Casey Connor
Shows
 Albums Bio/Press/Etc Lyrics
 Links
Last update: February 24, 2008

How to Build a Cajon
This page documents the design and creation of a cajon.

I am neither an experience cajon builder or an experienced woodwooker in general. I wanted to build a cajon and found a lack of material online to support that endeavour, so I have put together this web page. Comments/corrections welcome. This is a great project for musical people that are interested in instrument design but aren't quite ready to jump into something like a guitar.

It's properly spelled "cajón", but in case search engines don't pick up the word correctly, "cajon" is used here.

Here are the plans for the cajon depicted on this page, in various formats: pdf|png|jpeg|illustrator

This page owes the biggest thanks to Germán Ocaña who provided most of the information required. Check out his webpage, he makes professional-level cajons and other instruments. He has two main PDFs that I saw regarding the cajon. The english in the first is pretty rough, and the second is in Spanish, but even for non-speakers they are invaluable nonetheless. This page will assume that you have read these documents. This page is intended as a supplement.

Here are some other documents I came across that may or may not be helpful:

Audio recordings of this cajon will be posted ASAP...

In building this cajon, i had hoped to make an instrument that was a little bigger, a little bassier, and generally more burly than the standard cajon. I was worried about passing some point of diminishing return by enlarging the cajon without having adequate knowledge to simultaneously tweak other factors in the design (thickness of the front, shape of the box, size of the hole, etc), but i think it worked reasonably well. The tapa of the cajon presented by Ocaña is about 12.5" by 18". This tapa is about 14" by 21", with the other measurements similarly scaled. This increases the internal volume by a factor of about 1.4 or more, which is no kinda small potatoes.

Another unusual feature of this cajon is that i made a hinged door on the back. To avoid impacting the kick sound too much, i designed it to seal well, and to stay closed with significant force (as opposed to a simple latch, etc.) Since it was the first cajon i made, i knew that i would be inside it a lot, adjusting things, installing new gadjets, etc, so it was worth it to me (even though the door took a large percentage of the total construction time.) The risk is that the bass sound would suffer; instructions i read from Ocaña and others make a big deal about how well the box is sealed, glued, etc, and certainly such a door compromises that issue. My feeling is that the seal is not so great, but so far it sounds good. I may add a second layer of bicycle rubber to make it a tighter seal, or replace the sealing material altogether.

What i would do differently next time:

So, the wood. My impression is that the back, sides, top, and bottom of the cajon are to be made with the thickest heaviest wood you care to use, as if one were building a bass speaker cabinet. Generally 1/2-inch to 3/4-inch plywood is appropriate. I used 1/2-inch which, frankly, seemed plenty strong to handle any bass oscillations that came its way, and seems to be the most standard size used. The face, or tapa, must be rigid, but Ocaña points out that it shouldn't be too dense. The selection of the tapa wood is much more important. Sounds like the typical soundboard ideal of "strong and light" comes in to play here, perhaps with more emphasis on the "strong" than you would have with, say, a guitar top, since you will be striking the tapa directly. Note that the normal playing of the cajon is somewhere between a membranophone and a resonant top: in terms of the "kick" sound, you are largely relying on the force of the air generated by the compression of the space inside the instrument rather than the "ringing" of the top, but without the ringing of the top (both at the 40-60 Hz kind of zone and the higher overtones) you have a pretty lifeless thud. Since you don't play the tapa in the center (rather, more towards the top) it maybe doesn't have to be the toughest wood in the world. Many pro cajons use solid wood tops. I don't know how much that's about sound and how much that's about rare-or-fancy-wood fetish. This may be one sound application where plywood works just as well, but i really can't say.

The plywood for the tapa can seem hopelessly flimsy when you first evaluate it, but once it is all screwed down to the face it should do fine. This cajon used 1/8-inch birch 3-ply. There were alternative "strong and light" plywoods at the supplier. I would consider using a thicker wood for the face, but again, without playing lots of cajons, i'm not sure. All of the candidates seemed to have much more flexibility along one axis than the other (apparently because with 3-ply, the two outer layers of wood are aligned in grain, with only the single inner layer at 90 degrees to those). This undoubtedly affects how the top resonates (maybe for good, maybe not); i took a guess and cut the tapa so that the outer grain ran vertically. If thicker wood could be used for the tapa, perhaps 4-ply wouldn't have this issue, for whatever that's worth. "Aircraft-grade" plywood sounded tempting, but my rough information on that is that the grains of the plywood layers are not 90-degrees to each other; so it's not as rigid and even more biased to flexibilty along a certain axis (apparently because the wood is intended to be shaped.)

It seems that there must be some technique to shaping the tapa, but i found no information. Ocaña does have instructions on how to flatten the tapa of store-bought instruments, so perhaps flat is best. I did iron my tapa (low heat, no steam) before affixing it, as he recommends doing for tapas that are warped.

