View Full Version : Overarm Support for Atlas Mill

11-29-2010, 10:07 AM
It's time to ask for some input from the vast knowledge base here. My current project is bringing an Atlas MF horizontal mill back to life. The model I have did not come equipped with an overarm support and unfortunately the design won't allow me to adopt the factory designed type because the mounting area on the Z control on the knee is tapered. Apparently models that came equipped with overarm supports had a straight bearing mount for the Z axis control. So, here's my question: rather than have the overarm support go from the Z axis control to the overarm, would it work to make supports that go from the bench which the mill is mounted on (very solid) to the overarm in an "X" fashion? In other words, is it important to tie the knee to the overarm or just to support the overarm so it can't flex under load?
One advantage of having the supports go from the overarm to a mount on the bench is that they can be fixed. The knee can be moved up and down freely without having to loosen anything. Another advantage is that the already small Z axis travel available on a mill of the type isn't further limited by overarm supports that attach to the knee.
If possible I do not want to modify the mill itself from it's original configuration. I can easily make a collar to clamp on the overarm to which the supports from the bench attach.

11-29-2010, 10:45 AM
I can't decipher your post too well. The overarm support had a hole on it that went around the shaft of the Y axis handle,where it has a straight area right up against the knee. This would help keep the mill's table and arbor from separating in a vertical direction, but would do nothing for the sideways racking motion of the table and knee.

I started out with an Atlas. My advice is to find a REAL mill. My next one was a little Burke #4. A small machine. Table was about 4" x 18",with 1 T slot,but it was a real machine that would take decent cuts in steel. It wasn't so heavy that a few guys couldn't carry it.

11-29-2010, 02:39 PM
I like those little mills but as gwilson noted, they aren't the sturdiest machines around. I think that the biggest argument against trying this is the fact that these machines have been around for a l-o-n-g time and I'd expect that we'd have heard of a modification like yours by now.

However, if time is available, why not put some time into improving performance ? There's nothing to be lost because even if the idea doesn't work you will have gained some knowledge (which hopefully can be shared with others).

An obvious disadvantage is that the bench and the machine move with temperature (and with humidity if the bench is made of wood). Another far-fetched but possible disadvantage is introducing a mechanical resonance into the system that would affect the quality of the cut. I'm sure that you've considered some of these problems and I think the idea might work, for many applications.

If you have set up the machine and worked with the mill for a couple of hours and then needed to leave the setup overnight, there might be problems as the different materials moved and introduced stresses in the setup. However, if you re-established the relationship between all of the parts before commencing work the next day, this is probably a minor concern.

I like the idea of moving the knee without disturbing the overarm support. If the knee has a fairly beefy cross-section and a good bearing surface then I would suspect (as you are suggesting, I think) that the overarm is the more flexible of the two parts.

In any event, there should be a non-permanent way of testing the concept. Why not jury-rig something up, even if it takes a couple of days ? If the temporary supporting system shows promise, then you can tackle a full-up design.

Good luck, I hope that your idea works and that you can post your results - whether they are good or not so good. (It could easily help others in your situation or a similar one.)

Randy C

11-29-2010, 03:07 PM
Tha Atlas milling machine and shaper are stouter machines than the lathes, and are quite useful for their size.

Rudy Kouhoupt had several articles on recommissioning an Atlas milling machine. This video shows his machine and the support he fabricated. It might not be too difficult to adapt this style to your machine with the tapered control.


I sure do miss Rudy.

11-29-2010, 04:50 PM
Own a yet unused (by me) Atlas horizontal, the version where the overarm support was not part of the package.
My gut says turn a fitting that is identical in every dimension except for the exterior taper of the factory piece, there is about 27/32nds if you keep close to stock dimensions (between the hex and the lip closest to the operator), you should then be able to make something similar to the support in the video. In fact you could probably gain a little more if one made a counter sunk area for the base of the support that fit over that lip/edge. That is the plan since cutting a good/accurate taper is beyond me just now.
IIRC over on the Yahoo! site there are some drawings of various support pieces (dimensioned I think?).

Mine is a version that had a "stylistic" cover that looks OK but as far as I can tell adds little in terms of support and I am debating if there is value in somehow adding more support (as in another very small dimension overarm). That was part of what I found interesting about randyc's description of his Leitz mill.

I don't have all the pieces in front of me but how far out of plane are the mounting holes in the base (to fasten the whole machine down) and could there be something there that would be useful...?

