[sansbury]

I've had a CNC'd X2 for over a year now and have always had to "compromise" with the hideous state of the dovetails on the X axis. They seem to be bowed in at the center, so I can either adjust the gib to be properly tight in the middle, and lose 1-1/2" of travel on either extreme, or I can adjust it to be properly tight at the ends, and have it be visibly loose in the center.

Anyway, I recently got access to a full-size knee mill nearby, and have been thinking about re-doing the dovetails and gib strip on the table and saddle, and I'm curious about the setups others have used to do this. Mainly, I'm concerned about preserving squareness between the X and Y axes. My Y axis is great, so I'd rather leave it alone. Any suggestions appreciated!

[djbird3]

I've had a CNC'd X2 for over a year now and have always had to "compromise" with the hideous state of the dovetails on the X axis. They seem to be bowed in at the center, so I can either adjust the gib to be properly tight in the middle, and lose 1-1/2" of travel on either extreme, or I can adjust it to be properly tight at the ends, and have it be visibly loose in the center.

Anyway, I recently got access to a full-size knee mill nearby, and have been thinking about re-doing the dovetails and gib strip on the table and saddle, and I'm curious about the setups others have used to do this. Mainly, I'm concerned about preserving squareness between the X and Y axes. My Y axis is great, so I'd rather leave it alone. Any suggestions appreciated!

If I were you before you attempt to remachine the dovetails to first try correcting any gib problems and then hand lap the ways for a proper fit. The Column on my SX2 was pitifull, if I adjusted the gibs so that they were very loose at the bottom of the spindles travel and slid the head up the column it would start binding about three inches before it got to the top. My X axis was also like yours, loose in the middle and tight on the ends. I managed overcome these problems by correcting some problems with the gibs and then used a simple lapping procedure to get rid of the tight spots. I've been dragging my feet on doing a post about the lapping procedure but intend on probably writing it tonight. Check out my thread, I already have a post about fixing the gib problems and will follow that up with the lapping procedure and other pertinent info.
~Don~

[TXFred]

I recut both X and Y. Here is how I did it quickly while preserving squareness.

The base and table are easy, because there's really no alignment issue here. Bolt the item to the mill's table, indicate it in, and cut.

The saddle is the tricky one. A mistake here will take your axes out of square.

I mounted a vise to the mill and indicated it in.

I then mounted the saddle in the vise, right side up, and cut the X dovetails.

I then unclamped the saddle and rotated it 180 degrees about the Y axis. I then reclamped in the vise. This method keeps the same face of the saddle against the fixed jaw of the vise. I was then able to recut the Y dovetails and have them be perpendicular to the X dovetails.

Frederic

[sansbury]

I did spend a few hours a year ago polishing the gibs with fine sandpaper and a surface plate, and then using them to work the table in better. I think it helped a little, but there was still a really long way to go. In the center, I can get at least a half turn more on the gib screws, so there could be .020" or more of difference. That's a lot to lap down, and no guarantee the lapping won't introduce alignment errors.

I did think about cutting both X and Y; TXFred's advice about using the same saddle face to align with the vise is spot-on. I was hoping to avoid that since my Y works fine now, but it might be the only smart way to do it. I did consider using an indicator to make a plot of the saddle ways and align the X based on the best fit, but cutting both is probably best and may improve squareness over the original....

[djbird3]

I did spend a few hours a year ago polishing the gibs with fine sandpaper and a surface plate, and then using them to work the table in better. I think it helped a little, but there was still a really long way to go. In the center, I can get at least a half turn more on the gib screws, so there could be .020" or more of difference. That's a lot to lap down, and no guarantee the lapping won't introduce alignment errors.

I did think about cutting both X and Y; TXFred's advice about using the same saddle face to align with the vise is spot-on. I was hoping to avoid that since my Y works fine now, but it might be the only smart way to do it. I did consider using an indicator to make a plot of the saddle ways and align the X based on the best fit, but cutting both is probably best and may improve squareness over the original....