This cajon has the typical strong frame that goes behind the tapa. But there is also an additional frame in the back to form the frame against which the door can seal (only two pieces, shown, required for that). These pieces are all made of hickory (a hardwood is strongly recommended; also, in this design, standard zither pins were used as the snare tensioners, and such pins work best in hard woods.) Shown here are the frame pieces (not actually glued; depending on your methods, it probably doesn't make sense to assemble the frame separately.)

The unglued wall pieces are also shown freestanding to illustrate the basic idea.

Marking and cutting the hole. One way to look at the cajon is as a Helmholtz resonator. Not sure how accurate that is, really, since the volume of air in the cajon isn't doing to much resonating, but assuming that the air spring's oscillation after a strike influences the kick sound, one might infer that the size of the hole would affect the "pitch" of the kick. The original plan was to keep the disc cut for the hole and use it as an adjustable slider that could narrow or widen the hole. Testing it out, it seemed largely irrelevant. One might then infer that the size of the hole doesn't matter so much. Even so, intuition tells me for some reason that perhaps this hole was a little too big (it was scaled up a bit because of the general oversized nature of this cajon.) The exit hole for bass reflex ports is often shaped in certain ways. It may be worth experiementing with those ideas. It's also tempting to wonder if the internal reverberance of the cajon maximizes the energy delivered from the tapa: perhaps the frequency is low enough that it doesn't matter, but perhaps some shaping of the cajon or strategic insertion of wood inside would help keep those waves coincident (in-phase). One wonders too how the location of the hole affects this, if at all.

The hole on this cajon is on the side, close to the floor. The cajon has the most punch when played against a wall, or especially in a corner. This leverages the bass effect: when placing subwoofers or a bass amp in a room people often choose corners or, failing that, against walls, because bass frequency waves, with their long wavelengths, are less succeptible to canceling each other out when reflected off of nearby surfaces and redirected in a similar direction. Thus one can "collect" the energy emanating from the cajon in all directions, bounce it out of a corner, and focus it into the room as a unified force. If played in the middle of the room, the waves go everywhere, and their reflections of walls may intersect out of phase, thus reducing the bass sound dramatically.

Thus one sees the wisdom of having the hole in the upper back: you can play against a wall or a corner. When you get up against the wall and tilt back on the cajon (a common playing style) the hole is right where it should be. There may be advantages to having the hole in the side; I wondered if it somehow improved the bass by forcing the internal aerodynamics to a lower resonant pitch. Perhaps it is easier to mic on stage because a player tilting forward and back as they play can't accidentally change the distance to the hole from the mic (nearly as much, anyway). But having the hole on the side makes it difficult to find an appropriate place to sit to maximize the bass sound: you're always facing along the wall unless you're lucky enough to find a corner (harder than you think). Please lend your wisdom on this subject.

Gluing the frame together using the elegant custom clamping/roping techniques that my paltry toolset affords. A good time to make sure everything is straight, 90 degrees, etc. No nails used in the cajon, just glue. Given the way the boards meet, it's plenty strong enough to hold the player.

Gluing the frame into the cajon. Due to slight errors in measurement and cutting, the frame pieces were a bit short; thus slivers of hickory were inserted to fill the gaps when the gluing was done. Note the custom clamp improvisations and the stacked wood and old paint can (tip: pink is usually the heaviest) for bracing. Clamps would have been a lot better and easier, obviously. The image on the right shows the back side, where the door will be. Only two hickory pieces are required there. Hopefully it can be seen that the door will set flush against those two pieces and the two edges of the side walls. The overlipping top and bottom prevent shearing stress on the hinges from the player sitting on the cajon.

The hinges were sunk into the wood so that nothing would catch when the cajon was moved or slid on its side. It's also nice to take into acount the sensual aspects of touching the instrument. Supports were put in place to prevent the chiseling from fracturing any of the recent glued joints. The hinge that meets the door lays on the inside. This creates a small space between the door and the frame against which it will lie. All along the frame for the door was glued a strip of used bicycle inner-tube rubber, which was conveniently just about the same thickness as the hinge, the idea being that a good seal is important and a material with some give and sealing properties was needed to achieve that. It may not be quite enough: two layers of rubber may have been better: one glued to the door, one to the frame, and the clasps adjusted to appropriate tension. Otherwise another, thicker material may be used, but as it stands it seems to work well enough.

