Loose Z axis Problem

[erik_f]

Just wondering if anyone has ever had the cam bearings on your z axis loosen up...I was changing a bit today and notice a HUGE amout of play in my z-axis...I talked to Frank and shopbot and he said the z rails might be worn...but I find that either hard to believe or just piss poor. I would have to guess I have maybe, 100-200hours on the machine at best. I took the time to tighten the cam bearings...but how can you be sure the z is properly aligned?

[beacon14]

I found a surprising amount of up-down play in the z-axis recently while checking for side-to side play. Turns out the pinion was not tight against the rack. There is no spring-loading on the z-axis as there is on the x and y axes. Once I tightened the pinion to the rack my cut quality improved noticeably.

Also, there is a tendency for the z-axis to wear unevenly since a large proportion of cutting is typically done at the same area of the rails. In extreme cases it can leave almost a "step" in the rails where the worn area meets the no-so-worn area.

[paco]

Right David; one need to keep his Bot exercise every then or so to get the Z rails wear evenly...

About the spring loaded Z pinion; I have got this issue fixed!


8262

I know, it's look unfinished prototype but... it's engaged!

[kerrazy]

Eric,
Your question regarding aligning the Z, is some what easy.
Surface your table, and look for and step cuts in the surface. If you see some step cuts then examine them and try and determine the direction to adjust the z axis. The four bolts that you use to mount you z axis to the Y car is where you make your adjustments from side to side and front to back. I use a machinist square to aid me in this task. I lower the z with a NEW large diameter bit like a 1.5 or 1.25 cutter in its collet, then using feeler guages I check the space between the machinist square and each of the cutter tips. By using a larger cutter the misalignment is magnified greatly, so if you can get this trued up again on a large cutter, any further misalingment with a smaller cutter will be insignificant.
Paco, Your Z motor should have been bolted to your Z Car, so it should never need tensioning like you have done. Davids tightening of the pinion set screws is a very important step, that can sometimes get overlooked, and I now set these with lock tight and use a set of vice grips attached to the appropriate hex key to tighten these for all I am worth to help reduce the chance of them coming free.
Good luck,
Dale

[erik_f]

Its the actual v-bearings that were loose...so loose in fact that I could move the t-rail that moves up and down about an 1/8th of an inch...this must have just happend recently since I have had no problems cutting...I'm just wondering how often you guys have had to adjust the bearings on the t-rail...or replace this part...OR...has anyone ever had the bearings just vibrate themselves out of calabration? Thanks Dale...that is a similar method to what I use now...but not exactly the problem I was refering to...the 4 v bearings on the z axis aren't/weren't making good contact with the t-rail that hold the actual router...I eye balled it as good as I could...but I'm assuming that if those bearings aren't adjusted properly then you could get a diagnal motion out of the Z's T rail...which I guess you might be able to compensate for with adjusting the whole z-axis...but I'm using the Alpha hardware...which at least for me doesn't leave very much room at all for side to side adjustment. I guess I'm just wondering how I can tell if things just vibrated loose or if the rail is actually worn...The rail looks to be in pretty good shape...but my the past has taught me, just because it looks ok doen't me it is ok. I'm honestly a little annoyed that there is any problem with the Z. My machine isn't used heavily...so if wear is the issue I'm not going to be happy...I was told the t-rail is $250...but I read somewhere else someone paided $350....I little steep I think. I just don't know if there is anything special I need to look for while adjusting these bearings...everything seems to be nice and snug now...no side to side movement...a tiny bit of up and down...but I think I'm going to adjust how the pinion is set...I guess I just don't like it when things go wrong for no reason.

Erik

[paco]

Hey Dale!

I'm not sure I understand this part of your post "...Your Z motor should have been bolted to your Z Car,..." but it is bolted. It's the only way I have founded to have it stay there...

