Picture of Build Your Own Marimba and Wrap Your Own Mallets!
Let me start by explaining where and from whom I got help. I used this website to get some basic information on how wide and long to cut the marimba bars. In fabrication I got help from one family member who helped rough cut some of the bars. The rest is all my work.

I am a mallet percussionist who loves the marimba. About four years ago I had a problem. Though I loved percussion, I had no way of practicing at home. There were cheap options. I could have bought a bell kit. But I hate the sounds they make. I wanted a very large five octave marimba but didn't have the money to simply purchase one (7000+ dollars). So I decided against all common sense I would build one.

The goal: build a five octave marimba, without spending a fortune. Use whatever supplies are available to keep the cost low. (The keys are made from an oak tree which was struck by lightning several years ago!)

I hope this instructable will inspire others but I want to give a word of warning. This is an extremely ambitious project and will likely takes a year or two for the average individual (like me) to complete.

Some notes about the included audio recording:
The marimba was playing using the same mallet across the entire five octave range. for this reason, the mallet I chose was a little too hard for the lowest note, and a little too soft for the highest note.

1-40 of 102Next »
PaulF46 months ago

Regarding volume; I did some experiments today and a larger resonator seems to improve loudness and maybe smooth bass-ness... I'm just getting into a first project with no experience with xylophones other than some googling and testing. It's a small one octave instrument from high C 523hz to soprano C 1046hz. The 1046 key is pathetic without a resonator, noticeably improved with a half inch diameter tube, barely noticeable improvement with a one inch tube and I'd say a lot louder with a 2 inch tube.

Now, laying them out, I've staggered the sharps in between so there is a double spacing and there is room for larger diameter resonators. I've noticed some designs place the resonators off center to the center of the keys when there isn't room, so presumably this is OK. You have to strike the key in the center but the resonator doesn't need to be centered, apparently.

Perhaps there could be conflict if the adjacent key is resonating into the wrong tube as it overlaps the key's territory but I think I've got this solved by staggering the sharps so their nodal point (peg) coincides with the center of the adjacent key. This won't be true for directly adjacent naturals. Possibly those conditions could be fixed by blocking the opening parallel to the adjacent key. That would be like the small hole in a guitar using the larger resonator of the guitar body? I suspect it won't matter too much. Looking at school alto xylophones in a single shared resonator box, they taper the bottom of the box a bit but it's all open to all the keys and not very precise.

I think it'll be worth trying to cram in the largest resonators I can manage. Even if there is some cross traffic, I've made the effort to provide the largest resonator ideally tuned to each key and the cross traffic won't be worse than the single box resonator models.

In the attached image, you can see the three test blocks with different diameter test resonators at left. I've only finished tuning and testing the high and low C at each end and both improve with a larger resonator.

Also, off topic, note how the naturals and sharps are staggered/overlapped so that the nodal points mark the center points of the adjacent sharp/natural. Kind of a clever layout inspired by the 4-row European arrangements. This won't allow dragging the mallet in a long trill across the naturals but should make for easier reach between sharps and naturals. Actually it might work since dragging across the nodal points of the sharps shouldn't produce anything. This arrangement was really motivated by fitting the longest keys possible on the vaguely piano-shaped board, which was what got me going on this project. The keys are also splayed out a bit radially (not parallel). Fun project!

2015-02-02 18.42.06.jpg
can you mathematically find the nodes?

Unfortunately, no. You can get relatively close, though! Because the density and grain of each key is different, you need to adjust the arch and your nodes to both where your frame is, and where the key's individual node is. One method is the salt test, and that worked quite nicely for us.

eraticus1 year ago
Just wondering why the resonator length is 1/4 the wavelength of the note? Does 1/2 or 1:1 work too? I am not a marimba player BTW, just in case the reason is obvious.
wombat99 eraticus9 months ago

Short answer: Either 1/4 or 3/4 would work. 1/2 or 1/1 will not work.

Long answer: Visualize one cycle of a sine wave. At the 1/4 and 3/4 point, the sine wave is at maximum amplitude. At 0, 1/2 and at 1 the sine wave is at the zero crossing (no amplitude).

So, to get the resonators to resonate at their loudest, you need to either cut them to 1/4 or 3/4 wavelength.

Hope this helps! :-)

FYI, back in the 1990's my wife and I built a bass marimba, and she created a web site that has some good information on it about our experience building the marimba:


wombat999 months ago

Nice Instructable and video. :-)

By the way, I see a number of questions (comments) about what kind of wood to use, etc. When we built our bass marimba back in the late 1990's, we used padauk. There are different kinds of "hardwood" -- basically "hard" hardwood and "soft" hardwood. Woods like oak, maple, etc are "hard" hardwood. Wood like rosewood, bubinga, paduak, etc are "soft" hardwoods, and these are what are typically used to make bars as they resonate much better than "hard" hardwoods.