11-29-2010, 09:58 PM
Thank you to all who have responded so far.

Gwilson, I apologize if my post wasn't clear enough to be understandable for you. I appreciate your input, unfortunately I have limited space right now and already own the Atlas. The original overarm support you mention would be easy enough to fabricate if my mill had the correct Y axis control for that option. Apparently my mill was manufactured before the overarm support was available and the Y axis control has a tapered housing negating the possibility of fitting a support like the original.

JCHannum, Thank you for the positive comments about the little Atlas and the info about Rudy Kouhoupt's writing. Contrary to Gwilsons's opinion the Atlas seems to be held in fairly high regard by many people. I have all of the HSM issues pertaining to the Atlas Mill to which Mr. Kouhoupt contributed. The overarm support he made is a replacement for the factory item and his mill was equipped with the straight Y axis control. The difference between mine and his became apparent when I read his articles and made comparisons.

RussZHC, with regard to the Y axis control on mills like yours and mine, it did occur to me that I might turn the control housing straight to accommodate a support similar to the factory item. Research has led me to learn that the Y axis control is in fact completely different on machines that will accept the factory support. The Control Housing and ultimately the screw are longer to allow the table to move it's full range of motion with the support installed. When you get around to using your Atlas take a look at the table position in comparison to the Y axis control and you will see that there is no room for the support with the table moved toward the control handle to it's farthest extent. A possibility would be to make a new housing and screw; a considerable effort and a modification to the machine which I don't care to undertake unless all else fails.

Randyc, thank you for your comments and encouragement. I believe I have a plan that I can put in effect that will not alter the machine nor will it take a tremendous amount of time or material to accomplish. The plan goes something like this: Place a heavy steel bar beneath each end of the mill, perhaps 1" by 2", and bolt the mill through the bar to the steel bench. The bar will be longer than the base of the mill by a few inches. The area of the bar that extends beyond the base will be drilled and tapped to accept bolts that hold two supports in an inverted "V" fashion that will attach to the overarm at their apex. A clamp arrangement similar to the bearing support that fits the overarm will be attached to the overarm at it's outward end and the supports bolted to it at their apex. Spacers as necessary will keep them outboard of the Y travel at it's farthest outward extent.
My hope was that someone could provide me with personal experience in this area and knowledge of the importance of tying the table and knee together. The knee is supported by a pretty substantial screw and substantial ways. As you suggest, experimentation may be the only option.

Thank you to everyone for your input, when I complete my experiment I'll provide an update.

11-29-2010, 10:13 PM
Understand that I USED one for about a year. They are o.k. for brass,but mine wasn't worn out at all,and was a big pain to use for any cutter wider than a slitting saw. They are popular with hobbyists,and always fetch much more than they are worth,usually about $1100.00 for one in good shape. For much less than that,you could get a Burke #4. Probably for about $400.00.

I sold mine to a clock maker,who ONLY used brass. It was fine for him.

I think they are popular with hobbyists and older people,because they are afraid of weight,and have limited space. I knew Rudy. He had a TEENY little shop. I also think they are popular with hobbyists who haven't used a heavier machine that can really cut. The Atlas is a cool looking little machine,because it looks like a scaled down full size one.

I do not mean to be offensive about your mill. What I'm saying is just factual,and can save you a lot of trouble and hours of piddling with ANY Atlas product,except the drillpress. If you can find the #4 Burke,or other small,heavier mill,sell the Atlas and get it. You will see a World of improvement. I did.

11-29-2010, 11:19 PM
That is a plan that is far cleverer than anything I could have devised, although it's not hard to out-think ME, LOL. I think that you're on to something because the design takes the workbench out of the scenario, if I'm understanding your concept. (The only comment I could offer might be to replace the 1 x 2 rectangular bar with a section of channel, something with more vertical stiffness.)

Good thinking !
Randy C

J Tiers
11-29-2010, 11:58 PM
First, No, going to the machine table is NOT, repeat NOT, the same thing in any way, shape, or form.

It may do something, it can hardly hurt, but it is NO SUBSTITUTE for a proper overarm support.

the way the forces are oriented on a mill is as follows:

There is an up and "back" force on the cutter, which is partly supported by the spindle bearing, and partly supported by the overarm. The "reaction" to that force is down and 'forward", into the table and knee via the workpiece. The up and back force is substantially into the overarm, which is not actually that stiff.