I've spent hours trying to get all the length of my Z rack to pinion free of any (noticeable) play and just couldn't. I had this upper area that would just get off slightly to the pinion and all the rest was contacting correctly... so I needed a solution better than excessive force to make a good contact all the length of it. I now cannot found any play, it sound smoother, I can disengage the Z motor as easy as the other which help rack maintenance and I know it's not applying too much force on the motor. If one can't notice any play in it's Z and his happy with the default setup, I'd say don't change anything but I personally needed to investigated my issue by testing a such installation as I have made.

As for the set screw, I agree; I usually tighten enough to bend those little key...!

Now I don't say that SB is wrong with the fix z motor setup but I firmly believe it's better to have it spring loaded... just as all the other one. I personally think that they haven't founded enough room to fit the spring loading device where I did! Imagine if the X and Y motor we're fixed and not spring loaded...

[beacon14]

It wasn't the pinion set screw that was lose - in fact nothing was "loose" - the motor was firmly bolted to the z carriage, but in such a way that the pinion was not fully engaged in the rack. I have to assume it was snug from the factory and worked it's way un-snug, but I had never checked it before as the Alpha Z carriage comes fully assembled and as I bought an assembled machine from the diplay at the IWF it never occured to me to check it. If it happens again Paco's solution will be the fix.

This is different from Erik's problem but something to look at if you find slop in your z-axis.

[harold_weber]

Paco, is it possible that your rack gear is not mounted parallel to the Z-rails? That would give the problem you describe. For those of us that might want to try your solution, is there any chance of you sending a CAD drawing of the aluminum plate you show in your posted photo from January 27th? How about another photo showing where the plate pivots and maybe the dimensions to the pivot hole? Did you cut the aluminum plate on your ShopBot? Have you considered selling a kit of the parts with instructions??

[paco]

Hi Harold!

I've checked the pinion to rack alignment and it's correct. The problem I had was is related to the fact that the rack is glued on but not all the length; that's where the rack is getting away from the motor if it's fixed. This as to do with the way I installed (and use) the Z also; I use mainly the upper part of it...

I'll put that on my to do list for my personal web site I'm currently building along with other stuff too... Since I only have 5 Mb available for my web hosting both business and personal use, I have to make cruel choices... anybody know about some free web hosting?! I've considered "BLOGS" but I need something that allow me to backup...

As for the pivot, it's the lower bolt (on the Alpha) with two large flat washers; tighten just so it can pivot but there's about no torsion of the assembly. I should get my digital camera soon so I'll include details about it...

Yes, I cut the part on the Bot; I really enjoy to be able to make such part for GIGs and Bot modifications.

Yes, I did considered some kind of retrofits kits (to sell) but so far I haven't had enough demand to justify... I'll see the feed back I'll get from my personal site; there will be other installations and modifications I've made... I've improved the wire carriers, made an home brewed probe, made AL adjustable brackets for limits switches, a Z wires guide, added grounding, limit switch on the Z, "air diverter" for the router motor, installed compressed air for cooling/misting devices and accessories and still tinkering some other ideas!

More later...

[rhfurniture]

I think that the rack getting unglued could be a problem for many.
1. When I built a new y carriage for my benchtop, the old rack was partly detached "flopping in the wind".
2. When some rack arrived across the pond (I'm uk)from shopbot with tape pre-attached, the tape had almost completely fallen off.

Bless shopbot, ds tape construction may work a bit and be cheep, but as serious engineering construction, no. I drill, tap and bolt everything new, and will keep a serious eye on rack still taped to my bot.

R.

[rhfurniture]

Thinking about this issue, I suspect it's success is dependant on 2 things:
1. The climate, namely humidity and temperature.
2. Being used with pressure soon after installation.
What is the climate in Durham USA?

just my 2p

R.

[Brady Watson]

If for some reason anyone has their tape come off, by all means, put the tape back on it. It acts as a dampener and spacer for the rack. The rack should also be attached with bolts top & bottom, with spacer washers under them to space out for the tape. I tried it without the tape and it really was hard to adjust without the tape, so I put new FHB tape back on it. Wipe down with denatured alky before applying tape.

