Replaceable Insert Throat Plate

Each small picture is a clickable link that will bring up a larger sized picture in a popup screen.
If you keep clicking on images, they’ll appear in the same popup window, so no need to "X" it out until you’re done.

I’ve wanted one of these for a long time:
commercially manufactered replaceable insert throat plate That’s the throat plate (insert) with a replacable insert (what I call an oak insert insert) that Norm has had for some time. The guy that makes (made) them doesn’t appear to be marketing them any longer and has the business up for sale. Even when they were available, I thought they were a little pricey and have longed to craft one of my own for some time. “It can’t be too hard,” I thought, and it wasn’t.
The first need is a source for ½" thick aluminum plate. I’m very fortunate to have an excellent scrap aluminum yard nearby and got some stock. There’s no telling what the specific alloy is, but I would think that in the dimensions we’re dealing with here (≈13½" x 3¾" x ½" thick), there’s probably no such thing as too soft, and I would think that even a quite hard alloy would still be machinable with the usual woodworking tools.
plate stock to insert Using the OEM throat plate as a pattern, I traced the outline of the plate onto an appropriately sized piece of aluminum and rough cut the plate. My scrap yard cuts their plate stock with a K-12 type, gas powered saw with a diamond blade. I started out thinking one of my ready-to-be-resharpened carbide tipped blades in the Unisaw would do the job. It worked after a fashion for rough sizing, but generated a lot of heat and required much more carriage rigging and feed pressure than I thought prudent to employ. A serendipitous trip to the band saw resolved all issues—especially cutting speed. A plain old ½" x 4 TPI blade was perfect.
cutting at the bandsaw Rough cut to the line as you would normally do. Making the curves at the ends can be a little trickier, especially depending on the set of your teeth (on the blade—the set in your mouth will vary depending on the scary quotient of the particular task—this one has very little). If the cut starts to bind, just relax the turning pressure and cut straight to daylight, or at least a little ways away from the line (do I have to point out that this technique isn’t applicable to inside cuts?). Then come back and do more of the rough cut, getting rid of all the material that doesn’t look like a throat plate. Unless you’re a better bandsawyer than I (and that encompasses a very large field), you’ll have some places where you have more waste left than you’ll be happy about having to file.
Those who have been around tools a while have no doubt learned some of the same tricks I have, but just in case, here’s one that’s very useful in wood and more so with the aluminum plate. Just bring the work back to the blade, but register it gently against the trailing edge of the blade well behind the waste yet to be removed. Then rotate the work until the teeth come in contact with it. Continue to rotate and simultaneously feed the work into the waste using the rear edge of the blade as a guide. You’ll be amazed at the relative precision you’ll be able to attain, but rembember that you’re still rough cutting—do not try for finished dimension at this stage. In this operation you’ll want to employ a firm grip but a light touch. You will be consigned to plenty of hand work to make this a nice looking, nice fitting piece.
three sample throat plates-one in progress After rough cutting the plate into the general shape of an insert, I spent some time filing and fitting, filing and fitting, filing and fitting. Eventually, I wound up with a blank plate (center, left). The insert at the top is the OEM insert which came with the saw. It still has its uses. At the bottom is one of the Baltic birch inserts I’ve made, this one sized for a ¼" dado set. A little layout time (to make sure they were located properly), and a trip to the drill press yielded the pilot holes for the adjusting set screws you can barely see here.
my filing station I’m sorry you’re going to be disappointed that I don’t have many live shots of the hand machining (is that an oxymoron?). Suffice to say there were several subsequent sessions at the workbench (filing station)—a lot of back and forth between it and the table saw, and then the work at the drill press and router table described below. What? Your workbench doesn’t look like that? No horizontal surface is safe from me.
my file collection I have a great collection of files and rasps (inherited largely from an uncle who was a foundryman at Ellison Bronze in Falconer, NY, for those who may know something about that sort of stuff), which facilitated several aspects of the project. Real machinists will no doubt be horrified at the jumble shown. They’ll say I’m not worthy of the name. They’re right—I’m not.
With a nicely fitted blank in hand, the finger hole was rough fashioned on the drill press (up to ½") and then with my jig saw (they used to be saber saws—use the term with which you’re comfortable). Then back to the filing station as I worked it to rough size with rasps.
truing up the finger hole at the drill press Once I was very near the ¾" size I desired, I chucked a ¾" paddle bit in the drill press and s l o w l y (both speed and feed) trued up the hole. A light touch with a fine half round rasp and then I finished it up with a roundover bit in my trim router.
Note the far end of the insert resting against the column. There are two kinds of drill press operators, those who have had unsecured work grabbed by the tool and those who are going to. Secured doesn’t necessarily mean clamped. This technique absolutely prevents the work from twisting away in the event of the tool catching it.