Without trying to blow my own horn or take anything away from your Instructable (again, it's great), and the spirit of sharing information, I want to mention that my wife created a website that details our experience in building the bass marimba:


So, if I'm okay with a little more expense on the wood for the bars, where would i even begin to look for some good rosewood to purchase? would i purchase it in 4/4, 8/4, or what? how many board feet. it may be a dumb question, but i don't even know how to start acquiring wood.

cjs12981 year ago
Do you have any measurements on the frame of the 5oct marimba? Also, do you prefer wool or acrylic for mallet wrapping? I'm a freshman in high school, so I have little money for mallet wrapping.

wool wrapping is ideal because it is more resistant to wearing out. i have a lot of 100% acrylic yarn mallets that wear out within a year, which is just not worth it.

how can i tune the overtones to make it more professional sounding? also, can i sand the arches to tune the keys? i want each key to be a certain length so it looks pretty. im very picky about the look cuz im going to be using some padouk floorboards that are 3 inch wide by 1/2 inch thick (cut in half, of course) and theyre not cheap, so i wanna make sure everything works perfectly
RocketScientist (author)  luvtheteddie6 years ago
OK... To tune the overtones you will need one of two things,

A. Strobe Tuner
B. Audio Spectrum Analyzer

I suggest the Strobe Tuner especially if you are a musician. Personally I consider Peterson Strobe Tuners to be the best and they also make a Strobe Tuner app for the Iphone/Itouch. I only included the Spectrum Analyzer to give you an alternative.

Yes you can tune the fundamental frequencies and the overtones of all bars just by shaping the undercut of the bar.
Before we get into how to shape the curve, let's review the proper ratios between the overtones. For Marimba and Vibraphone builders it is 1:4:9.88. For Xylophone builders it is 1:3:6.

Now, to tune the overtones you should follow the template in the pictures below.
To tune the Fundamental frequency (1), sand in the center.
To tune the second overtone (4), sand just outside the center
To tune the third (9.88), sand close to the ends of the arc.

Now here is where it gets tricky...Changing one overtone, will change the frequencies of the other two!
For this reason, you have to first get the ratios between frequencies correct, and then sand evenly across the curve until you arrive at the fundamental (hopefully with the ratios intact.)

If you mess up and tune something too low, you can not (to my knowledge) fix the problem without reducing the length of the bar.

Additionally, you will notice the overtones become increasingly more difficult to tune as you begin to work with higher and higher notes. This is because the sample size/duration/sustain of the higher notes become shorter and shorter. This also makes the overtones more difficult to hear. So do you want to tune the overtones of the upper register? I don't know. You'll just have to play it buy ear. (I'm sorry... couldn't resist a bad pun)

I highly recommend you read this research article:
Nonuniform Beams with Harmonically Related Overtones for use in Percussion Instruments
by Felipe Orduna-Bustamante
published in the Journal of the Acoustical Society of America
December 1991

Finally... Good luck on your build. If I forgot to answer something just let me know. Be sure to tell me how it works out!


(Seriously I'm competing against some spectacular projects and I need all the support I can get.)
Hi! First of all I want to thank you for this good instructable. I decided to build a marimba. I cut all the 5 octaves from white oak, maple and cherrywood. I started tuning the bars. I started with C5. I have one problem. can't find the nodes, because the salt doesn't make a line. It bounces off the bar after a while. So I decided to use 2/9 of the bar length as the nodes. But there is one more problem I can't hear any overtones. What's wrong?
Thank you,
Lukas Cerveny (15 yrs, from the Czech rep.)
you got my vote. one more question... how bog should i make my arches?
RocketScientist (author)  luvtheteddie6 years ago
In general, the arch extends the full length between the nodes of the bar. (But this rule does not apply to the higher bars). So instead, I would suggest sanding the arches wider as you tune the third overtone. When you sand the "3" sections as in the pictures above, sand the arch wider as well as deeper. As a side note for anyone interested... You can tune the fundamental and second overtones just by sanding out a rectangular shape rather than an arch. The only reason for the arch, is to get the third overtone.
handers3 years ago
how did you calculate the legnth of the bars for oak ?
RocketScientist (author)  handers3 years ago
I used lengths similar to the La Favre marimba I linked to in step two. You can experiment with lengths and graduations but it's much easier to reuse the lengths found there. Those lengths worked fine for all the frequencies I needed from the oak.

Did you use the measurements for the La Favre marimba or for the graduated marimba? what is the difference?
Bianca13062 years ago
Would Red Oak, Oak, Poplar, or Maple work best for this?
cpatgardner2 years ago
I noticed your comment that you cannot raise the pitch back up (to your knowledge) if you go to low without shortening the length of the bar. I've made a few (not as nice as your) but when I go to low, I raise the pitch by removing material close to the ends giving the same effect as a shorter bar. I use a 5/8" or 3/4" forstner bit and drill these holes on the underside of the playing surface where they wont be seen. I drill them equal distance from end on both ends of bar. I drill to equal depth on both holes until I reach desired pitch. I try avoid this because you have remove quite a bit (drill deep) to raise pitch just a little. Just thought I'd share that trick I figured out.
kpalvin963 years ago
When you say push the needle through the top, do you mean through the core? Or just through the yarn?
RocketScientist (author)  kpalvin963 years ago
Just the yarn.
i was wondering if you ever got around to building a vibraphone? that would be cool. i am really interested in how you would build the dampener system. as an engineer i am sure that you could come up with something that is more practical (or at least more simple) and doesn't add so much pressure to the frame (the pushing and pulling unevenly).
RocketScientist (author)  1bad_drummer3 years ago
A vibraphone has actually been turned into an honors thesis project believe it or not. (I'm an engineer too) I've spent the last few years focusing on the keys but I have had a few thoughts on the damper.

