That Big Dog RAS Needed Some Work

[jteneyck]

You might remember my posts about the Dewalt GA 24" crosscut RAS I recently bought from a cabinet shop, where it was used for rough cutting stock.  After getting it tuned as best as I could, the ways were worn enough that I could not get cuts of the quality I want for cabinet quality parts.  The cuts would wander and wouldn't stay square along the length of the cut.  Remachining the ways is not a simple nor cost effective process, so I thought of other ways on how I might bring them back to a high-quality condition.  I thought of filing them down but decided that it would be difficult to file off probably 0.020" and keep them flat and coplaner.  Then I thought about filling in the low/rough spots with epoxy and decided to try that approach.  Worst case, it wouldn't hurt anything.  

This is what a section of the way looked like prior to repair.  



You can see the roughness and the low spots.  I used JB Metal Reinforced Epoxy to fill in the defects, then filled it back flat to the metal.  It looked like this afterwards.



After I made the repair, the carriage rolled beautifully smoothly in the arm, for a while, but then I started to feel clunking and when I inspected, I found that the bearings had started to pull up the epoxy.  I patched the epoxy and let it cure for a week, but the problem remained.  That led me to think about adding a thin piece of steel on top of the epoxy for the bearings to ride on.  I found some 0.033" thick x 3/4" wide stainless steel strapping material (the stuff you band pallets with) and thought that might work.  I roughed up the side to be bonded to the ways and glued on the two pieces with more JB Weld.  To make sure the banding would be flat on the ways and coplaner, I made a clamping caul to fit into the saw arm.



With the strapping in place it looked like this.



Clamping force was applied with turn screws against the cross pieces that rest against the bottom of the arm. 


 
After curing, I removed the caul and cleaned up any stray epoxy.  The ways no look like this.



The strapping is flat and smooth, with no bumps or low spots.  

While I was at it, I decided to change the bearings in the motor carriage.  Removing the stub axles from the bearings required the use of a frame press, which I don't own, so I made one.  



It worked great to remove the old bearings and press on the new ones.  



All good, except the saw carriage wouldn't fit into the arm.  I had guessed that adding the strapping might take up all the slack adjustment for the bearings and that turned out to be true.  OK, now what?  I decided to see if I could find smaller top bearings and I was able to.  The OE bearings are 40 mm diameter.  The new ones are 35 mm.  The new ones are narrower, too, but that doesn't matter for the upper bearings, but it did require that I make some spacer washers to take up the slack on the axles.  It's nice to have a CNC for things like this.  I used 1/4" acrylic and had them made in a few minutes.  UHMW plastic would have been a better choice, but I didn't have any.  



With the new top bearings installed it looked like this.



With the smaller top bearings, the carriage fit into the arm and after some adjustments it rolls very smoothly now.  Moreover, cuts are square and with much less wavering now along the length of the cut.  Still not perfect, but some of it might be blade related, and I'm evaluating different blades.  This photos shows the cut quality before and after the fix.  The before specimen is at the bottom.  You make a crosscut, then paint the cut edge black, then sand it a few times on sandpaper glued to my jointer table.  Low spots stay dark, high spots get sanded light.  



The after sample on top is marginally better, more so because it has none of the really light and dark spots at the beginning of the cut on the right side like the lower before sample.  Both cuts were made with the same blade to eliminate that variable.  

If any of you have a saw with a similar problem, this might be a good solution.  


John

[Hank Knight]

Impressive! Nice work, John.

[stav]

Wow, very inventive. I love stuff like this.

[FrankAtl]

Agreed, very impressive! I would have never thought of this solution or been able to pull it off if I had thought of it.

Thanks for sharing!

[jteneyck]

Thanks.  I'm still testing different saw blades to see if I can improve cut quality, and thinking about what else could be involved.  This is about the best I've been able to do.  The blade enters the cut at the bottom left corner, so if you looked at the photo upside down it would be correct.  




That's a big improvement over this cut before I fixed the ways.  



but it's still not as good as what I get from my little Dewalt 925, and that has me wondering why.  The cut is sometimes high at the very start of the cut, and sometimes low, as you can see from the light spot in the first photo right at the bottom left corner, and the dark area in the second photo.  I can feel tooth marks sometimes, too.  Cut speed seems to have no effect other than the tooth marks are closer or further apart.  The problem is worse or at least more obvious on thicker material.  

The little saw uses a 8.5" thin kerf negative tooth rake blade.  I have tested 3, 12" blades on the GA.  One 60 tooth thin kerf (probably the best cut quality) and one 60 tooth full kerf with negative rake, and one full kerf combination blade with positive rake.  The combination blade was the worst.  I also tested a 100 tooth monster kerf 14" blade.  Cut quality was in the middle.  On Melamine, however, that thing is amazing, with almost zero chipout top or bottom.  So I'll keep it for cutting Melamine and high quality plywood.  

Blade runout is generally 0.002 - 0.003" for the 12" blades and 0.005" for the 14" one.  I'd call that acceptable.  Measured kerfs are about the same as tooth width.  I've eliminated any healing so there's no burning on either side of the cut.  The only thing I haven't yet checked is axial arbor runout, but I will.  Does anyone have any other ideas of what might be causing these less than perfect cuts?

Thanks,

John