It doesn't even seem possible to me that there could be anywhere near that much error in the dovetail and it be useable at all. I'd take that table and clamp it on the mill and run a indicator along the dovetail and see if it really is that screwed-up. I agree that you can't fix that much error with lapping. I wouldn't attempt more than a couple of thousandths, besides if you aren't old and gray now you would be by the time you lapped that much. I'd say anything over .0025 on a table that size definitley should be remachined but remember not only do you have to recut the table and saddle but also the Y axis dovetail on the base if you want it to properly mate with the saddle. The Chinese don't use a standard 60 degree cutter but rather some goofy non standard angle, at least thats the case with mine.
~Don~

[hoss2006]

They use 55 degree dovetails, you can get cutters from a couple places but they aren't cheap,
$133 US here.
DOVETAIL CUTTERS Dovetail Slot Mills Metric Dovetail Slot Cutters Inch Dovetail Cutter Metric 13mm to 50mm head diameter Inch 5/8” to 1-1/2” head diameter
which is why most just change to much more affordable 60 degree.
You can get some procedures for lapping and scraping here and the arguments for both.
http://www.g0704.com/Shop_Info.html#dovetail
Hoss

[sansbury]

It doesn't even seem possible to me that there could be anywhere near that much error in the dovetail and it be useable at all.

Right now if I grab the ends of the table and twist I can see movement, so I've got at least a few thous of slop in there, probably more. Yet a couple nights ago I milled a 1" hole in 6061 and it came out within +/- .0025". I do have to be careful on feeds and speeds, though--that hole was milled at .030 DOC. Much more and the chatter got nasty. I think what saves me is I have a hefty tooling plate and vise that add at least 10-15 lbs of mass. Not the most elegant solution but I've been making decent parts the past year (chair)

[TXFred]

I do have to be careful on feeds and speeds, though--that hole was milled at .003 DOC.

Oh you poor, poor man. Unless that is a typo, I think you just won the award for the worst X2 mill ever. 0.003" depth of cut is just embarassing.

With my recut dovetails, column brace and belt drive, I can achieve 0.06" DOC at 4700 RPM using a 1/4" carbide mill. I'm not saying this to brag, but to make you aware of just how much improvement you will see from new dovetails and a column brace. In this case, your machine will become 20 times as fast as it is now.

One word of advice. Order more than one dovetail cutter. If your cutter hits a piece of embedded casting sand, it will blunt the tips. After that, it will no longer cut good dovetails. So a spare is a handy thing to have around.

Good luck with the machining!

Frederic

[sansbury]

Oh you poor, poor man. Unless that is a typo, I think you just won the award for the worst X2 mill ever. 0.003" depth of cut is just embarassing.

Yeah, that was a typo, it was meant to be .030" (chair)

Ordering a couple extra cutters is a good call--thanks for reminding me of that.

[TXFred]

One other thing. If you find that your axes are out of square after you finish your upgrades, you can use Mach's formulas to compensate. I cannot remember details, but it's in a thread here somewhere. Essentially, if you're out 0.001" in Y per 1.000" in X, you can tell mach to move the Y axis as needed to compensate.

Frederic

[strohkirchw]

Where is this located in the config?

[TXFred]

I cannot remember details, but it's in a thread here somewhere.

Frederic

[wam]

You can also troll ebay for deals on the dovetail cutter.

1 New HSS Dovetail Cutter 35mm x 55 deg Milling - eBay (item 290538888539 end time Mar-02-11 21:11:31 PST)

[Johann ohnesorg]

Or you can buy a cheapo 60° standard one, remill the dovetails and make new and thicker gibs. But do yourself a favor, buy an american made milling cutter. You´ll lose approximately 0,1" of material from the dovetails. This allows for approximately 5/16" thick gibs.
That doesn´t sound like very much but it adds a lot of stiffness to the machine. Change the set screws, too. The factory ones are undersized by 0,008-0,012" and wiggle in their threads.

A 10mm high dovetail will shrink by 1,23mm in width if you change it from 55° to 60°, both sides reworked will yield 2,46mm or just a hair under 0,1" The gib thickness increase is smaller due to the 60° angle. Take into account that the counterpart will get wider, too.

Cheers,
Johann