When my blank was sized to fit the throat and had a finger hole in it, I installed an 8" blade (a blade I’ve had for years which is actully designed for cutting non-ferrous metal) to make a plough cut which established the initial groove in the new plate. Then it was time for some milling. For that, I turned to the router table.
my router table—'Router Workshop' style My router table setup (right) is your basic Router Workshop style table (flat, no slots, wood or UHMW fences secured with clamps) with an Hitachi M12V router installed (pay no attention to the Bosch 1615 hanging there—this is an old picture). Normally, my fence is to the right of the bit, and I feed by pushing front-to-back…
'New Yankee Workshop' style router table …which is comparable to the right-to-left feed in a New Yankee Workshop style table (right). I used three bits: ½" solid carbide, 516" solid carbide, and 14° dovetail (all with ½" shank).
a push stick is essential Without overstating the obvious, a push stick should be considered an essential element of the task of machining this material. The prime thought when planning and executing this sort of task especially, but equally applicable in all woodworking, even down to the Neandertahl level (that’s not what it sounds like), is, as my late friend, Steve Lamantia (famous author of Scary Sharp TM) says, “where will it go if it slips?” Ask yourself that question at every juncture in your stock preparation and you will improve your safety regimen multi-fold.
multiple jigging to keep the work and the tool together And don’t by shy about setting up supplemental fencing and stops. The goal is to keep the work in the tooling environment. Besides being wasteful of energy (having to keep picking the work up off the floor), it’s inefficient of time to have to file the dings out of your work after it’s concluded another demonstration of Newtonian physics.
watch your hands Naturally, the full bit is exposed in these cuts as the plate is fed with the face side down. As in any blind cut (dadoes on the table saw, is one example—jointing is another) be vigilant of where the bit is in relation to your hands. In fact, I learned somewhere (in regard to jointer operation, but equally applicable here) to never allow your hands to pass directly over your cutter when handling the work—drives me nuts when I see pros do it. This is a staged shot with the router off. No palms were harmed in the making of this picture.
usual operation against a fence For usual table routing—against the fence (bit partially buried in fence)—you would simply run the work right-to-left (NYW style) or front to back (RW style)—as with the setup shown here. However, for this project these are inside cuts. Be sure and keep that in mind as you’re doing the various setups and making the various cuts, otherwise you’ll wind up having to pick the plate up off the floor away from the table. You might have to do it two or three times, if you didn’t think it through enough, like me.
Hogging out the bulk of the material with the ½" solid carbide bit was relatively straightforward. How you register the first cut doesn’t matter so much (since both leading and trailing edges are cutting), although the left edge-left side fence might be preferred as the leading edge of the cutter will tend to force the work toward the fence. And you should make it in the approximate center (although not critical) of the intended channel…so the critical edges you’ll develop later aren’t at immediate risk.
watch the rotation of the bit You must pay attention to the rotation of the bit. Set the fence and the work so you’re feeding against the rotation of the bit. This mostly meant my fence was to my left with the work positioned left against it, and the material I was removing was on the left (inside) edge of the plate. This would be a left-to-right feed on a NYW style table—on mine, it’s still front-to-back, but on the opposite side.
feed required in the opposite direction However, when the right edge (face side still down) needs work, the plate needs to be fed in the normal direction—still with the same reference edge against the fence and to the left of the bit, but now fed from back to front. It’s quite easy with my table to work from either side (front/back, I mean). Therefore, it was no problem to change feed sides when I needed to. It’s even easier on a NYW style table.
I wish I had paid more attention to this, but it didn’t occur to me until after the accumulated experience(s) of the preceding paragraphs—establish one reference edge, and then always machine with that edge against the fence. It’s very important that the two dovetailed edges be absolutely parallel. This means really paying attention to your setup. The funny thing is, I’ve always been conscious of reference edges—I just got caught up in the project this time.