Another thing to point out is that the T-rail is going to wear, especially if you do a lot of 3D work with a lot of Z moves. The Z rails and rack are considered consumables as well as the pinions on all axes. Yes...I wish that they lasted forever, but they don't. They are cheap to replace and should be a regular maintenence item. There really isn't anything else on the machine besides the router that needs maintenence, so I don't balk too much about it. You can adjust the roller bearings a little to tighten it up, but take care when you do this. You can have the pinion too tight or too loose & knock the Z to table perpendicularity all out of whack.

-Brady

[gerald_d]

When Erik started this thread, I hoped that it would turn out like the first thread he ever started here.....(link (http://www.talkshopbot.com/forum/messages/312/1742.html)).
Ah, the good old days!

[erik_f]

Are you suggesting that gnomes are responsible for my problems?

[robtown]

I think they like to be referred to as "little forest people with pointy hats" ...;)

Funny thing about that thread is that I think it's the first time I posted as well...

[gerald_d]

Erik (Jimmy-the-Thief), find the washers that the gnomes pinched and re-assemble the Z-axis to spec....

We also saw the "wear" on our z-axis "T-Plate" and then fixed it with judicious use of a fine file. Also added a second set of V-rollers at the bottom so that the slide is carried in double the amount of rollers and that the wear is now spread to two points. We keep the bearings tight so that it actually flexes the tower tubes open - this eliminates any play. (Bearings like to be lightly loaded all the time). We tapped threads into the tube towers so that the eccentric adjustment became a lot easier. Hands up those that don't know about the eccentrics.......?

[gerald_d]

Pic (http://www.mechmate.com/Forum/messages/47/255.jpg) of doubled V-rollers at bottom of slide.

[rhfurniture]

Good idea Gerald. Could have em all the way up!

R.

[mikejohn]

I do believe that all the good advice in this forum since Eriks very first post has caused some idleness for the wire elves.
In my opinion it is these idle hands that are causing the problems of loose V-rollers.
Can somebody stay late in the workshop some evening, try to catch them at it?
(Not me, -10ºC at night is considered warm here at the moment!)
...........Mike

[gerald_d]

Ralph, the double roller idea came from SB, whose first pics of the Alpha showed doubled rollers top and bottom of the slide.

[erik_f]

I do have some left over bearings that the little bastards must have missed on there last late night rampage through my shop...I think I'll try it out. I was thinking the best way to re-adjust the Z is to set the eccentric bearings all the way in on one side and then adjust only the opposite side until everything is nice nice. Any opinions on this..?

[gerald_d]

Erik, fiddling with those eccentrics will influence 3 things:
1. Tightness of the bearing on the rail
2. Mesh between the z-pinion gear and rack
3. Squareness of Z-travel to table.

No. 1 is the least of your hassles. For no.2 you can remove the z-motor and reset it right at the end. For no.3 you can read this (http://www.talkshopbot.com/forum/messages/7/5876.html#POST23108) thread.

[mikejohn]

I mistakenly loosened these when I was realigning everything recently

I corrected by first squaring things, clamping it in place, checked the z-pinion, then adjusted and tightened
It seems to be ok, better than before. But I wouldn't recommend it!!
.........Mike

[jamesgilliam]

David, I too ran into the problem of the wear on the t-rail, and mine did have a step in it. We usually run in the same area of the rail and this was causing just enough bind to throw the z-zero off. Not a problem cutting through material, but a real headache cutting small v-carve text. Replaced the rail and upgraged the assist spring for upward movement and this corrected it. As for the old t-rail I am thinking of having it reworked by grinding down a portion of the edge and having it hard metalled and reground to the original size, or maybe .030 oversize, as the replacement I put on was almost too narrow to adjust the eccentrics to. Anyone else have any ideas on reworking a t-rail, or success with reworking one?

[gerald_d]

James, it would be more economical for me to build a new "T-rail" rather than trying to rework an old one (except for a little hand filing/grinding). The standard rails are of a low-carbon steel that cannot be hardened by simple heat treatment. Something (carbon,etc.) has to be added before a hard skin can be created. I would rather start with a high-quality steel flat bar and work from there.

