Making rabbets with a router

I’m making another rack. I promised I’ll write about the finish of the first, tung oil, and shellac, and I will get to it eventually, but I want to write this before I forget.

As I mention in the rack post, all the joints are rabbeted, and with the first rack I did some with a table saw and a dado blade, and then some were impossible to do on the table saw, so I did them with a router. Also, the dado blade was tearing out some wood, so it really makes sense to do it all with a router. Here are my best rabbets:

Some aren’t as pretty and smooth but they’re pretty close, and with the glue and the piece that goes in the joint, it’ll all be invisible. These were done separately, meaning that the long edge was done first, and then the shorter one. I then squared the rounded corner with a chisel. I thought I took a picture, but I guess I didn’t. It came out pretty well I thought.

So here’s my method for making rabbets, but before I get into it. Yes, it’s safe to use a router on Baltic Birch plywood. I do not know if this will work with regular hardware store plywood, and I don’t know if it’ll work on Oak plywood that has a very thin top – I can see that tearing like crazy, but it works very well on the Baltic Birch, as you can see.

Width of the rabbet:

Pretty self explanatory, it is the horizontal distance in the pictures. I hemmed and hawed about how to set that distance in the best way, and here’s what I came up with. I measured the distance from the edge of the router base to the edge of the bit. It measured exactly 2.5″, and I’m sure it’s specific to this router (Porter Cable 690) and bit. So if I set my guide/fence at 2.5″ from the edge of the piece, then the bit will just slide along it. So now I add the thickness of the piece of birch. This birch is pretty close to being 3/4″ thick, but it’s not – it’s about 0.715″. Now I add the two measurements together (2.5″ + 0.715″ = 3.215″) and set my digital caliper to this length. Then I use my caliper to set the depth on my adjustable square. So basically, I use the caliper like I would measure depth (of course the measurement is locked in) and adjust the square so it’s perfect. Then I make two shallow passes with the router and check if the width is right by placing a piece of birch on it. I like it to be dead on (it almost never is) or for the piece to be a little shallower than the edge of the rabbet. So if the piece is really 0.715″ thick, then the width of the rabbet is maybe 0.715″. The excess can then be trimmed away with a flush bit with a ball bearing.

Depth of the rabbet:

First and foremost, do multiple shallow passes. I probably do 5-10 passes to get the depth. Try and remove a 19″ stretch of 3/8″ thick of Baltic Birch in one pass and your blade will dull and your joint will suffer.

I set the depth like I set the width. Measure 3/8″ on the caliper, and copy that to the square. My first rabbet will be exactly that and I’ll measure how far the bit sticks out by placing the square on top of the bit and see how the ruler end touches the base. When I have rabbets that are a corner, I do one edge first (mentioned that earlier), and make it 3/8″. Then for the depth of the other end of the corner, I just keep lowering the bit down, and I see how it sits on top of the rabbet I have already done. Obviously, power is off and the router is unplugged. I keep comparing and I like erring on the side of caution, meaning that it’s ok if the new edge is a little taller.

Double cutaway Telecaster pickguard

A friend is putting together a double cutaway Telecaster and needed a pickguard. As far as I know no one offers production doublecut Telecasters, which means that pickguards for them aren’t available either, so I offered to make him one. Again, because of their scarcity, that meant I had to make the template from scratch. It was my first time making a pickguard and I hit a few snags that could easily be avoidable.

He sent me a drawing of the guard that I believe he made by copying a regular Tele guard. The drawing is available here. The scale is 1:1 so anyone should be able to print it and use it, which is what I did.

I got a 3/4″ thick¬†piece of plywood and that was my first mistake – it’s too thick. I had to use a lot of sanding drums (handheld and stationary), and the thicker the piece of wood, the smaller is the chance that all of it will touch the sanding drum at once. It also meant that it was impossible to make the straight edges straight. Like the back edge of the pickguard that sits against the bridge:


See what I’m talking about? It’s tough squaring those edges on such a thick piece.

Unfortunately, I didn’t realize this until I was almost done making the template.
So I went ahead and got a 3/16″ piece of MDF/particle board. It was much easier to work the contours and straight edges with this piece. For instance, here’s that same back edge on the MDF piece.


Not perfect, but well within the margin of error. Meaning that it’s easy to smooth those bumps on the pickguard itself.


Here are the two templates side by side. See how thick the plywood is!



