and I imagine what many of us will be modelling ours after...

Just a few things I noticed:
1. The cabletrack/cablecarrier/cable chain: Looks like the mounting attachment point on the Z-axis might be low causing it to bow up at the extremes - causing an additional bend point at the cables in the exit of the chain. I believe best design practice is to have cables come in/out parallel at the exit points of the cable track. This is probably fine for the use case, but it also appears that the cable carrier is full of cables, I believe typical design practice is to not exceed 60% occupancy by cross-sectional volume. At a greater than a certain percent fill occupancy, many cables will be pushed in or out from the neutral axis where they will be binding in either compression or tension with each pass leading to abrasion of the cable insulation. You will want to make sure that the cables in the carrier are firmly secured on the stationary side to avoid ratcheting of cables as well. There are also rules for when one should secure cables/tubing on both ends of the cable track, but they escape me right now. In any case, it would be wise to inspect the cables and cable insulation to look for any of this abrasion and adjust accordingly. I have seen numerous cable chain failures in industry, including abrasion through PTFE tubing, ratcheting, link failures, etc. You may already know all this, but I figured I'd share what I could

2. It looks like signal cables and VFD power cables are both running through the cable chain? If you have noise issues down the road, I might inspect here first. Low voltage square wave signals (5V) from your BOB/Mach3 are running in a cable parallel to the VFD cables powering the spindle? These may be even more susceptible to noise than a standard stepper pulse from a stepper driver considering the lower voltage.

Just some thoughts, it's exciting to see it running!

-UA

[wado1971]

@Wizard & jonathlee - Thanks for the links and suggestions. That gives me a bunch of materials to consider so I'll likely make a short list and try a few out.

[wado1971]

Great video and update, Wado! Your machine is awesome and I imagine what many of us will be modelling ours after...

Just a few things I noticed:
1. The cabletrack/cablecarrier/cable chain: Looks like the mounting attachment point on the Z-axis might be low causing it to bow up at the extremes - causing an additional bend point at the cables in the exit of the chain. I believe best design practice is to have cables come in/out parallel at the exit points of the cable track. This is probably fine for the use case, but it also appears that the cable carrier is full of cables, I believe typical design practice is to not exceed 60% occupancy by cross-sectional volume. At a greater than a certain percent fill occupancy, many cables will be pushed in or out from the neutral axis where they will be binding in either compression or tension with each pass leading to abrasion of the cable insulation. You will want to make sure that the cables in the carrier are firmly secured on the stationary side to avoid ratcheting of cables as well. There are also rules for when one should secure cables/tubing on both ends of the cable track, but they escape me right now. In any case, it would be wise to inspect the cables and cable insulation to look for any of this abrasion and adjust accordingly. I have seen numerous cable chain failures in industry, including abrasion through PTFE tubing, ratcheting, link failures, etc. You may already know all this, but I figured I'd share what I could

2. It looks like signal cables and VFD power cables are both running through the cable chain? If you have noise issues down the road, I might inspect here first. Low voltage square wave signals (5V) from your BOB/Mach3 are running in a cable parallel to the VFD cables powering the spindle? These may be even more susceptible to noise than a standard stepper pulse from a stepper driver considering the lower voltage.

Just some thoughts, it's exciting to see it running!

-UA

UA_Iron - thanks for the tips / suggestions on the cable tray. I've been subconsciously ignoring this a bit... looking at your post I realize I must address it. I definitely underestimated the required bend radii of the spindle cable and the amount of wires running thru it. I'll have to replace this little cable tray with something more appropriate. I have plenty of room to double the width (+height) of the cable tray which will give the VFD power some room.... or is it better to have VFD completely separate in it's own tray? - or that's overkill?

BTW - next project is to take your suggestion on changing ratio of the pulleys and get some more speed - I realize I'm not utilizing this design to it's full potential at the moment. I've gone through all the motion calculations in detail and adjusted my goals - In my original setup I based everything on a targeted accuracy I wanted (~3 tenths) and assumed 2 microstep setting of the MDrives, this resulted in a speed reduction ration of 0.66. As you pointed out, I was asking too many RPMs from the steppers and I realized that the motors run much better at ~8 microstepps. So... ratio will change to 1.25, still achieve my accuracy goals due to microstep at 8 (testing will tell), and push the top end rapids/cutting to double what I have now. Pulleys on order - will keep you posted.

[UA_Iron]

UA_Iron - thanks for the tips / suggestions on the cable tray. I've been subconsciously ignoring this a bit... looking at your post I realize I must address it. I definitely underestimated the required bend radii of the spindle cable and the amount of wires running thru it. I'll have to replace this little cable tray with something more appropriate. I have plenty of room to double the width (+height) of the cable tray which will give the VFD power some room.... or is it better to have VFD completely separate in it's own tray? - or that's overkill?

BTW - next project is to take your suggestion on changing ratio of the pulleys and get some more speed - I realize I'm not utilizing this design to it's full potential at the moment. I've gone through all the motion calculations in detail and adjusted my goals - In my original setup I based everything on a targeted accuracy I wanted (~3 tenths) and assumed 2 microstep setting of the MDrives, this resulted in a speed reduction ration of 0.66. As you pointed out, I was asking too many RPMs from the steppers and I realized that the motors run much better at ~8 microstepps. So... ratio will change to 1.25, still achieve my accuracy goals due to microstep at 8 (testing will tell), and push the top end rapids/cutting to double what I have now. Pulleys on order - will keep you posted.

