Rong-fu (round column) mills

"Michael Koblic" wrote in news: snipped-for-privacy@news2.newsguy.com:

You aren't missing anything. You have to find a way to re-register after moving the head on the column. Some people modify the mill to have the head keyed to the column. I read somewhere (HSM probably) Someone drew a plumb line on the wall opposite the mill. They had a laser pointer mounted on the mill head with the laser dot pointed to the line. When the head was moved the alignment was good as long as the dot remained on the plumb line on the wall (and the mill level). It's suprisingly accurate if you do the math.

I used one for a while and wasn't highly impressed, 0.5mm accuracy was about its limit. The quill was tilted, the table grooves had been milled at an angle and the fine downfeed was almost impossible to adjust closer than 0.005" because it shifted when locked.

It was a good-enough drill press and accurate enough for the electronic prototypes and control panels I was making on it. It was noticeably stiffer than my Clausing, removed metal faster, and had a larger work envelope. The table travel was enough to reach any point on a 19" x 5" relay rack panel. The 5" quill travel was enough for what I was doing, as long as I pre-tested the reach of all the tools the job would need. The face mill cut steel smoothly with the quill fully extended and locked.

If your sundials are no more demanding of accuracy than control panels it might be OK. It was NOT good enough to make close-fitting machine parts, I took those jobs home to the Clausing.

jsw

That sounds like a Rong-Wae.

I used one for a while and wasn't highly impressed, 0.5mm accuracy was about its limit. The quill was tilted, the table grooves had been milled at an angle and the fine downfeed was almost impossible to adjust closer than 0.005" because it shifted when locked.

It was a good-enough drill press and accurate enough for the electronic prototypes and control panels I was making on it. It was noticeably stiffer than my Clausing, removed metal faster, and had a larger work envelope. The table travel was enough to reach any point on a 19" x 5" relay rack panel. The 5" quill travel was enough for what I was doing, as long as I pre-tested the reach of all the tools the job would need. The face mill cut steel smoothly with the quill fully extended and locked.

If your sundials are no more demanding of accuracy than control panels it might be OK. It was NOT good enough to make close-fitting machine parts, I took those jobs home to the Clausing.

jsw

I've got one of these and it's very difficult to get any sort of accuracy in "Z" without a digital scale. Once you've added the scale then plot the digital readout against the mechanical dial, you get to understand three problems - backlash, the quill sticking and non-linearity of something...

Steve

I felt it would be a decent hobby mill for the price with the inaccuracies fixed. Someone here once described shimming the column base to square up (tram) the Z axis. I wedged the locating key I made for the vise tightly into the tee slot and milled shoulders in its top edge, automatically parallel to the X travel, then slotted the underside of the vise be to a press fit on it. The table itself indicated substantially tighter than 0.001" over its full travel.

It wasn't mine so I didn't take it apart. Do you think a cleanup would fix the quill feed and lock?

jsw

This laser alignment idea is someone's ill-conceived fantasy. You don't get any accuracy without projecting the spot a long distance. But laser pointer spots are big blobs over such a distance (many thousandths), so there's no accuracy to eyeballing a blob onto a plumb line.

A) You can reduce the size of the dot with a tiny hole in a thin metal plate. (Also the brightness.)

B) Consider that the spindle is perhaps 8" away from the column (not too unlikely, given the size of the machines which typically have round columns).

Assuming that the wall is perhaps five feet from the column (60") then 0.001" at the spindle becomes 0.060" at the wall. (And five feet to the wall behind the operator is not that unreasonable, especially if there is another machine tool, such as a lathe, on the other side of the aisle.

Enjoy, DoN.

Uh no. The accuracy of the spindle is the ratio of the column to spindle divided by the column to wall. For your example of 8" and 60", .001" on the spindle would be .0055" at the wall.

That is why you use a Cross + split beam that draws a line up and down. I have a short level that shoots that out the end. I mount it on the head once I have it trammed - My professional vice is set - I set a marker.

The real trick is looking 30' for a line. I figure spotting scope or binoculars.

Mart> Charles U Farley writes:

Mirror on a stable stand or wall (the surface table would do in my shed). Then you can look at the wall close to the mill :-)

Mark Rand RTFM

Drill / Mills are an economical solution for many hobbiests. They are stiff compared to minimills. And will handle large pieces.

As far as the round column..........Almost all of the time it is no problem. You do not have to raise or lower the head very often. When you do have to raise or lower the head, a lot of the time losing the x- y position makes no difference. And the few times you need to keep the x-y position, well you just swing the head so it is back in the same x-y position. If you get the y position anywhere close, the x position is right on. You only have to get the y position correct.

Dan

A mirror that reflects the beam back to a target attached to the laser should work, with doubled sensitivity, and be visible up close while you are tapping and tightening the head. You could confirm the flatness of the mirror by double checking with a square and test indicator as you move the head down the column.

I just tried it over a 10' path using a 50ths ruler as the target and equalizing the spot width on both sides of a 1" line, which is easier than estimating the center. I have to hold the button on this laser pointer so it wasn't as steady as possible, but I could align it within 0.050 easily, translating to about 0.005" at the quill. I think with a clamped instead of hand-held laser I could center the beam repeatably to 0.020", 0.002" at the quill.

I should have written that the RF-31's Z axis clamping shift was

0.1mm, not 0.5.

jsw

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 Dan

I just remembered that when I had to move the head I centered a small hole under it first and then afterwards located the head on that hole with a center finder. The center finder shank fits in a collet when the head is down and a drill chuck when it's up, effectively adding the height of the chuck to the accurate Z axis range of the machine.

jsw

Here's a link to some articles written by an RF30 user:

Simple, no laser required.

Dan

Depends on the laser, of course. One can get under .050" spot size, and with two yards distance to wall (there should be at least one wall that far away) that angle accuracy is 0.04 degrees. I'd be tempted to use one of those hologram-symbol pointers, and get all the lines in a star to match up to the wall engraving; when it's spread over a large viewing area, the alignment would be easier to eyeball.

[ ... ]

O.K. -- not my 0.060", but not your 0.0055" either. I get

0.0075" when I recalculate it a couple of different ways. So -- add a mirror on the wall behind the operator, and bounce it to another wall more distant. Sum the distances.

Enjoy, DoN.

The mirror has to be *very* flat over the height of the column travel. You might find one that's good enough in a scrapped scanner.

jsw

Don't be naive. Pinholing a big blob of light doesn't make it any more pointy. You can't improve coherence outside the device, at best you preserve what exited the laser device aperture. A pinhole will just diffract and degrade coherence.

All of the follow-ups to date of my critique ignore the optical principles that prevent this entire scheme from working. Specifically, Gaussian beam optics, beam waisting, and beam divergence. Laser beams are narrow but they still start at millimeters wide and spread out to inches as they project. Have a friend stand 50 feet away and see what kind of spot you get on a target from a good-quality laser pointer.

A simple scribed vernier scale with mechanical pointer lever and magnifier would work better.

Lasers do not create precision out of nothing. They have to be attached to precision mechanisms, which isn't the case for this axis on a round- column mill-drill.

You're essentially trying to build a precision transit or goniometer without a mechanism or scope. You need more than just a laser pointer.