The door was a lot of work, but for an experimental cajon, it seems worth it. Of course a cruder, less-easy-to-use door could have been fashioned, although screwing into the edge of plywood is generally not realistic. The clasps, though they may appear to be casually located, are actually very carefully placed so that when the door is closed and the clasps engaged, the door is under tension against the frame. It is nonetheless possible to notice the door moving slightly when the kick is hit, which indicates that it is absorbing some of the sound despite the clasps. The shape of the clasp is such that maximum tension is reached before the clasp is fully closed, at which point the tension backs off. This keeps clasps closed when used normally, e.g. to keep a cabinet closed. In this case, they could be bent out, so that maximum tension is sustained when they are fully closed. Then they could be placed and adjusted appropriately and a better seal attained. Additionally, one could place more clasps along the top and bottom of the door frame. Make sure you give the top and bottom clasps enough space from the frame pieces so they can function.

The edge of the door that meets the hinges needs to be shaped, of course, in order to allow it to open (in this picture, the door is the piece on the left, and the visible hinge is attached to the side wall). For fine-tuning the snares, it is recommended that the door be able to open wide enough to allow fiddling with the snares against the tapa while the cajon is tilted back at a playing angle, since the tilt can affect the force with which the snares meet the tapa. On the other hand, your snares may not be so sensitive.

I couldn't find large, tough, rubber feet with holes, so i used rubber stoppers with a screw and a large washer. I drilled a small hole for the screw, then drilled with a larger bit halfway into the stopper, sorta forced the screw and washer in there, and voila. Drilling into rubber is difficult and probably dangerous. Perhaps freezing the rubber first would help. Can't honestly remember if I tried that. :-) Traditionally, the feet in the back are large to enable the player to tilt back. The feet in front are less important. When designing your feet, watch where you're drilling and plan ahead for the tapa screws in the front and the door frame in the back. I intentionally mounted the screws for the feet into the hickory frame inside, for added strength.

Ah, the snares. Perhaps the strings weren't thick enough (classical guitar "A" strings), perhaps they shouldn't have been so far up in the corners (longer strings = more mass), perhaps my notion of the "snare" sound is different from typical cajonares, perhaps my tapa is warped; but the snares seemed ineffectual, even after considerable tweaking. Here, anyway, is how they are set up. The image on the left shows the tapa before screwing and the various pilot holes in the tapa and the internal frame. The red lines show where the tapa screws will hold it (there are only three along the top of the tapa, the middle one was already screwed in once, thus the hole looks larger). Also visible are six large holes in the frame in two sets of three (which will hold the zither pins from the inside) and, above each of them (actually below them in the image; one is indicated with the blue arrow) is a small hole through which will pass a snare string (i broke and lost drill bit in the string hole rightmost in the picture, thus a secondary hole had to be drilled). Leading from the exit of these snare string holes is a groove where the string will lie (as demonstrated by the yellow line.) The string passes along the inside face of the tapa, through the groove, into the string hole, and finally wraps around the zither pin on the inside-face of the hickory frame. On the other side of each string, it passes through a similar groove and through another small hole; it is simply knotted on the other side of the hole to hold the string. The grooves are cut to both prevent the string from pressing or bulging the tapa (which could also affect the seal) and so that the string is not unduly stressed as it is tensioned or released. It may be a good idea to grade the entrances to the string hole so there isn't such a sharp edge for it to pass over. I have broken one string so far in my snare tweaking, and it was at this edge.

There are many, many variables in the isntrument that affect whether or not the snare works well, not to mention a wide variety of opinions as to what a good snare sounds like, so it's hard to say what kind is best.

Standard zither pins (a.k.a. "tuning pins") are available online or, better, from your local friendly music store. The proper way to use them is apparently to drill the hole and then hammer them most of the way in (screwing them in can apparently cause them not to hold their tune as well over time.) It's probably not crucial in this application since the snares will never be at particularly high tensions. Ocaña describes the tension as just enough to hold them against the face to cause a rattle.

Ocaña recommends a small piece of tape along the middle of the snare to help ensure contact with the face. This didn't help me. Perhaps my tapa doesn't resonate enough after a strike, but there didn't seem to be enough leftover energy to drive a snare that required driving. Thus i changed over to a "rattle" system: the rattles get their energy from the initial strike, and the mass of the rattles (coupled with the bounce of the snare string which was repurposed to mount and tension the rattles against the face) allows that energy to continue sounding after the strike. It worked well (after a lot of time spent trying different shapes and sizes and careful tensioning of the strings to get just the right rattle without lingering noise, etc.) The wire used is 12-guage copper grounding wire from some extra electrical cable (just what was around, and it seemed massive enough). The finer wires (from old twist ties) are just there to hold the rattles in the optimum locations.