Wado, I am living vicariously through your project here so I'm definitely excited by all your work - I just moved states so my project will continue to be pushed off indefinitely. As an engineer I always have to push for the super-most ideal application cases - maybe its an incurable disease!

Now that I got that statement out of the way: if you're not experiencing issues with the system, it might be best left alone...
Here is a great resource link for the application of cables within the cable chain and sizing: igus

Cables and hoses should always be strain relieved at the moving end and, whenever possible, at both ends. The exception is that hydraulic hoses should only be strain relieved at the moving end.

I recall why the quoted piece above is true now (for at least one reason): constraining fluid and pneumatic lines at both ends will cause volumetric differences in the tubing due to deformation - if I were dispensing fluid, I would have a different fluid pressure at different lengths of travel.

Are your cables shielded and specifically meant for flexing applications? A shielded cable not meant for flexing will most likely lose shielding coverage gradually over the number of flex cycles. If you have shielded cables you want the shield grounded to the chassis at both ends (or even more points if it is possible) to properly drain any undesired signals it is carrying - the end result goal is an equipotentially electrical/hardware bonded system. I have uploaded the best document there is regarding proper system grounding - it was given to me by a seasoned electrical engineer who designed elevator safety systems and then medical devices (medical devices being my background as well).

Best shielded cables for flexing applications: igus® Chainflex® continuous-flex cables: Chainflex® CF140US control cable PVC

In short, to answer your question, yes the ideal case is to run the VFD cable as far away from the signal cables as possible - if this means another cable track, then that may be the solution. Ideally they would never run parallel AND if they must cross-paths then they should cross completely perpendicular to each other. This ideal case just may never be a possibility due to geometry constraints, and you aren't even sure that there is a problem... so maybe it is best left alone at this point right?

There a lot of considerations when it comes to shielding, grounding and electrical systems, it can easily become overwhelming. I was (or am) in the process of designing an electronics control box for all my future CNC hardware... and it's been quite a ride configuring and reconfiguring to meet the most ideal electrical situation possible. If you are planning on building an electrical box i'd be more than happy to share my experiences and insight that I have learned along the way.


I still think you're being shy with the microstepping on those M-drives Unless of course, this is a function of your pulse generator (sorry, I didn't see if you were using a smoothstepper or just parallel port).

[wado1971]

Update on the build.... I've posted the latest and final version (v0.3) of CAD models & price list for this design on my website so go grab them if your interested (Direct Link). Below is a list of changes made from the last v0.2 files.

1) Added Enclosure & table axis protection cover for rails and screw.
2) Modified Home & Limit switches of X table axis to accommodate protection cover.
3) Changed pulley ratio on all axis to improve performance - increased axis speed to 280 IPM rapids & 220 IPM cutting (was 150/110). This new ratio also allows easy swap with SERVO motor alternative which will allow 370 IPM rapids and 350 IPM cutting.
4) Added electrical hardware - elect. enclosure with basic layout inside.
5) Move z-axis non moving geometry up ~1" - I was too low in original design because I was paranoid about stiffness in this area.
6) Added Mist Coolant system - a modified Wesco mist system with Noga articulating arm.
7) Fixed Cable Tray bend radius.

CNCzone was an invaluble resource during the build and for that I'd like to thanks the forum members.
Cheers
Wade'o

[v2tiller]

Your build is great.
I used it as inspiration for my build.
How are you liking your teknomotor?
I got one too. Not as powerful but im really happy with the quality and service from damoncnc. Like his custom collet tools.

[wado1971]

Your build is great.
I used it as inspiration for my build.
How are you liking your teknomotor?
I got one too. Not as powerful but im really happy with the quality and service from damoncnc. Like his custom collet tools.

Thanks v2tiller.
Really like the Teknomotor and you're spot-on about the support from damoncnc - I also had a great experience dealing with them. I'm not using the machine in a production environment and with no past alternative to compare to I can't give any definite feedback; but have been very happy with the spindle so far. Collet tools were a nice bonus.
Just checked out your build thread... looks nice. I'll post my questions over there.
Cheers,
wado

[footpetaljones]

I have to say that this is one of if not the best documented builds I've come across, so thank you.

I am wondering if you have had any problems with deflection of the spindle at full extension.

[wado1971]

I am wondering if you have had any problems with deflection of the spindle at full extension.

It's funny you should ask that question - I posted a new blog entry on my website about 2 hrs before you asked the question about this very subject. About a month ago I read an article from CNCCookbook titled/link "Set up G-Wizard for small CNC Machines". In the post he talks about spindle power and rigidity of the machine and that set me out to perform a test to determine how far I could push the spindle HP before I had machine stiffness issues (including the spindle deflection of course). I ran into an issue mid test with the steppers not able to handle the big cuts but so far I think I can say the spindle, and for that matter the entire machine, is holding up rather well. I'll likely have to spend some moola before I can finish the test but you can check out my progress so far (link to blog post).

[footpetaljones]

Was it your X axis that was stalling? If so, the weight of the Y axis might be an issue.

[wado1971]

Was it your X axis that was stalling? If so, the weight of the Y axis might be an issue.

Actually this router is a fixed gantry design so the Y axis is not riding on the X. I think I've accepted the fact that the steppers are not keeping up and will be getting Clearpath servos... I'm actually excited about placing the order... New toys!

[footpetaljones]

Doh. Meant Z axis, not Y.

[hacker7]

Soooooooo whatever happened with some of the issues with this machine?????
Did you figure them out ? And if you were to build another one would you do it this way?????