The only "connection" between the up and down forces is via the column, which supports both the knee and overarm. Quite a lot of "prying" force can be exerted in a direction that wants to tear the knee and its base-mounted screw support off the column, or break the overarm off the top of the column.

Spring in the overarm, and the knee/base/column system allows the cutter to push away from the work.

By tying the overarm DIRECTLY to the table, this set of forces is more "confined" to the overarm and table system. The knee and its dovetail is no longer the only part involved, and the "prying" force is much reduced.

The upward prying force (and some of the "back" force, depending on support type) is "carried into" the knee, table and workpiece directly. The system is hugely stiffer, and a much larger cut can be safely taken.

Mr Wilson is an excellent craftsman, and I respect him. But he cannot have ever used the machine with a proper overarm support, or he would never publicly make the statements he has made about the Atlas being a "brass mill only".

A relative had an Atlas, and I have a similar sized mill. The difference that tying DIRECTLY to the knee has made is profound.

As for how to do it........

I do NOT think the Atlas/Kouhaupt system of using the crank bearing is worthwhile. it supports the "up" force, but does not address the "back" force adequately. Study of most heavy duty commercial machines systems will show that they usually brace for both the "up" and "back" forces.

What I did was to tie to the knee dovetail on my (non-Atlas) mill. A clamp bar slides over the dovetail and crossed arms forming a triangular support go up to the end of the overarm. The cutter shown is a 2" wide slab mill taking a substantial cut in a steel workpiece.


A posed overall picture taken some years ago before some modifications to allow more table travel by reducing the size of the clamp. Cutter is merely representative, you'd never need the support with it.


11-30-2010, 12:43 PM
Well, there are a few reasons for NOT doing it, LOL. Doubtless nobody would argue that attaching knee to overarm is the best technique, it's fundamental that a closed section is always more rigid than an open one. But let's be fair - shouldn't we make the initial assumption that firbkrhd considered these issues ? I got the impression that he did and that he also had considered a drawback or two to the X-brace configuration.

I think that he has come up with the next best idea - other than the fact that I've never seen it done, I can't find fault with it, LOL. Mind, I can't speak for the OP but since I am GUILTY of encouragement, I feel that I should say something in support of his project :) (And of course we are bystanders, it's easy to offer advice when our time and money are not involved.)

In any event, I'll look forward to the project, should it continue - the results should be very interesting.

Randy C

11-30-2010, 02:04 PM
The Control Housing and ultimately the screw are longer to allow the table to move it's full range of motion with the support installed.

I have an Atlas mill and I have the original overarm support casting. I am missing the yoke that binds it to the overarm but thats a longer story.

I don't agree with your statement above unless your mill is different than mine. Currently they sound similar if not the same year (maybe).

I have the tapered Y axis spacer that supports the Y feed screw. I can attach the over arm support casting to it but the rigidity is poor due to the taper behind the od of the support nut. When it is attached the Y movement is within the design of the machine. The saddle controlling the Y axis was not meant to hang over the support spacer. And if it is moved into its furthest position, there is room for the support on my mill as well as room to work the X axis lock and the Z axis handwheel.

The support has a 0 with an indication line on it for reading the Y handweel because the support uses up all of the Y axis spacers od.

I have a photo that I took today of my project in mid stream. I am making another Y spacer to fix the problem. If you have the overarm support casting why not make the Y axis spacer. Its not a difficult project and it will fit the look of the mill as well as give the mill more rigidity.

If you do not then a cross brace support as described in this thread may be your best choice because your starting from scratch.

Click for larger photo.
http://i4.photobucket.com/albums/y128/katiecat222/projects/tools/th_IMG_8997.jpg (http://i4.photobucket.com/albums/y128/katiecat222/projects/tools/IMG_8997.jpg)


J Tiers
11-30-2010, 10:32 PM
Well, there are a few reasons for NOT doing it, LOL. Doubtless nobody would argue that attaching knee to overarm is the best technique, it's fundamental that a closed section is always more rigid than an open one. But let's be fair - shouldn't we make the initial assumption that firbkrhd considered these issues ? I got the impression that he did and that he also had considered a drawback or two to the X-brace configuration.

I think that he has come up with the next best idea

"Next best"? maybe so.