It should come as no shock that you shouldn’t try to hog out the full depth of the slot in one pass. I went about 18" at a time, which worked well. My final depth is about 516" , but there’s nothing magical about that—I wasn’t trying to duplicate the exact dimensions of the original (which I couldn’t anyway, without one at hand). Since I’m going to be the one machining replacement inserts, it can be any depth I wanted—516" seemed like a nice dimension.
'engine turned' finish But I wouldn’t go to that depth with the ½" bit. First, it leaves too rough a surface—something like an engine turning, but rougher. I found the dovetail bit left a very nice surface—still visible, but smoother. And I think I wound up maybe taking .010" or so. The principal adjustment in cutting the dovetail is by moving the fence, anyway, so don’t raise that bit too much when you go to it.
The turning is a little rougher than it looks in this picture, but not really unacceptable. After all, it’s buried under the insert and will only be seen by whoever replaces the insert from time to time.

Also, I did lower the speed of the router for all these milling steps. I don’t know if I had a particular reason to past intuitive, but it seemed prudent. I can’t give a quantitative value but it was down near the panel raising speed. Subsequent discussion indicated I’d made a serendipitous choice—in fact I probably could have used a higher speed. Also, there was no melting. Some of the chips were tenacious, however, particularly in the dovetail recess, and needed to be scraped out with a pointy tool.
oops There is one critical dimension I found by accident. You musn’t go too far toward the short side of the insert—the point of the dovetail gets perilously close to the setscrews and it isn’t immediately obvious as you look down at the upper edge of the groove.
If you see a slight spark, you’ll know you should’t take any more off that side. You might also have to touch up or replace your bit. Better yet would have been a few more nights tossing and turning as I planned this and I could have avoided milling the setscrews altogether.

I hogged out the major slots using my 516" carbide bit. It didn’t plunge nicely like the ½" and at first I couldn’t figure out why. Upon examination, it turned out to be a down spiral bit, which doesn’t have any aggression at all in the bottom cutting edge. I wound up having to cut a couple of clearance holes in the slot and then hogging the whole thickness of material at once. Slow feeding and an additional fence on the other side of the plate to prevent any kicking out to the side made it work okay. I also clamped a stop onto the table (one nice feature of the overhangs on the RW style table) to limit travel.
finished throat plate The insert here is a piece of walnut. It just happened to be an already thicknessed piece at hand and I was in a hurry, once I had the dovetail groove ready, to get something in there. I suspect eventually I’ll use a piece of maple, birch, or ash (Hint: some ten years later, I still haven’t replaced it). To follow Norm, of course, I could make one of oak, but it doesn’t really strike me as the appropriate zero clearance insert wood. Visible here are the enlarged and tapped holes for the adjusting set screws.
Incidentally, I agonized for some time over the angle to use for the dovetail. My first thought, naturally, was the traditional 14° used in the dovetailing jigs. Then I thought 45° would be better. Finally, not having a 45° bit, I opted for 14° and in retrospect, there shouldn’t have been any question. It’s just right.
throat plate with insert removed Here is a view with the insert removed so you can see the machined substructure. With the exception of the plough cut done early on with an 8" blade in the table saw to establish the initial groove, all of the machining was done on the router table.
The through openings in the plate are essentially the same as the OEM insert, so I may be able to fit inserts for miter cuts as well, although probably not for a full 45° angle (thinking about it, the OEM plate is only ≈332" thick—this plate is ½"—that effectively narrows the slot significantly). Still, since I had already made a BB insert for 45°, I’m covered there. The through openings will also accommodate the OEM splitter/guard and an aftermarket splitter, so long as I cut the insert at the back for clearance.

I had some concern during the planning stages about the prospect of aluminum chips getting into the router, even to the point of thinking of placing a snug fitting circle of plastic (milk bottle gauge) on the shaft of the router bits. In the end, when I got to the routing stage I was so excited to be this close to my dream, I just threw caution to the winds and let ’er rip. There were certainly no fireworks, and during two intermediate bit changes, I expected to see at least some debris falling out of the bottom, but there was little or none, even with banging the router about a bit. I suspect there was some centrifugal chip ejection involved.

There is one humorous side story: Once I got the initial sizing done and had a nice fit—probably just a few thousandths of slop…er, clearance—I was gloating about it to a friend and mentioned that I wanted to wait until another 90° day (we still have some coming) before deciding the fit was good enough, due to the differing coefficients of thermal expansion between aluminum and steel.

So I proceeded to the next step, which was to cut a groove in the plate by running the blade up through it (you’ll need a smaller diameter blade than 10" to do that). After just a few fractions of an inch into the cut, I went to adjust the plate and recoiled from the heat. I figured I’d better get the plate out of the hole but it was jammed tight. Coefficient of thermal expansion, indeed. It would not budge.