The next step was the toughest – cutting the hole for the neck pickup. Centering it isn’t a big deal, but how do you cut the hole at that exact width and with that exact profile (meaning the circular edges). The width of the hole as it’s drawn is 9/16″ exactly, but I wasn’t sure what was the radius of the sides (i.e. the half-circles). I practiced on the thick template by placing two straight pieces of wood 9/16″ apart and cutting with a router while avoiding doing the half-circles. I don’t remember why, but it didn’t work as well as I hoped. Then I found a 9/16″ spade bit, so I used that in a drill press. I spent a lot of time aligning it every time before lowering the bit down to drill.


The dimensions are marked so I can have a visual reference when aligning the bit.

I didn’t take any pictures after the drilling, but I cleaned out the hole with a router – again I placed two pieces of wood 9/16″ apart and cleaned the route. This didn’t work as well as I hoped because those pieces of wood were kind of soft, and apparently had a couple of bumps, so the cavity didn’t turn out perfect.


I mean, it’s not that bad.

Cutting the guard with a router wasn’t so bad. Although I did use a table router that only works at a frighteningly fast speed. Also, remember to move the piece against the direction that the bit is moving in. So if the bit rotates CCW, move the piece from right to left, or down-up, and not left to right. Going left to right, the router will just throw the piece away from you.

Then came my biggest blunder. Maybe you can tell from this picture:


The half-circle for the control plate is on the left. I made a lefty pickguard and my friend is right-handed. I actually didn’t notice that at all until I sent him this picture and he asked if the picture was inverted. Of course, the template wasn’t cut for lefty. The problem was that I finished making the template at midnight after a really long day, so I wasn’t paying attention and taped the plastic to the wrong side of the template. So now I have a lefty double cut Telecaster pickguard. Very useful. So yeah, pay attention.

A couple of days later I went and cut a new pickguard. It actually turned out better than the first (nicer beveling, especially around the horns, better location of the pickguard screw holes).


Lastly, the countersinking of the screw holes – not sure what’s the proper Fender countersink bit, but I drilled a center hole and then used a 3/4″ countersink bit. I drilled a little, checked with a screw, drilled a little more, checked again, and repeat until it was the right size. It came out really well this way.

The template took the longest to make, so I wouldn’t recommend making a one-off guard for money. It was a worthwhile project and I learned a few important lessons, but god damn it took so long and I’m not at all interested in making another pickguard. Maybe I’ll learn how to use a CNC machine and make the next one with that.

Fixed the MIM Strat with the broken truss rod

Back in February I wrote about how the truss rod on this guitar broke.

I hemmed and hawed about what should be the replacement neck. The priority was to get as close of a match as we can to the broken neck. So 9.5″ radius, medium C profile, and rosewood fretboard. I looked at Warmoth and Allparts, and Warmoth are expensive and have very long wait times. Allparts are more affordable, especially the unfinished necks, but I wasn’t ready to get into spraying lacquer again. Then it hit me that I should be looking for an actual MIM neck from around that same time (2004), because that’ll be the closest match. A few weeks ago I finally found it, and they’re almost exactly the same:

The last three pictures show the fit of the brass nut on the new neck. The new neck is slightly wider.

The first thing I did when I received the neck was to check how straight it was. With the truss rod loosened all the way and no string tension, the neck was straight when sighting down from the headstock. A ruler showed a slight bow, but it was pretty minimal. I’m mentioning this because the broken neck has a pretty serious bow with no string tension. Like, very visible from the headstock. No wonder the truss rod broke. I should have taken a picture of it when I still had it.

The frets on the new one are in a better condition and it plays better than the old one past the 12th fret. My friend is very happy with it, and so am I. I’m really happy this project is finally behind me.

One last picture from when my friend picked it up:


Recording at Electrical Audio Studio B (April 30th-May 1st)

Last weekend my friend (I’ll call him NM) and I recorded a band (he also played drums on the recording) at EA’s studio B. We were both interns at Electrical, so we’re familiar with the studio and feel comfortable running sessions, my internship ended almost two years ago, so I had to refresh my memory on the signal flow in B.

The Neotek Series II in B was modified to have a fader reverse function, which makes the transition from recording to mixing smooth. This schematic (which is also hosted on Electrical’s website) shows it:


Steve himself explains the mod very well in this recent thread.

Basically, with the “mixdown” switch in the up position, the mic signal (or line if “line” is pressed) goes through the channel strip (Phase, EQ, Gain) and comes to the “monitor” pot. The monitor pot controls the output signal to both the output buses and the direct out. By the way, set to 7 on the dial the monitor pot is at unity gain.

In this situation with the mixdown switch up, the tape returns come up at the faders. As can be seen from the schematic, the Cue sends are fed from those line returns (i.e. the signal coming to the fader). On the other hand, the Echo sends are post-fader, and while E2 can be switched to pre-fader, it actually takes the signal from output of the channel strip.

