Benchtop Monster is Born - Harbor freight 44991 / 33686 Hybrid CNC Conversion!!!

[BobWarfield]

In the above discussion, the pitch accuracy is what is used to determine the AMOUNT of float that will be necessary. The whole point is several posters were commenting that there was no float, but in actuality, the system isn't nearly stiff enough to prevent the float that is needed - therefore the problem that was theorized is non existent (which is consistent with A.S.' experience).

There are two reasons for float.

One is as you and A.S. have said: to deal with irregularities in the ballscrew. I will grant the 125 millionths figure and forget about this issue.

But the second is to deal with inaccuracies in the bracket that fixes the nuts together. Consider a typical wavy washer from McMaster Carr: 9715k92

It has a total deflection available of 0.069". That's about 120 degrees of rotation on a 0.2" lead ballscrew, so the couple of degrees of rotation the flex allows is fine. But the accuracy of those fasteners on that aluminum strip is pretty iffy. Are the forcing the ballnuts closer or further apart? Do you have 100lbs of preload, 75, or 50?

I'm sure there is probably some preload there and things are working relatively well. It's just that's it's extremely easy to make a floating assembly and it makes things a bit more predictable. That's a pretty light mill, so the cutting forces will likely never overcome whatever preload is available.

To each his own. The mill is working, its a nice simple assembly, rock on!



Cheers,

BW

[HawkJET]

But the second is to deal with inaccuracies in the bracket that fixes the nuts together.

You got me on this one Bob. As a master craftsman mold maker, I forget that not everyone has the skill to work to the tolerances I routinely use.

From your comments, I see there is an implication that one could build their anti-backlash system this way. But if they don't get it just right - it could be trouble.

Yes Bob, your point is well taken. Using a different system can make the required tolerances much easier to manage.

[SScnc]

A.S.,

How long did shipping take to get this table from China ?

And I use flood coolant, are there "acordian style" way covers available for this size table that you know of ?

[CROSSHATCH]

Keep us updated, Love the work!


-Jason

[Crevice Reamer]

A.S.,

How long did shipping take to get this table from China ?

Already answered:

It took about 7 weeks - I ordered parts by calling harbor freight 800 number, they shipped them from factory to warehouse, and delivered on a semi.

CR.

[asammons]

The truth is, as was stated earlier, deviations and inaccuracy of these screws is insignificant - rolled screws are smoother then ground screws, and the rolling process by default is repeatable over short distance. The preload is far greater than what the bracket can resist due to lack of stiffness. Can the design be improved? Sure it can... Will it yield a better performance? Not at these loads...

Keep it simple, stupid (KISS).

As for springs, I have used two 81lb springs in each assembly (1/16" deflection of full load). McMaster part number 9717K62.

[SCzEngrgGroup]

That's all well and good, but I agree with Pete. Using the mounting holes from the return tube looks like it will constrain the motion of the nuts such that they aren't allowed to float back and forth lengthwise on the screws. If they can't do that, then they might as well be rigidly attached to each other and ditch the springs.
A detailed picture would be worth a 1000 words here, as it's possible I'm missing something.

You see, if I LOOKED at the pictures it might be better. I reaffirm what I said, it constrains the system in a way that isn't the best. You want the second nut to float axially while being constrained radially. I used a dowel pin against the side (two actually, but it only technically requires one).

Here's how I did mine....

Regards,
Ray L.

[escott76]

Here's how I did mine....

Regards,
Ray L.

Nice! Very simple and compact.

[asammons]

Here's how I did mine....

Regards,
Ray L.

Looks good! How much force are you applying between the two nuts?

[SCzEngrgGroup]

Looks good! How much force are you applying between the two nuts?

I don't recall exactly, but in the neighborhood of 250 pounds. These are 1" screws on my knee mill. This arrangement was trivial to machine, and works very nicely.

Regards,
Ray L.

[acondit]

Here's how I did mine....

Regards,
Ray L.

I don't see any Belleville washers or springs for tension. How do you deal with any inaccuracy in your screws?

Alan

[SCzEngrgGroup]

I don't see any Belleville washers or springs for tension. How do you deal with any inaccuracy in your screws?

Alan

There's a stack of Bellevilles inside that aluminum sleeve between the two nuts. They're not in the pictures, because I hadn't received them yet when I took the pictures.

Regards,
Ray L.

[caleb105]

Sammons,

How did you keep the ballnuts from rotating? Looks like you crossdrilled holes in the coupler....were those for set screws? Wouldn't that be hard on the threads?

[SScnc]

I'm going to be ordering this table from HF shortly. I currently have 270 oz-in steppers on my X2, will those still work reliably with this setup considering they are direct drive right now but with this new table there will be some reduction with pulleys ? Or do I really need to go up to more like the 480 oz-in ones like asammons has used ? thanks

[jfseaman]

I saw on LMS that you can get an X3 head and column. Anyone though of putting an X3 on this table?

[Crevice Reamer]

I saw on LMS that you can get an X3 head and column. Anyone though of putting an X3 on this table?

With a 20% off coupon you can buy the whole MILL from Harbor freight for $50 more than what LMS wants for just the column/head.

An X3 or SX3 column should bolt up even easier. If I added my SX3 column with 18 inches of Z to this base/table--Now THAT is a mill!

Yeah, I'm thinking about it.

CR.

[CommandoJoe]

Great job on the project - I have a couple of questions:
So this is basically just an X3 table on an X2 mill right (Not putting down your project, just simplification)?
For someone who doesn't already have mill to work with (me), would it be better to go this route or just buy an HF X3 for $720 ($899 - 20%)?
Is there any advantage to the X2/X3 hybrid over a "normal" X3, or did you create this to expand your current mill's capabilities vs buying a new mill?

Thanks

[jalessi]

Joe,

Thats not a X3 table, its a Harbor Frieght 33686 round column mill table.

Its much bigger than a X3.

Where did you find a X3 listed for $899? Please post a link.

I have the 20% percent off coupon however the current price for the x3 is $999

http://www.harborfreight.com/cpi/cta...emnumber=93885

Jeff...

[LeeWay]

There are several ways to handle the double nut system. I had plenty of room on my lathe and router builds, so I went with simple compression springs.
I mill a channel in some aluminum and fixed this fast to the first nut. The second nut can't turn, but can move relative to the first. I think I have about 95 pounds of force on these. I don't see any backlash on my lathe, but do see slight irregularities on the router. Likely due to different types of bearing mounts.


http://www.cnczone.com/forums/attach...2&d=1230561899

[kanton]

Is anyone willing to make up a kit and stock the parts to help a newby do this conversion?