The rattle setup was a good start: an edgy, barky snare, but the rattles are, well, rattly, and the snare sound was still on the short side and lacking in higher-frequency shimmer. I took old "E" steel guitar strings (.52 or .54 inch diameter, not sure which it was), carefully unwound the wrapping from the core, and gradually made 10 5-inch long pieces that i joined with some hammered copper 12-guage wire and electrical solder to form a somewhat fancier custom snare that went in the other corner (the strings were slackened and pulled out of the way). Two zither pins were used to mount the snare assembly: the top pin could be used with a string tied to the strategic bend in the wire (near where the actual snares connect) as a tensioner of the snare, but it didn't seem necessarry, and is not done in this picture. (The old classical guitar string snares with some rattles are still hanging around loose in the picture, they are not contributing to the sound.) A whole lot of time was spent straightening and tweaking with the very finest degree of accuracy the lie of these snares so that they sounded but did not rattle excessively, etc. The combination of the two snare systems sounded nice, and for now that's where it lies. (Note that the snare tweaking was done after construction, since you need to test it with the real instrument, but it's included here out of order.)

There are, of course, many other snare designs for the cajon, which you can dig up on the web. Some use snares all the way across the center of the face, which i would expect to reduce the separation of the kick and snare sounds. Many cajons (such as those from Meinl) allow the snare to be disengaged completely, to allow a more hollow, ringing high sound. I didn't bother, given my personal intentions for the instrument.

Ocaña suggests you take a swallow before you hit it the first time since you may well be disappointed. It's true. But play around with it, stick it in a corner and play it, get used to your tapa and how to strike it, and it's true character will likely reveal itself to be a lot better than your first impression. After giving it some time and practice, i have grown to love this one as it is, and i haven't taken it much further than the original design (e.g. by coating the inside with glue as Ocaña suggests, trying other tapas, placing wood inside to direct the waves, whatever). The snare, though: You will probably want to spend some time on your snare?

Ocaña recommends not gluing the tapa. Apparently some builders glue it, probably because it's easier to make but possibly because it seals better. Many cajon players like the top corners of the tapa to be free (thus the lack of screws there, or with glued tapas, the lack of glue in the top third) in order to get a slapping sound as the tapa corner hits the frame (Inserting paper shims in the corners to raise them up a tad may help you get a good corner sound.) The seal may not be as gravely important as it's made out to be, but perhaps the overall seal over the perimeter of the tapa really does matter. Having the corners lifted with shims in conjunction with slightly loosening the screws that abut the corners didn't seem to affect the bass sound much, as compared to no shims and all the screws tightened.

Ocaña recommends being generous with the number of screws. The tapa of this cajon lifts off a tiny amount at the edges, probably because of a slight, natural deformation from the tension of the screws. I am confident that the seal is nonetheless good. Drilling pilot holes is probably a good idea for the tapa screws to as to avoid splitting the hardwood frame. Before affixing the tapa, of course, you want to ensure (through sanding, if necessary) that the internal frame is appropriately flush with the edges of the plywood sides, top, and bottom. (I may even have tapered the plywood back a hair to make sure that the tapa contacted the frame fully, which may also help explain the slight gap between the tapa and the plywood.)

Straight-up organic locally-grown homemade fresh-squozed beet juice was used to dye the wood. As for the tapa, Ocaña recommends rigid finishes only as oil would soften or dampen the resonance. Not sure if he feels the same about the other faces: I think he advocates using similarly rigid finishes on the interior at least (or the glue method he described). It's hard to imagine that it matters on the external non-tapa faces. Anyway, given the horrid environmental and health aspects of finishes, beet juice was used. As for finishing the tapa, it was left unfinished and we'll let nature slowly break down the plywood (which, of course, is itself full of toxic glues and will probably last lifetimes.) After the beet juice dried, a light pass with unrefined walnut oil, which is purported to be a "true drying finish", and is about as natural as it gets in that department, was applied. The inside is also unfinished.

Beet juice fades very quickly. A test piece faded dramatically after a week or two of mild sunlight. But it's beautiful, and you can always add another coat. Once totally dry it does not rub off on your clothes, though if it got wet it may well. I don't know if the walnut oil will make it harder to re-juice. I keep it under a cloth when not playing it in the hopes that the fading has more to do with ultraviolet light. Maybe the oil will help if in fact it has more to do with oxidation. Time will tell. After three weeks it still looks very strong, as opposed to the test piece, so maybe the combination of oil and storing under a cloth will work.

The finish work basically involved making the edges meet up through the magic of sanding. All three of the top non-tapa edges were all generously rounded - it's worth it. After sanding, the moisture in the beet juice will cause the wood to regain a rough texture, so don't spend too much time making it perfectly smooth, as obsessively fun as that is.

Isn't it so beautiful?

Micing the cajon:

I mic the cajon in my home laptop studio with an AKG C535-EB, which is not a kick drum mic. Sensitive condensers may be damaged by the high SPL, but the feedback at the TapeOp message boards assuaged my fears, some saying that a kick drum will generally have a higher SPL than a cajon anyway. The best placement for the kick sound seemed to be with the head of the mic just inside the hole (in line, essentially, with the side). Haven't experimented much with the snare side of things, but it's easy to check that out on your own.

Other Links: (see top of page for main info)

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 Unported License.