Official Luthiers Forum!

Peter Havriluk's question about measuring the runout on the PC310 in the collet%#@&!!! discussion has me thinking about all the years I've used trim routers and how often I've had a problem with runout. Except in the case of a worn out router with failed bearings or an extremely cheap Chinese HF trimmer which had many problems, I can't think of any.
When I have had problems with out of round or varying width channels it was usually a problem with the bit, or more usually, my own poor set up or technique.
I am curious to know what problems others have noticed with router runout and how you dealt with them. I know some have bought precision collets for their routers - is this the principal cause of runout and has the collet cured your problems?

Runout on a lam trimmer right out of the box is not a good thing and although it may be fine for counter tops .005" run out on a lam trimmer with .050" thick bindings is a 10% error rate..... I would not accept a 10% error rate on my top or back thickness or my rosette or anything for that matter. In the context of what we do lam trimmer run out is a big deal and to be avoided. This is not about technique and I wish Link was here to describe his test methodology. He is who Krenov willed his hand made planes to, Link is so technique is not in question.

We produced the Collins saddle mill and that .005" runout in a saddle slot would have been visually noticeable and even led perhaps to sloppy, leaning saddles.

We were the ones who posted here on the OLF first about Precise Bits and the better stuff that they sell to tighten up the runout of lots of tool products. The Dewalt with a precise bits bit and we average around .001 runout which is good enough for the intended purpose. Can't remember if we replaced the collets too but we may have. It's been a while.

Runout is runout Clay and when a lam trimmer is strapped down and imobilized and then runout is measured it's not about technique, it's about a ***** consumer product lam trimmer that is not suitable for Lutherie work unless you accept sloppy, substandard work.

Speaking of precise bits, Ron has an excellent article there about calculating runout. I've followed this for my CNC spindles. https://www.precisebits.com/tutorials/s ... runout.htm

There is actually a lot of great information on their site. The navigation is terrible. In terms of discoverability and UX design, it's very bad. That being said, click around in there. Lots of gems.

_________________
Insta - https://www.instagram.com/cbcguitars/
Facebook - https://www.facebook.com/cbcguitars

been using BQTools it is key to use good stuff , cheap stuff is a cost good tools are an investment

_________________
John Hall
blues creek guitars
Authorized CF Martin Repair
Co President of ASIA
You Don't know what you don't know until you know it

Runout is not good because it stresses the bit, however, it is fairly consistent during the cut. So it will not cause a lack of precision during the cut if it is allowed for. A channel cut by a 1/4" bit will turn out to be 1/4" in width plus 2x the runout (dependent on the length of the bit). But like I said, this channel will be consistently over width, not uneven.

With a spiral flute bit, runout will cause the sides of the channel to be slanted, concave or convex. The shape will change every time you insert the bit. With a straight flute bit, the cut is scalloped, but the sides are straight. I use a straight flute bit to cut saddle channels for that reason.

Hesh,
I can see where those who fit saddle slots to saddles may require a fair degree of accuracy. For those who fit saddles to saddle slots perhaps a little less precision in slot width can be tolerated, so long as the depth and width are uniform.
Working with HPL materials can require more precision than working with wood. Unlike wood, repairs can be difficult to affect, and sometimes require replacement of the entire work piece.
I agree Barry, when the channel width has been uneven it is usually because of slop in the setup or something has slipped. The slight amount of "over width" runout may produce has not been noticeable to me with the routers and techniques I have used.
But I am curious as to what others have experienced, and Hesh mentioned at least one case where with what they do, runout makes a difference. What other problems have people experienced and how did they remedy them (better routers, precision collets, precision bits, change in methods?)

