So, how am I measuring my router wrong?

[Wade Lippman]

I am building a new router table; all done except an enclosure around the router.

I put a Milwaukee 5625 in and found it had terrible run out.
I posted here about it and all agreed the router must be defective.
http://www.sawmillcreek.org/showthread.php?202141-My-new-Milwaukee-5625-shudders

I returned it and Toolup.com sent me a replacement. It is exactly the same as the old one. I tested three router bits and all three show runout proportionate to their length.

Assuming all Milwaukee routers don't have the same defect, and further assuming Toolup.com didn't repackage my router and send it back to me (they wouldn't do that, would they?) then I must be doing something wrong.

I installed a router bit with a bearing on the end. I clamped a piece of steel to the table and put a magnetic based dial indicator on the steel. Putting it up against the bearing I turned the router bit. The dial went back and forth.
I did that with three bits and got movement proportionate to their length.

To me that indicates that either the collet or the router itself has run out. But it did exactly the same thing with the first and second router/collets.

I don't see what I could be doing wrong. If the table or the lift was out of square, then the bit would just be tilted; it wouldn't show runout.

So, what do you think?

[Martin Jodoin]

Hello Wade

I have a Milwaukee router but never measured if there was any run out.
How many thousands of an inch do you mean by "terrible run out" ?
Is it possible that there is some play in your bearings ?

[glenn bradley]

I have a couple 5625's. I'll try to do a quick measurement when I get home. If my numbers are like your numbers then I would say you are OK as I have never had an issue with either router in practice. If my numbers are better than yours it will make me sad to think that what was one of the best router available 5 or 6 years ago has fallen so far.

[Chris Padilla]

I guess you don't have other routers to measure, Wade? I mean, really, only one router in your whole shop?! ;)

[Richard Coers]

Do you have the collet installed correctly? Some of them are self ejecting. You have to remove the nut and collet, put the collet on the bench, then push down the nut until it snaps on the collet. Have you checked runout with a bare piece of steel bar? Cheap bits, will mean more runout. Your last post listed runout with 100ths. Was that a real number, or was it in 1000ths? I've never heard anyone describe runout with 100ths unless it is really that far out. Have you tried deflecting the bit back and forth to see if the bearing has play in it? I would also take the router out of the table, and clamp the steel to it's base. You may have a little clearance on the mounting plate. More info needed.

[Mike Goetzke]

Wade - Have you tried making the measurement on the router alone w/o the lift. I didn't have your same problem but I found out after having fit problems with some raised panel doors that my WP router lift wasn't clamping my router collet square to the table. Had to loosen it and tap it into position. But in my case the runout was good.

Mike

[Wade Lippman]

First, let me apologize for a bit of premature dementia. My report of 8/100ths is almost a .1". Giving it a little thought, that just doesn't seem right; so I had another look at the indicator. The indicator says "0-1.0 In" and has a scale from 0-100. I made the assumption that one full revolution was and inch, so each mark was 1/100th of an inch. No so. Each mark is 1/1000th of an inch. (maybe the 0-1.0 In means it has 1" capacity?) So, I am actually getting a runout of 0.008" with a 3" long bit. That is with both the first and second router.

I have a 8 year old Bosch 1617. I put the same 3" bit in the Bosch and measured .003" run out.


So, the Milwaukee has nearly 3x the runout of my Bosch; but is it excessive?

[pat warner]

"So, the Milwaukee has nearly 3x the runout of my Bosch; but is it excessive? "
******************************************
Have >30 machines (all brands) for many reasons; I study runout. I would say that I have seen more runout in Milwaukee's. Started with 5650's and 80's; they were miserable. Have had bearing problems with 5615's and 2 5625's. Their components (including collets) are good to excellent. But in my anecdotal sampling, hi. runout is common. Get some ground .500" drill rod, say 2.5" long. Put that in the collet and remeasure. .004-.007" is not uncommon. Starting at 8 is too much squiggle. Appreciate that it is not uncommon for all sorts of router bits to spin in funny circles.

[Myk Rian]

Measuring the bits at the bearing can give erroneous readings. You need to measure the bit, preferably the cutter edges.
Are you sure the bits are straight?

[glenn bradley]

Sorry Wade, I almost forgot to do this. I don't have anything super sophisticated but, the dial reads in .001" increments. I tried to setup a worst case by using the 1/4" collet. I chucked up a pin designed for centering template collars on handheld routers. I found the lowest read point and zero'd the gauge. I placed the gauge tip at the extreme end of the centering pin and turned the collet by hand. The highest reading is about halfway to the first mark which would be .001". From there it returned to zero. I could measure other ways within the scope of the tools I have if it will help.

261344 . 261342 . 261343

[Wade Lippman]

Measuring the bits at the bearing can give erroneous readings. You need to measure the bit, preferably the cutter edges.
Are you sure the bits are straight?

I agree the cutter edges would be better, but I had real trouble getting repeatable figures. With the bearing it is easy.
But the Bosch had less runout with the same bit, so I think much of the problem must be in the Bosch.

Of course I might just have gotten lucky with the Bosch; it might have had runout opposite the router bit. I should have tried it in a few different orientations.
Maybe I will try that tomorrow.

[Jim Neeley]

Wade,

I recommend removing the collet and cleaning the inside really well and blowing it out. A small amount of residual anti-rust would be magnified by the length of the bit.

[Mike Goetzke]

Wade - Do you have a base centering cone? I use a centering cone to check runout. Also, do you just drop the bit in the collet or do you have a rubber o-ring in the bottom of the collet or do you lift the bit a little before tightening? This is so the bit doesn't bottom out.


Mike

[Wade Lippman]

I don't have any drill rod, but I have a 1/4" bit with a really long shaft. I chucked it up and tested the runout at 1" from the collet. 0.001". I tried it in three orientations and got the same result.

