[davida1234]

Outlawtaz, I am not sure I can follow your thinking.

First, the 1515 is leveled to the X- rails by going with one single dial from
one end to the other. How thick the paper is, is irrelevant. I had to use
2 pieces of paper and a strip of packaging tape to make both ends the same.
Any tram issues are not important at this point. You just move the dial
end-to-end and make the 1515 totally parallel to the X linear rail.

Once set, now you can test the tram with the dual-dial.

[davida1234]

We need more speed

What's wrong with you, wasn't one inch per minute not fast enough?
Now you want two inches per minute?

- - - Updated - - -

[bhurts]

It looks like you used the same setup for the rotating ball nut as used on northwood machines. I posted a similar design here somewhere many years back. It looks really good. Very professional.

Ben


Sent from my E6810 using Tapatalk

[davida1234]

bhurts, thank you for your encouragement. I don't know
the setup of Northwood machines, could be similar.

This is 150 ipm:
https://youtu.be/UzcWURk_xhQ

[davida1234]

300 ipm:
https://youtu.be/C1JPkXVcRBk

[davida1234]

450 ipm:
https://youtu.be/PpC9ugCJ388

[Ntl]

Crank it up to 1000.

[Outlawtaz]

Outlawtaz, I am not sure I can follow your thinking.

First, the 1515 is leveled to the X- rails by going with one single dial from
one end to the other. How thick the paper is, is irrelevant. I had to use
2 pieces of paper and a strip of packaging tape to make both ends the same.
Any tram issues are not important at this point. You just move the dial
end-to-end and make the 1515 totally parallel to the X linear rail.

Once set, now you can test the tram with the dual-dial.

Yes and no. Let's see if I can put this into words... I will put out a hypothetical situation...

You place the 1515 right in the center of the table along the X axis. One piece of paper is needed to level it. The spindle tram is checked and found fine. Next, keeping the 1515 along the X axis, you move it edges of the table. On one side it requires 2 pieces of paper to level and the other side requires no paper to level. The tram is found fine in each instance. So, now you have three different "levels" and corresponding trams. Which one is correct?

This is what I mean by twist. While you were checking the tram in X axis, you have discovered there is a twist as evidenced by more or less shims needed to level the 1515. Even after skimming the table, it still wouldn't be flat. If you ran a large part it would show a 0.010" twist if removed from the machine and put on a granite.

Again, its probably overkill on a homebuilt machine. But, I drilled and tapped the base on my Bridgeport and made feet for it so I could take the twist out of it.

So you see, its more about my own OCD than your design. You just seem really educated and have really thought the machine design through. And like I said before, I have entirely too much time pondering stuff for my own good

[Outlawtaz]

And I need to qualify the above post. It could be that the table isn't flat and needs to be skimmed. But, unless something that is precision ground is placed across the linears, you could never know which was the problem. And if it is the linears themselves not being true, skimming the table will only continue the problem.

[davida1234]

Outlawtaz, now I see what you mean, thank you.

First of all, I wouldn't call it a twist, more like a slant or not level. A twist in my mind is something 3D.

Second, we don't know if the 1515 was dead straight. Could be a factor why I had to shim it. Third, the stringers underneath could be a little off, which would create that delta too. Or the simple fact that one Y rail ended up a bit higher than the other one during assembly. You should know that that frame was maybe 8-9 times fully disassembled/re-assembled while doing all the revisions. Looks like I didn't check everything close enough during the last assembly.

I tend to think that I ended up with one Y-rail being a bit higher than the other Y-rail. Particularly through the fact that in the front-to-back-direction, the 1515 didn't need any shims at all (not documented in my post).

IMO, that is an acceptable condition.

The only relevant thing here is that the 1515 was parallel to both the X-rail and the Y-rails. After all, the spindle moves on those rails. Which it was when I shimmed it. Once I install the spoil board, it will be a bit higher on one end than the other relative to the X-rail. Like 0.02" or so. Skimming it will make it parallel to both the X-rail and the Y-rails.

Let's see what else you got!

[davida1234]

But, unless something that is precision ground is placed across the linears, you could never know which was the problem.

Sorry, posted my response above before reading your new post.

Yes, you are right, the 8020's could be off a bit. Or the cuts of profiles, performed by humans. Or the linear rails could have an unrecognizable bend or hump.

I think this is a factor that has to be accepted within the realm of $2,000 to $5,000 DIY machines.

[GME]

Sorry, posted my response above before reading your new post.

Yes, you are right, the 8020's could be off a bit. Or the cuts of profiles, performed by humans. Or the linear rails could have an unrecognizable bend or hump.

I think this is a factor to have to be accepted within the realm of $2,000 to $5,000 DIY machines.

Little known fact: 80/20's cut tolerances aren't all that precise. In the catalog, they advertise a cut length tolerance of .015" and a squareness tolerance of .002"/inch maximum. See page 850 of the current 80/20 catalog (Catalog 22). You won't find the spec anywhere else on their site. I've bought my share to 80/20. I've never received a cut piece that was better than .002'/inch for squareness, and it's sometimes worse. I've seen it as bad as .005" in 1.5".

In the last few months, I bought enough 80/20 to build a 4x4 of my own design. Virtually all of it is the 3030 profile, except for the gantry, which is the metric version of the 3060. Every one of the 40 cut ends was off the full .002"/inch for squareness. I haven't bothered to measure lengths yet, because I am going to mill the ends square. I'll ensure all of the lengths are equal during the milling process.

You have a very well thought out and executed machine. I wish I had the background to design to your level of sophistication. But then, my machine will use rack & pinion drives, which were pretty straightforward design wise. Your ball screw drive appears to work well, and as far as I can tell, you don't have any whip in the screws. Very nice. Correct me if I'm wrong, but aren't you at, or very near, the length limit for screws.