If your t-rail is too narrow for the eccentric adjustment, you can bend the tower tubes towards each other in a vice, or you can drill and tap new screw holes. (the tubes themselves don't bend, it is the lighter material connecting them to each other that gives)

[rhfurniture]

I am going to just bolt on bwc hardened rails (from hepco, uk).
r.

[erik_f]

I was thinking of doing something similar buying some type of harded steel and bolting to the existing Z t-rail...I guess I don't really understand why shopbot hasn't done something about this problem...$350 is way too much for a replacement rail that wears as fast as this one does...I'm going to check out the link you put on your post Gerald...for anyone that has any better ideas...please post them.

Erik

[erik_f]

Actually the way I was figuring on keeping the up and down action of the t-plate square was to actually adjust one side of the z axis's eccentric bearings all the way "in" and then adjust the other side tight...this way I have one side that is going to be used as a "square" to work against...does this sound like a good method...then the rack and pinion can be adjusted after.

[gerald_d]

If you adjust the one side eccentrics all the way "in" you will be lucky if that happens to be square to the table. If SB had drilled those holes square they wouldn't have bothered putting eccentrics on both sides of the slide.

[rhfurniture]

Could I just add (and I am sticking my neck out here - it is very non u in shopbot circles) that I would never run shopbot without oiled felt wipers on the rails.
1. According to hepco(who supply bwc in UK) it reduces wear by over 50% on both hardened and unhardened rails - I wouldn't disbelieve it.
2. It keeps the dust that falls on the rails lubricated so it squidges out the way easily, and doesn't form hard spots.

R.

[gerald_d]

This thread (http://www.mechmate.com/Forum/messages/11/185.html) applies.....

[erik_f]

So without being a mechanical wiz...here is my new method that does not assume the holes are square. Go back through the 4 corners of my table and make sure the table is level in x and y directions (to earth)...then level the sub-assembly of the Z axis (to earth) and then level the T-rail (to earth)...I don't have any experiance with measurement devices other than a square and a tape...and a microphone or two...I know eventually I will get things working right...just from tinkering...but I want to know how I did it...so I can do it again when/if problems happen again with the Z.

[Brady Watson]

Erik,
Others have different ways of doing it, but the method that I like best is to level your table perfectly. Then clamp a good quality (no aluminum Mickey Rooney edition levels) to the ground side of the T-rail. Get it tweaked both ways, gradually tightening the bolts as you go. On Alphas it is helpful to loosen the AL extrusion that sandwiches the Z axis and move it in tightly against the 2" 'towers' that the t-rail rolls against. This will eliminate much of the tendency for the bolts to pull the Z axis to the + or - Y side. Moving the Z axis in the X direction (if laying down top in X direction) is easily adjusted...so concentrate on the Y direction 1st. A rubber mallet is handy for bumping the Z one way or another. Don't hit the t-rail itself, only the frame portion (and don't beat it to death...this is only to tap it a little one way or another when fine tuning)

Read the top bubble of the level. When it is close, tighten the bolts, alternating sides. Keep your eye on the level every 1/4 turn or so. It likes to move on you when the bolts are tightened.

I have used this method to setup about 10 bots so far, and none of them have ridges in the spoilboard when I run the flattening routine for the 1st time. I find this method the most reliable for me; Your mileage may vary...

-Brady

[mikejohn]

Brady
Ridges in the spoil board.
After I recently re-squared everything I surfaced the table.
Close your eyes and run your finger tips over the table, you can feel absolutely no ridges
Open your eyes and you can see where the tool cut.
When you surface, Brady, can you see nothing whatsoever, or can you also see a visual difference?
For all practicle purposes, my table was surfaced spot on.
................Mike

[gerald_d]

Mike, also see here (http://www.talkshopbot.com/forum/show.cgi?tpc=26&post=25354#POST25354) and here (http://www.talkshopbot.com/forum/show.cgi?tpc=26&post=21181#POST21181).