So, I am going to conclude that I don't have significant runout. If I had tested that first, I never would have done anything further.

In the previous post I tested runout with three bearing bits and found high runout with all. It is hard to explain how 3 bits can be consistently off, but I am just going to assume that you can't get meaningful measurements with bearing bits.
It would a lot of work to figure out why I have so much less runout with the Bosch than the Milwaukee, but if you can't get meaningful measurement on a bearing bit, then it would just be a silly waste of time.

I appreciate everyone's help with this. (and I apologize to Milwaukee for returning what is likely to be a perfectly good router)

[David C. Roseman]

Late to this thread, but I think Myk nailed it by saying don't measure off the bit's bearing. When held firmly against the template/workpiece during cutting, the outer race of the bearing isn't intended to rotate. So while it would be nice to have it, perfect concentricity of the outer and inner races isn't as critical in the manufacturing process. Hadn't really thought about that until Wade posted about his issue.

David

[Jeff Duncan]

Yup, as David said, you don't want to measure at the bearing as they have their own play in them. Here's the thing to remember when checking tools for runout, you want to eliminate as many sources as possible. So if you try to measure the runout of bearing on a bit in a router that's mounted in a table, you have so many potential sources of error it would never be reliable. If the router base is off to the table by a fraction of a degree your going to have runout, the bearing as mentioned has runout....too many places for it to add up. The only reliable way to check IMHO is to attach the dial indicator to the base of the router, insert a known straight piece of rod, and check the collet to the base.

Anyway it sounds like you have solved your problem for now, and learned a little something along the way;)

JeffD

[Joe Hillmann]

What was the reason for returning the original router? Did it actually not cut well. Or did the runout just bother you?

[Rick Fisher]

I remember reading an article somewhere that said the Milwaukee routers had the most run-out .. It might have been on Pat's website, can't remember..

[Bill Bukovec]

How does the Milwaukee router cut?

[Wade Lippman]

How does the Milwaukee router cut?
Just finished the router table today; haven't wired up the switch yet. Should know in a few days.

[Marvin Wilson]

Have the woodworking magazines driven us all mad? Some of the most beautiful pieces of furniture ever produced have moldings produced by a wooden plane with a hand shaped cutter and we are talking about 3000th's of an inch. You never mentioned whether you simply tried to rout an edge or not. Did you? Was the finish acceptable? Then it is a good router. For crying out loud men we build furniture not jet engines. We are woodworkers not machinist. My suggestion is to lighten up, stop obsessing over the mythical zero run out and go build something useful. It's much more enjoyable.

[David C. Roseman]

Have the woodworking magazines driven us all mad? Some of the most beautiful pieces of furniture ever produced have moldings produced by a wooden plane with a hand shaped cutter and we are talking about 3000th's of an inch. You never mentioned whether you simply tried to rout an edge or not. Did you? Was the finish acceptable? Then it is a good router. For crying out loud men we build furniture not jet engines. We are woodworkers not machinist. My suggestion is to lighten up, stop obsessing over the mythical zero run out and go build something useful. It's much more enjoyable.

Point well taken, Marvin. But if we pay top dollar for a tool, don't we want to be sure it's built to its own manufacturing tolerances? I think that was the OP's concern. He wasn't sure he was measuring correctly, and turns out he wasn't.

David

[Joe Hillmann]

I read an article a few months ago (If I could find it I would post a link)in which they were doing a story on a high end hand to manufacture. The owner of the company said back when he started only machinist owned accurate measuring tools so when someone bought a tool from them if it worked it was a good tool. But now with cheap measuring tools from China anyone can own them and think they can accurately measure down to .0005 and they expect there tools to be as accurate as they think they can measure them. The problem with that is just because the tools may read out to five ten thousandths of an inch doesn't mean the person using it can measure to that accuracy. Because of that the company made a large investment in large slow speed grinders that act almost as random orbital sanders to make the surfaces of there tools flatter than what can be measured without specialty tools. Although the article said even with the large investment the tools they are making don't work any better but it is required to stay in business in the modern world.

[Chris Fournier]

If you're using a router bit to measure your routers runout you might as well stop right there. It really doesn't matter where on the router bit that you measure, this is not a useful exercise. You are measuring both the routers and the bits runout at once. A drill bit is no different really. You need to buy a standard of known accuracy and chuck that then measure the routers runout. Any other efforts are a waste of time and the data you get is not terribly useful unless you're using that router bit, chucked that way in the router - then you know that you are running out 0.008" for example.

[John Piwaron]

I guess you don't have other routers to measure, Wade? I mean, really, only one router in your whole shop?! ;)

That *is* funny - one router. :) Those seem to collect like stray cats to a tray of milk.

[John Piwaron]

Mr. Hillmann,

For sure measuring is a science. I'm frequently impressed with the things my employers QC dept. has to measure things. Some are exceptionally clever, all are exceptionally accurate. All need training to be used to get valid results. Some training takes 10 minutes (micrometer) for some (CMM) you go to school for a while.

[Myk Rian]

Mr. Hillmann,

For sure measuring is a science.
It's more common sense than science.
I did precision tool repair/calibration for 20+ years. If you study the procedures, it makes perfect sense.

[Pete Bradley]

Chris Fournier has it right. Measuring at the bearing of a router bit is a dead giveaway that it's time to put away the dial indicator.

There's way too much emphasis on machine measurements in certain woodworking mags, and some of the "precision" measurements they publish demonstrate a complete lack of understanding of machine precision. Likewise there's endless blather about checking for "play" in new bearings when the worst quality offshore radial bearings now have runout that is unmeasurable by amateur methods.

Most likely the bit that you're using was not very accurately manufactured, but it still may work just fine. If the router works, enjoy it.