I particularly enjoy following your mill design work. Do you have a designed CNC mill for cutting, or a conversion? What make?

Gary

[davida1234]

Gary, thank you for your encouragement and comments.

I've never received a cut piece that was better than .002'/inch for squareness, and it's sometimes worse.

I found out as well that one could and should not
rely on squareness or cut lengths of extrusion pieces.

Therefor, I designed the frame such that any tolerances
have no effect on the squareness of the machine.

my machine will use rack & pinion drives, which were pretty straightforward design wise.

It is true that rack & pinion drives are quite easy to integrate.
One of the reasons I pursued a different path is that I am
not comfortable with their power transfer situation which
creates different load conditions for one direction and the
other. I also wanted to avoid protruding motors.

Your ball screw drive appears to work well, and as far as I can tell, you don't have any whip in the screws. Correct me if I'm wrong, but aren't you at, or very near, the length limit for screws.

I am not driving the ball screws. I am driving the ball nuts
which I call "power ball nuts". As the screws are not rotating,
there is no limit on length of ball screws.

Do you have a designed CNC mill for cutting, or a conversion? What make?

I am cutting everything on my previous DIY machine.

[mactec54]

Little known fact: 80/20's cut tolerances aren't all that precise. In the catalog, they advertise a cut length tolerance of .015" and a squareness tolerance of .002"/inch maximum. See page 850 of the current 80/20 catalog (Catalog 22). You won't find the spec anywhere else on their site. I've bought my share to 80/20. I've never received a cut piece that was better than .002'/inch for squareness, and it's sometimes worse. I've seen it as bad as .005" in 1.5".

In the last few months, I bought enough 80/20 to build a 4x4 of my own design. Virtually all of it is the 3030 profile, except for the gantry, which is the metric version of the 3060. Every one of the 40 cut ends was off the full .002"/inch for squareness. I haven't bothered to measure lengths yet, because I am going to mill the ends square. I'll ensure all of the lengths are equal during the milling process.

You have a very well thought out and executed machine. I wish I had the background to design to your level of sophistication. But then, my machine will use rack & pinion drives, which were pretty straightforward design wise. Your ball screw drive appears to work well, and as far as I can tell, you don't have any whip in the screws. Very nice. Correct me if I'm wrong, but aren't you at, or very near, the length limit for screws.

I particularly enjoy following your mill design work. Do you have a designed CNC mill for cutting, or a conversion? What make?

Gary

When using a Rotating Ball Nut there is no length or diameter that is affected, you can use any length ballscrew that is manufactured, the Screw is fixed and does not rotate, this is a common practice, in machine building,the manufactured Rotating Ball nuts are expensive that is why you don't see it on many hobby machines, but those that get creative make there own Rotating Ball Nuts and the cost then works out to be not much different than from the mounting of a regular Ballscrew

[wmgeorge]

Interesting. But that means the motor is traveling along with the wiring. Correct? I guess its all the same. Duh I just watched the video. Novel and good idea but it looks like the motor and gear drive size is limiting the travel of the axis. BTW Excellent job.

[davida1234]

Thanks mactec54. You are right, it is somewhat close to the cost of a similar ball screw + end supports + coupler.

Interesting. But that means the motor is traveling along with the wiring. Correct?

wmgeorge, yes, the motor is traveling with the carriage.

The same as with rack & pinion. Or any gearbox direct drive.

I think that the rotating ball nut combines the advantages of
rack & pinion drive (unlimited lengths, higher rpm, low rotational
mass, compact package) and ball screw (smoother operation,
less wear and noise, probably better backlash).

it looks like the motor and gear drive size is limiting the travel of the axis.

Sorry but, no, it doesn't. Where do you see that?

Btw, the rotational mass of the power ball nut is just
4.4 ounces compared to several pounds of a ball screw:

[wmgeorge]

David since I have never seen the complete side of the machine with the ball screw drive as above I had assumed the drive needed to be inside that frame. In any event , I am like Dan I am impressed enough to want to convert my R&P over to the ball nut drive system like yours!

[davida1234]

wmgeorge, thank you, what an endorsement!

As you can see in the pic below, the power nut
installation is very short. I don't think anybody
has a carriage narrower than that:
Attachment 415532

[Outlawtaz]

First of all, I wouldn't call it a twist, more like a slant or not level. A twist in my mind is something 3D.

Yes, it is a 3D problem. I actually had a customer that cut some insanely small profiles on parts and I went in every six months or less and checked / adjusted the machine. They were quasi threads that were 0.005" wide and 0.003" deep. I used a ballbar tester that measures down to the micron to make very tiny adjustments. Again, well outside the precision that anyone shoots for on a homebuilt

I just haven't seen anyone show their tramming before so it just came to mind. Mostly just curious about your thoughts on it...

Love the videos of power up. Can't wait to see you cut some actual parts on it. Now I'm curious to see feeds and speeds, depth of cut, etc...

[mactec54]

Again, its probably overkill on a homebuilt machine. But, I drilled and tapped the base on my Bridgeport and made feet for it so I could take the twist out of it.

He already has adjustable feet on the base of the machine, and most do, not over kill it is need for any type of machine build, talking about Bridgeport mills I have done that also to dozens of Bridgeport machines, mainly just for leveling more than anything else

On his machine build he could easy adjust any of the feet if he as a twist problem, on these extrusion builds you have to adjust between twist and level

So if you make it level and it has a twist, then you adjust for the twist then it won't be level, so you have to figure what part is the most important