How flat should router casting be (for table) - Milwaukee 5625

[Tyler Keniston]

I am in the process of mounting a Milwaukee 5625 to an Incra aluminum router plate and starting to worry about the flatness of the casting. The plate (which is flat so far as I can tell against any shop flat surface) rocks slightly on the casting. A straight edge also indicates that the casting is not flat. Kind of hard to say by how much, but lets say in the 5 thou ballpark range if pushing on a low spot and reading across from it at a high spot with feelers.

Obviously what matters in the end is the flatness of the plate, and its security (i.e. does not rock or deflect). My concern is that the plate will be drawn down onto the lowest attachment point.

Should I expect the casting to be dead flat? Since it is not should I:
A) contact Milwaukee if it is indeed something they would give two hoots about
B) Lap the casting myself (my only fear being preserving squareness to the chuck)
C) Shim under the plate
D) Get a router lift instead so as to avoid the casting all together
E) just tighten the screws and get on with it.

Last point/question is in regards to the design of the clamping mechanism of the 5625 (which makes me consider option D more seriously). The casting is clamped in place to hold the motor, and there is no separation that I can tell between the casting that clamps and the base of the casting, thereby inducing distortion when clamping the motor. Is this an issue anyone else would pay any mind, or would consider it a shortcoming?

Thanks,

Tyler

[pat warner]

It should be ~ +/-<.001" all over.
If it isn't, for portable routing, you'll bend the base plate whence applied.
And it will rock.
Who knows whether you'll achieve spindle normality (90°) in the router table.
There are just too many variables.
***************************************
Wet sand the casting? I would. On a granite surface plate.
The high spots will waste before you screw up the normality.
Moreover, you can't sand too much aluminum.
You think you're removing a lot of metal but you're just altering the mill marks.
But indeed, the high spots will sand down because the unit forces will be high on those points.
************************************************** *******
Is this a Milwaukee exclusive? Hardly.
**************************************************
Motor normality at any position in the casting? I doubt it.
Routers are not precision tools. The best of machines for motor normality in its casting: PC 7518.

[Rich Riddle]

One thousandth of an inch for a casting for a tool dealing with wood? Good luck finding that. Manufacturers just don't mill them that exact for some reason.

[John Lankers]

I own several routers and not two of them are of the same color and they all struggle with their accuracy when clamping. I replaced a Freud with the 5625 in my table and no matter what, router setup is by trial and error.
The casting is aluminum and simply screwing on the base plate will twist it somewhere to a degree. Mount the router in the table, make sure the insert plate is flush with the table and go from there.
Btw. The Milwaukee is a pretty darn good router IMHO.

[Prashun Patel]

"Obviously what matters in the end is the flatness of the plate".

In addition, the bit must be perpendicular to the plate. If the casting on the base of your router is out of flat enough to inhibit this, then I would either shim it or ask Milwaukee for a replacement.

If it were me, I would cut some 45 degree chamfers, and some mortises and some rabbets and measure the results for accuracy. If they are good enough for your joinery, then I wouldn't bother.

[Art Mann]

I do not think the plate is a casting. The description says it is a machined plate. Aluminum plate is normally rolled and that is probably the way you want it to be. If you are getting deviations of only 0.005" over the whole plate, that is probably as accurate is you are going to find on woodworking equipment. That is only the thickness of a sheet of notebook paper. If you go back to high school trigonometry class and start calculating the angle error produced from such an inconsistency, you will probably find it is irrelevant in comparison to other factors. Beyond that, the flatness of the table you are using with the insert plate is probably not to +/- 0.005", even if it is machined metal. Table saw tops certainly aren't. In the past, many "experts" have even advocated a router table that is slightly convex where the bit protrudes through the table. All the plates I have ever installed included jack screws to level the insert plate with the rest of the table and this plate is no exception. They are there for a reason. It is not reasonable to assume that woodworking equipment is built to the same accuracy as machine tools. Pat's stuff is but that is an exception. If you want that level of precision, you need to look at his stuff.

[Tyler Keniston]

Thanks all for the replies.


I do not think the plate is a casting. The description says it is a machined plate.
To clarify, it is not the Incra router plate that has the issue, it is the router casting itself. My worry is what will happen to the plate and the bit 'normality' when the plate is screwed to the not-flat casting.

I may see about a replacement, may lap it, but I am also tempted to start looking at lifts the more I think about it. I know its woodworking and not precision machining, but I like to at least strive for highest reasonably achievable accuracy in any given tool so that error variables are reduced and less time is spent fussing on set-ups and tweaks.

[Art Mann]

Sorry I misunderstood you initially. Here are my comments in light of my better understanding.

For one thing, thin cast aluminum is not that stiff. If you screw it to an aluminum plate, the casting will conform to the plate to some extent. For the sake of argument, let's assume that the plate will do nothing to flatten the casting and that aluminum plate warps by the maximum amount (highly unlikely). From the first worst case assumption, the 0.005" is then an accurate error term. Lets further suppose the radius of your router casting is 3 inches. Then the worst case angle caused by the defect can be no worse than ARCCOS(0.005/6) = 89.90 degrees. 0.1 degrees is then the error between perpendicular and the actual router bit shaft. Now lets suppose you are using a picture frame or raised panel router bit that is 1 inch tall. The error position at the very tip of the cutter would be 1 * COS(89.9) = .0017 inches. That is 1/3 the thickness of a sheet of notebook paper. Is that really relevant? Are you sure your fence is that accurate? Can you be sure your feed technique doesn't vary by more than 0.0017 inches? My conclusion is that the lack of flatness in your case is absolutely dwarfed by other variables that you have no control over.

I am not trying to be argumentative. I am just commenting on the fact that most woodworking tool manufacturing errors aren't relevant. Many, many times I have seen beginning woodworkers distract themselves with insignificant machinery problems and neglect the real goal of making things with wood.

[Wayne Lomman]

Art is spot on here. There is far more to be gained by good technique than by stressing over the machine. Also, accuracy is why commercial machinery is big and expensive. Have a go and see what happens. Cheers

[Myk Rian]

Lap the router base if you desire.