Better than nothing, certainly. But it appears to be an "around the barn" solution to what is fundamentally a simple problem.... the prying stresses created by heavy cuts. I suspect that the inclusion of more parts, arms, etc in the system also allows more flex, more stress on the rest of the mill, and more chance of breakage.

It isn't that it won't work, it is just that it does what I consider is provably a worse job of avoiding stress on the mill. That is, after all, why the supports are made.... to allow doing a heavier job without exceeding the safe stresses on the unit, by containing the forces and keeping them as much as possible straight-line forces with no "prying".

However, looking at the picture.... apparently I have forgotten how short the Atlas knee is... with the overarm support on the dovetail, it looks as if the arbor would have to be short and the crossfeed would be limited.

The pictured single arm support would be better than nothing, and allows the maximum crossfeed distance. Alternately a bracket for double arms could be attached to existing bolts in the knee, but the Atlas is singularly lacking in any visible boltheads on the front of the knee.

Posibly at the end of the day, the single arm system is as good as the "bottom beams and tension rods" approach, and much simpler. if the existing potential seat is unsuitable, a new one might be made, or the existing one modified, at least as easily as the more complex system could be made.

12-01-2010, 12:26 AM
Firbkerhd1: Thanks, I have most of the sub-assembles within easy reach (won't say what my apartment looks like!) but pieces are apart so the movement of the table "into" the support would not have dawned on me.

That said, and I am certainly not doubting your statement as every photo I can find (I could not believe I missed something that
obvious :o ) clearly shows a much extended piece, I came across this


not sure if someone can link directly, so that is a jpeg file called, "Dials for mfb mill" on the Yahoo! Atlas mill site, and one of the parts of this individual adding larger dials was to make that bushing with straight outside walls specifically for fastening an arbor support yet a photo shows it placed next to the stock item and they are the same length...so I am very confused :confused: :confused: as what you said makes so much sense and even in Rudy's video is is plain that bushing is longer than some models had stock.

As a related "aside", at one point someone was supplying all the pieces to "upgrade" if you did not have the arbor support, apparently Atlas back in the day offered a 'retro fit' kit with all the pieces and someone in 2009 decided to more or less duplicate that kit but there were almost no sales.
I am waiting for an answer as there was mention of a set of drawings as well...I'd be "after" those.

12-01-2010, 10:26 AM
Thank you to all who have given input. You have given me many things to consider and more research on my part needs to be done before I move ahead. The input of each poster has given me new insight on other lines of thought through which I may eventually come up with a "best" solution, which is what I had hoped for when I originally posted. I also have other projects underway (back gear knob for the mill) at the moment so I have time to contemplate my next move.

JTiers, what you have said is sound and makes a lot of sense. I suspected as much or I probably wouldn't have posted my original question. Now in consideration is making a new cross feed housing with straight rather than tapered area while maintaining the original length. I may be able to do one of two things with this. First, I may be able to make an offset support similar to the factory item, offset away from the knee to allow full movement of the cross feed or an even more rigid support incorporating the cross feed with the X bracing; perhaps incorporating something like you have pictured on your mill but continuing the braces on to pieces at the base of the machine as originally planned. It seems that this would tie everything together in a most secure fashion. Some time measuring, drawing and tinkering will help me decide whether this is a workable solution.

Rockrat, you may be right about your assertion, I have only pictures to go by, however the Atlas parts manual does list different part numbers for the control housing on different models. The parts drawings aren't clear as to whether they are tapered or not. It may well be that the control housing is the same length for units that were equipped with overarm support and only the housing was changed from tapered to straight. Unfortunately I do not have the original overarm support so I would be making the entire thing from scratch.

RussZHC, unfortunately I was unable to use the link you provided but I believe I have seen the picture to which you refer. I believe it is one that incorporated some dials form a mill drill into the Atlas. Not a bad idea at all. As mentioned above, pictures and a parts manual are my only resources and I am getting mixed impressions regarding the length or the control housing. It may well be that Atlas offered a retrofit kit back in the day. One problem with working with old machinery is how often it has been modified by previous owners and whether the mods were professional or just something to get it to work for a little longer. Trying to put things back as they should be can be a daunting task when limited information is available.

Thanks again to all, I'll continue to monitor in case anyone has more to add.

12-03-2010, 01:38 PM
Thanks again to all, I'll continue to monitor in case anyone has more to add.

If you come up with a plan, let us know what it is. I'm curious to see how it all turns out.