When “mixdown” is pressed, the tape return signal goes only to the “line” switch at the top of the channel strip, and then pass through the channel strip to the fader and pan (and bus assign). In this instance, the monitor pot does nothing.

The console has 8 output buses and those are normalled to inputs 1-8, then 9-16, then 17-24 on the multitrack. So let’s say if the kick is bused to bus 1, it’ll show up on tracks 1, 9, and 17. This means that the direct outs aren’t normalled anywhere, so to use the direct outs, the engineer has to manually patch them to the tape inputs. This is very important and got me stumped for a little bit.

Headphone mixes are done using a Furman system so each musician can set up their own mix. The drum mix should be made using the Cue sends (remember, those are real pre-fader), and the engineer can hear the Cue sends mix by pressing the “solo” switches in each Cue section. Everything else (like guitar, bass, vocals, etc.) should be patched from the tape outs. The tape outs are half (full?) normalled, so doing that doesn’t override where the tape outs are going.

As far as microphones and preamps, it was all very standard. We tried an AKG D112 on the kick reso head and a Senn MD421 on the beater side, and they didn’t capture enough attack and “oomf”, so we swapped the D112 for an EV RE20 and the MD421 for a Shure SM98.

We used Josephson e22s on the toms (both top and bottom) going through the Neotek preamps, but when NM used mallets on some songs, the preamps were clipping. I don’t think it was audible, but I ended up swapping things around and passing the e22s through the MX-35s because that was the only thing I had left with a pad. I’m not sure it was a better choice, but I didn’t want to take any risks. I tried listening to the toms and see if they sound distorted as I was getting levels, but NM has crazy dynamics, and his playing was always getting more excited while recording. So even though I wasn’t hearing any distortion while getting levels, it was possible that it would distort as they were recording, and then the take is ruined. With that being said, I think I should have been more sonic-ly¬† vigilant – I was watching the levels, but I should have been listening more.

My biggest misstep was not taking better notes of the signal flow. It wasn’t so bad when we were tracking basics, but when we started overdubbing and I had to change inputs in Pro Tools, I found myself following patch cords on the patchbay. I should have used a track sheet, labeled the preamps, and just write where everything was going. Another important thing is to keep a table of what’s being fed into the Furman system inputs. I fumbled with that a bit too. Lastly, I kept bringing back the overdubs to the first 24 tracks on the console, but it really would have made a lot more sense to just bring them up on the 24 and above faders. It’s a 36 channel console!

We’re going to record in studio B again in July, so I’m going to implement all these things then.

Oh, and here are some pictures:


This is the spaghetti mess I was trying to follow. It was a bad idea. I made some mistakes!


NM at the board!


A fluffy coffee that was a joint effort between NM and me. The coffee-art looks like my cat Frostie!

Redoing a couple of cables (Canare GS-6 black insulation)

I always use Canare GS-6 instrument cables because they coil very easily and are 21 AWG. I discovered this week that the black insulator (under the shield) is conductive and should be stripped! It can short the cable if it comes in contact with the hot lead, and can also help the cable act as a capacitor, which rolls off high end. Man, I had no idea about that damn insulator, because when I made my cables ~6 years ago I didn’t RTFM.

I re-did them (some I was able to just strip down the insulation), and I don’t hear any difference in sound, but that doesn’t mean it shouldn’t be stripped away from the core conductor. Here’s what’s Canare is saying about it:


P-Bass fret level and sending off my Jazzmaster

My friend brought his P-Bass in a few weeks back for a setup. I set it up and eliminated most buzz, but it still had some he was unhappy with. It wasn’t anything that come through an amp, but he likes playing it unplugged, so I suggested I’d level the frets.

I leveled the frets and still buzzes. I leveled it again and added more fallaway, and it still buzzes in the middle of the neck. I tried swapping saddles, I tried shimming the front of the neck pocket, and even the back for the heck of it (even though it shouldn’t really help), and nothing really does it. I tried a straight neck and add some relief – nothing. Only thing that helps is raising the action. I think I did a good job leveling the frets, so I don’t really know what’s up with it. I’m going to need more time to think about it and figure it out.


In other news, my Jazzmaster sold on eBay. It was kind of bittersweet because it’s a nice guitar and I enjoy(ed) playing it, but not enough to keep it. USACG’s wrong corners on the body really get to me. So yeah ,that got sold and packed and out the door. So long, my first build! Maybe we will meet again. I hope you end up in the hands of someone who’ll be really excited to play you!