Runout in your router becomes more noticeable as the bits get smaller, that's no new news to you guys. I don't find it to be a problem on any operation where I'm using a 1/4 or larger bit. In the past I was doing mostly tiled rosettes but have switched to spaced multi line rosettes. Running a .060" bit with a total indicated runout of .005" starts to get a bit too sloppy for my taste. Since switching my DW611 stock collet to a Precise Bits precision collet (which you'll notice are not available for a lot of routers because the routers are not accurate enough) and using their bits I am much happier with the results. The other two operations where I'll use the precision setup is milling my saddle slots and of course in my cnc mill for inlay work. And, no, I don't fit my saddle slot to my saddle, no one does. I do get a more accurate slot which simplifies fitting a saddle.
For my logo inlays I use a .020" bit. Try that with .005" tir, hope you've got plenty of those $5 to $15 bits laying around because you'll be breaking a lot of them. I've attached a close up photo of a test inlay, corian in maple, actual size is about 1 3/8" square. No way to get those results with a router with excessive run out.

That's an incredible fitting inlay, Steven.

Ditto: if you can route like that, who needs cnc?

We do a lot of work for Martin owners since we are a certified warranty center. Martin provides dealers with the ability to purchase the saddles that they use in production.

Normally we only provide high quality bone saddles that are individually crafted and intonated for the specific instrument. But for Martins it's way easier for us at times when a new saddle is needed to use a Martin OEM bone, compensated saddle. It's cheaper for our clients too since we charge more to customers make one from last weeks pot roast.....

Long story short Martin bridges are CNC produced with usually very accurate saddle slots that the OEM saddles nicely and snugly fit into the slot with zero gaps. I usually have to swipe the sides of the new OEM saddle a few times on 320 paper on a surface plate and I always when replacing a saddle draw a pencil line on the bottom and then mill that saddle bottom on a surface plate with paper on it to be sure it's absolutely flat for pick-up response reasons.

Another topic but we often mill the bottom of saddle slots because they are not level and the complaint is uneven pup response. This fits that usually and is a unique things we do with the saddle mill. We need to be able to mill the slot bottom only without touching the slot sides, that's not possible with a sloppy .005" runout lam trimmer.

So these OEM saddles can't be made thicker..... well they could but it would be hack work.... so the precision of the slot is important here to use OEM parts where desirable.

Throw in the fact that in the early 70's many Martins had their saddles improperly located (bad jig) and I speak of 18's, 28's and 35's. To this day they are still showing up and often it's the case that the cowboy chord players never noticed that their D-18 had poor intonation which is hard to notice if you live your life playing a G, C and D...... But then they die, the guitar changes hands, the new steward likes to play the greatest song ever written, Smoke on the Water and then they notice that the intonation is off.

So it comes to us and Martin will cover this under warranty had this guy been the original Smoke on the Water player but since he is a second owner he has to pay us.

We use the saddle mill to true up the existing saddle slot, find a matching blank wood/material and make a plug. We plug the bridge saddle slot shooting for invisible even matching grain when we can. Then we mill a new slot where it was supposed to be all along. It's a pretty cool fix the bridge never has to leave the instrument either since our mill is vacuum clamped on the guitar top and immobilized completely. We've done this on near priceless pre-war Martins too and it's safe, predictable and never has failed to do a great job.

Now with very little runout we can cut the slot to be a near perfect match for Martin's OEM saddles.

We can also do other cool tricks such as tilted saddle slots or for through saddles where a UST (under saddle transducer) never would have worked we can mill a special pocket for the UST under the through saddle preserving the vintage and traditional look and bringing modern electronics to an old dog with a minimally invasive and reversible mod.

With all of this said we also use the mill to route pick-up cavities and such. Runout of .005 would have been a non-starter for us and much of what I described above would not be possible with professional, very high quality results if the dang tool was sloppy as can be.

Going back to bindings. I don't know about others here but for me I was obsessed with uniform thickness binding everywhere on my guitars including on the back by the neck heel a place where traditionally bindings can get very thin. That 10% variance in a .050 thick binding would be a show stopper for me. I'll go further that on a multi colored purf 10% of slop might remove an entire color in certain places.

So if anyone is trying to make the case that .005" of runout on a brand new tool and for how and what we use lam trimmers for with Lutherie in my experience this is an area where the utmost precision is beneficial and always worth shooting for. This is also very much a question of personal standards, YMMV and I don't care (actually I do or I would not be here....)

My ex boss Jack Welch was famous for saying be number one or number two in all that you do or go do something else..... He also made 80,000 share holders millionaires. Neutron Jack RIP.