After cooling down for a while it popped right out. Frankly, I don’t think 90+° will be a problem—it was sure hotter than that from the blade friction. That happened again when I hogged out the blade slot to a full kerf dimension prior to using the router bit. It made me laugh again.
I wish you could hear the 'clunk' I’m so pleased with the result. I have to say it looks good and it has received some nice comments. The plate sits in its place, rock solid, and with a satisfying heavy metallic “clunk” as it goes in. It’s too soon to rate its performance—all I can report at this point is look and feel (and sound). It rates 100% in those categories.

Important Caveat

This isn’t a beginner’s project, I don’t think. I suspect some experience with tools and a grasp of mechanics will prove necessary to replicate it. I’ve been doing this kind of tinkering all of my life, so I have a pretty good idea of potential hazards (even if I didn’t always deploy against them in a timely fashion). Be forewarned that, as with any project featuring powered machinery, there are tasks which are inherently risky. Don’t attempt this just becuase I did.


Bruce, a MN galoot: That’s really helpful, and congrats on a great result! I’d appreciate a little exposition on routing the aluminum. You indicated you did it all at the router table, and while I’ve heard it’s possible (obviously), I’ve never done it myself. How did it go for you? Any hints you’d care to pass along?

Rod Peterson: Thanks, Bruce. I’ve incorporated thoughts on your remarks into the article.

Garrett Lambert: I’ll have to make one of those for myself. Did you do anything unusual on the router table, e.g. slower speeds? Did any of the alumnum melt onto the router bit?

Rod Peterson: Thanks Garrett. I’ve incorporated thoughts on your remarks into the article.

Mike Fitterling: Wow, Rod! Great Work! I have always liked that style of plate but the price has kept me from getting one. Yours looks every bit as nice as the commercial one. I might just have to try that.

Thanks for posting.

Rod Peterson: Thanks Mike. Let us know how it comes out. I’m sure you’ll be pleased.

john b. southwest Chicago burbs: The cutting speed for aluminum is about 3000 sfpm.using carbide cutters. That means that you can spin a ½" carbide cutter at 24,000 RPM. john b.

Rod Peterson: Thanks, John. It’s good to know I accidently came upon the right milling parameters.

Carol Reed: When I was using a CNC to rout aluminum, I had a discussion with the Onsrud people about cutters, speeds, lubes, etc.

Upon their advise, I used 17,000-19,000 RPM on the router, moving about 2-3 inches per second, and lubed with aluminum cutting fluid. I simply painted it on the cutting path with an acid brush before cutting and then every other pass or so. The cutter was an up-spiral solid carbide two flute, right hand twist end mill. Before that, I had galling and busted cutters.

Worked well. Kept things cool and the cutter lasted.

Rod Peterson: Thanks Carol. I actually have cutting fluid for aluminum (A9), but it sure would have been messy in a router table! Fortunately, the alloy is reasonably forgiving, and my speed was at least lower than the ones you used. I couldn’t measure my feed rate, but it was probably much slower than yours, as well. I also incorporated lots of fencing and supplementary feeding apparati (push sticks).

Bruce, a MN galoot: Carol, did you rout the aluminum with the router in a table?

Carol Reed: Well…probably not in the sense I think you are asking.

The info above was what I derived when routing with the CNC. Before that, I made special hold down jigs for the material and a specific guide system for the router, depending on what part I was routing.

To rout in the conentional router table is something I would think twice about. But that is me. When wood gets away from me it smarts. When aluminum gets away from me, it really hurts and always results in leaking blood all over everything. :)

Since at the time I was in production, I was motivated to get the CNC. I was routing a 1½" hole and a 5½" curved slot in the aluminum. I was also routing a long (10-12"?) ¼" wide slot. The CNC reduced set-up and dramatically increased my success rate.

My sense was that the travel exertion of the router was significantly harder that when routing wood. It stood to reason (for me) that a very positive hold down and guide system was in order. So that is what I did. Also aluminum was, and is, expensive, so waste was not an option if it could be avoided.

Eddie Pacheco, San Jose, Ca.: Brilliant piece of shop work, Rod! Doing that amount of metal work with woodworking tools, and not getting hurt, is a feat in itself.

Rod Peterson: Thanks Eddie. Maybe when we don’t know we can’t/aren’t supposed to do it, we’re less intimidated and then less inclined to get hurt. Maybe?

Dan Donaldson: Did you have any issues with the aluminum chips getting into the router? It would seem that with the router upside down, the chips could fall into the router and let the smoke out.

Rod Peterson: I hate when you lose the smoke. I’ve incorporated thoughts on your remarks into the article.

Last updated: 31 December 2017

Valid HTML 4.01 Transitional Valid CSS!