[Ranscon]

I have Thompson 5/8 linear bearings and 5/8 linear rails for my z axis. The bearings have a slight amount of slop in them. (all are brand new) I Just finished attaching the z axis mounting plate and the router. The slop is now magnified greatly. I can easily wiggle the router up and down. This is terrible, I can move the end of the bit 5/16 to 3/8 of an inch with 1 finger. This is definitely not acceptable. my old machine used abec-7 skate bearings and had far less sloppiness. How do you use linear bearings with success. I'm ready to scrap the z axis and build another with skate bearings. The rail and linear bearing design is garbage. Sorry but I am disgusted with the quality of this set up. If any body has any advice on how to tighten this slop up It will be greatly appreciated. (I tried adjusting the bearings with set screws and when slightly snugged the bearings will not even move on the rails). Thanks for any help on this, Jeff

[RICHARD ZASTROW]

Are these ball bushing or sleeve type bearings? What are you using for Z-axis motion, ball screw, ACME or ????

Dick Z

[Ranscon]

They are ball bushings I believe. Look like a sleeve with several rows og little balls in little tracks. Lead screw is a 1/2-10 Acme.

[RICHARD ZASTROW]

5/16 - 3/8" right/left? Up/down? Either way, there is something very loose.

Are all fasteners tightened properly?

Dick Z

[Ranscon]

Left/right is good. Up/down is sloppy. I think I have found a good bit of it in my mount. I will redo my mount and let you know where it stands. Thanks, Jeff

[Ranscon]

I just tightened the set screws in my bearing mounts and it is much better. Now it is between a 16th and 1/8. The majority of this seems to be coming from my 3/4 rod rails on my Y axis flexing a little bit. Any recommendations? Thanks again, Jeff

[harryn]

Hi, pictures always help of course. I am curious - are your Y axis rails supported or unsupported 3/4 in rail ? How long is the rail ?

If it is unsupported 3/4 in dia rail, and even a few feet long, it will flex. Amazing and annoying.

[Ranscon]

I don't have any pics yet. Rails are 41" long unsupported. They do flex a lot easier than I had expected. Is there a way to harden them (with heat and cold water maybe) without affecting their straightness?

[harryn]

I don't have any pics yet. Rails are 41" long unsupported. They do flex a lot easier than I had expected. Is there a way to harden them (with heat and cold water maybe) without affecting their straightness?

Heat treatments can substantially affect hardness and somewhat affect stiffness of a material. I don't know of any heat treatment that would change the stiffness of a steel rod enough to significantly help reduce the observed flex.

You will not want to hear this, but for a given material, it is mostly a depth effect. You can observe this easily at your local hardware store. Go to the rack of pipes and slightly shake a 20 ft long pipe. Keep going up in size until you find one that looks "stiff" when you shake it. (doesn't just wave around visibly). In my case, this was somewhere around 2 in dia.

More or less, a supported rail helps because it it uses its attachment to the base to imitate being larger in diameter.

You can get a modest improvement in stiffness if you can put the rods under tension. There are some DIY plans that call for putting a nut on each end of the rods and tightenting them to put the rod under tension. I am not sure that this will really help enough or not.

[vger]

With the 3/4" rods that long and unsupported, you will also find there is a fair amount of sag when the Z is out in the middle. How much distance is there between the Y Bars? The greater the distance between them the better off you will be. It's all about leverage, rigidity of the lever, and stabiliby of the fulcrums.

One possible fix would involve some re-design and hacking. Get a square or rectangular piece of tubing, aluminum or steel, say 4" X 6" or so, and mount the 3/4" rods lengthwise along the corners. Slice your ball bushings in half lengthwise and mount them in a box surrounding the tube and rods. I know its a hack... but you would only have to buy 2 more 3/4" rods.

Steve

[Ranscon]

Thank you everyone for the replies. I have tried to stiffen this up and cannot get it it within acceptable tolerances. I have just trashed the entire linear bearing and rod set up and replaced it with 1/4 steel plate and roller skate bearing carriages. MUCH, MUCH better. I will never mess with linear bearings again. They may be ok for short spans but not for mine. Anyone interested in purchasing some very slightly used(about 10 minutes) linear ball bushings? I have 2 thompson 5/8 ID and 4 3/4 ID one's I'd really like to sell. Thanks again, Jeff

[handlewanker]

Hi Jeff, sorry you had design problems with the linear ball bushings.

Steel bars will deflect and even if they are supported at both ends they will flex in the middle, but that can be overcome by proper design, and without figures to make a design, guesswork is like walking along a tightrope blindfolded...LOL.

The steel bars would need to be hardened or case hardened in any case, (.020" depth of hardening) and also ground exactly to size if you are using ball bushings, otherwise the balls in the bushings will make grooves in the soft steel bars, and there is no latitude for sloppiness in the fit.

To work correctly the ball bushings MUST be mounted precisely and accurately in rigid housings rigidly attached to the carriage, with the steel bars also rigidly and firmly fixed, and there can be NO allowance for flexing in the housings or the steel bars.

Even if you are contemplating using ballraces (skate bearings) directly on the steel bar surface, the bars must be hardened otherwise the ballraces will dent the bar surface and make a "bumpy ride".

You can offset the wear rate by machining flats, same width as the bearing outer race, along the length of the steel bar where the ballrace/skate bearing contacts it, but you are still liable to denting due to point contact, but not so much.

That's a long Z axis at 40" + long, guaranteed to flex in the middle.

The z axis is usually a short length slide with a main frame slide bringing the cutter/spindle assembly down to within reach of the job.

If you are going all the way with a long Z axis in one slide only, you might be better off with a linear slide way design such as the Hiwin type that is attached to a heavier backing plate.

A Bridgeport vertical mill has a short Z axis slide in the form of a quill in a fixed head assembly, and the table is raised and lowered on dovetail slides attached to the machine body to get the job within striking distance of the cutter.

Table vertical mills have the vertical head Z axis moving in square or dovetail slides attached to the body and the head has the quill inside it.
Ian.