[Brady Watson]

No...not tooling marks...I mean angled ridges like the Z is way out of whack. And no...you cannot catch your fingernail on my spoilboard at the intersection of where the toolmarks overlap. I can't say that I was able to do this the 1st time I did my Bot...but you do it enough times and you get the hang of it!

-B

[jdearborn@mac.com]

New guy needs some help. Ran my first test instrument panel tonight and ended up with flat spots at the 4-6 o'clock position on all cutouts. Checked the T on the Z axis and it has some play that seemed excessive to me. Ideas. . .

[paco]

Hi MossieJim!

You mean you have some areas that are'nt cut through?

Which way is the play; horizontaly (V-bearing guides) or verticaly (motor pinion to rack)... or both (both!)?

[mossie_jim]

Hi Paco, I have enjoyed reading your posts.

I tried to upload a photo, but no luck. The vertical (depth) seems to be okay. The actual circle is fine except it has a flat spot running from 4 o'clock to 6 o'clock. The "out of round" was cosistant in 9 circles.

The play is horizontal. I can get the "T" to wiggle between the V-bearings. I couldn't find a way to adjust them. Supposed to be set by SB. Mine is new, first use. Should be nice and tight. . . but isn't.

Thanks,
Jim
(mossiejim)

[mossie_jim]

Hope this works


8263

[gerald_d]

Jim, have a close look at your z-axis v-rollers, a really close look...

You should notice that the 1/4" bolt is not in the center of the roller shaft. And there is another (bigger) hexagon under the bolt head that looks as if someone used a nut as a washer. Well, that big hexagon is part of an eccentric bush around the 1/4" bolt. Slacken the bolt, turn the hexagon and watch the roller move around....

[paco]

Jim,

even if the Z have a play, have you check your CAD design and the toolpath just to make sure there's no such flat spot in 'em?

Yes, I think it should be set by SB when shipped... but @#!% happen... mine was set, but too tight if you ask me. Keep in mind that it's assumed (I believed) that even if the tool is supposed to be "set", you as a Botter, should be able to rectified such things as a play in somewhere... unless something is broken. If you really don't feel confident tinkle the tool, SB will help you...

Now, let fix this play. You'll need two hex keys (three if the nut in the tube choose to turn while you hope it's not). You can see that there is a smaller bolt inside a bigger hex; the bigger one is an excentric to adjust the bearing position in regards to the rails. When you want to make an adjustment, you "slack" (keep in mind that I speak mainly french and I get bored sometime to look in the DICT) the smaller one while holding in position the bigger one. Just untighten it so you can gently turn the excentric for what it need. You may want to practice on the Y carriage holding roller to get used to how this assembly work and you'll may adjust 'em too while at it (that part is in the book I believe).

Now the critical part about the Z rails bearing guide is to adjust 'em tight but to get the rail parallel to the Z assembly too. I would start by looking if there is one side that is already set parallel then work the other side tight.

Ok now, do all of this make sense?

I would have post pics but I don't have my digital camera yet... still watching at Ebay...

[mossie_jim]

Gerald and Paco,

Thanks for the input. The drawings checked good. All circles are a different plot, but showed the same problem.

Gerald, I'll check the 1/4" tomorrow.

Again, thank you.
Jim

[erik_f]

Shopbot told me to make sure to turn the eccentrics in a counter clockwise motion so I wouldn't loosen the moutning bolt...so maybe you don't have to "slacken" the moutning bolt at all?

Erik

[gerald_d]

That is not valid advice. To tighten and loosen you need to be able to turn in both directions. And, depending on which side the eccenter is lying, counter-clockwise may mean loosen or tighten - you don't know what is going to happen until you start turning.

I totally hated the loose nuts up inside the square tubes needing 3 wrenches/spanners (of sizes not easily available in most of the world). Very early we tapped 6mm threads into the tubes (ours are pretty thick) above and below the 1/4" holes. The y-axis "hold-down" rollers serve as the double z-rollers. Eccenters were made from bolts. The 6mm screws are allen key type, so that the wrench doesn't fall off while fiddling.

[mossie_jim]

Made the adjustments. Working well now. Thanks Gerald and Paco.
Jim