I've never liked that most Adams vibes had only 1 spring in the center of the damper. I'd prefer 2 because I think you'd get a more even force distribution. And as you've probably noticed, Adams vibes have their damper pivoting on arms. That works but those arms are ugly and I'd rather use some linear bearings underneath the frame.
i am definitely looking forward to what you find in terms of the bars!! but as far as frames go i am really interested in as portable a frame as possible, lightweight and simple. i was thinking something along the lines of how i could streamline the design as much as possible for a gigging instrument as opposed to a concert instrument. replacing the resonators with a pickup, using cables like a remote hi-hat uses to construct a dampener system that didn't need a crossbar to support it....

but still i would love to see what you come up with and i hope that you publish it!
TruthHunter3 years ago
Great Project...Very ambitious. I liked your string supports.

I was wondering how Oak would work for a marimba, so I was very glad that you posted a sound file. It is obvious that oak isn't as good as Rosewood or Paduk. Still it makes a nice practice instrument.

BTW, have you considered going back and upgrading the bars? You could get enough Paduk from http://www.exoticwoods.com for about $300 to do it.
I see a lot of discussion as to the best woods to use. Cherry and Walnut might be a little better than Oak. Most USA hardwoods don't have much resonance. The only 2 that I've seen that rate with the Tropical woods are Black Locust and Osage Orange. If you can get these locally, they could do the trick, but to buy from suppliers, they are as expensive as tropical woods.
I thought at first you were making a thumb piano (which some call a marimba). Would it be right to call yours a type of xylophone or is it different?
A larger kalimba could also be called an 'mbira, which sounds quite similar to marimba. Or at least, that is what I have read somewhere, so take it with a few grains of salt.
The "thumb piano" you refer to can be known as a Kalimba. But yes, this is similar to a xylophone. See below for more info. http://www.differencebetween.net/object/difference-between-xylophone-and-marimba/
dabad4 years ago
Well you drill the hole, are you supposed to drill to the center? Or just enough to place the dowel in?
RocketScientist (author)  dabad4 years ago
Drilling to the center would be best but that's more difficult in some materials than others. In the end you're just trying to get a good 'hold' between the head and dowel. It's very embarrassing to have the head suddenly fly across the room in the middle of practice, trust me.
Am i blind or is there no video of the finished piece in action?
RocketScientist (author)  nodnodwinkwink5 years ago
You're not blind! There is no video! I've tried to make a few videos and audio recordings in the past but unfortunately my video equipment is cheap! Every time I've tried the notes always come across as brittle and clunky, almost like a xylophone. In person the instrument sounds very "warm" and seems to hum like a marimba should. Thanks for the question, hopefully I'll be able to upload a video one day when I have better video equipment.
maybe Darmani can instruct you in the proper modifacation of video equipment to use for a musical instrument .... oh as long as it is hand made, non electronic and made out of Venus bogon moth tree which as it turns out is the third largest tree in the 5th smalles section in the portion known as kruzijenkle on Venus .... I should know cause I've been there!!!
Xzen214 years ago
When wrapping mallets, should I use more wraps for a softer mallet or less wraps?

I am using a rubber bouncy ball core and acrylic yarn like you would typically find in a craft store.
RocketScientist (author)  Xzen214 years ago
The more times you wrap it the softer the sound will be. Depending on the yarn and core you use, 150 to 200 wraps will give you a "medium" mallet.
Well, I guess I basically need a set of mallets to use for the lowest 2 octaves, any suggestions for that?

Thanks for the quick reply on my first question by the way.
RocketScientist (author)  Xzen214 years ago
No problem! For the lowest octaves you'll probably want a soft mallet. I would suggest maybe 175+ wraps. In that range of the instrument softer mallets will give you a very warm tone, sort of like humming. When you reach about 150 wraps, you can wrap some yarn around the stick and strike a few bars with it. If you like the way it sounds, go ahead and start the crowning step. If it's too soft it will sound harsh and "clunky" so just play with as you're wrapping it until it sounds like you want!
Thanks, will give it a try.

I had been getting a loud "clunking" sound while tuning the bars (had been using a hammer from the inside of an old piano) and the actual notes were not very audible, I am hoping that using an actual mallet designed for marimba bars will correct this issue.
JcBeaver4 years ago
what exactly is the difference between diatonic and chromatic?? just wondering
Dhha JcBeaver4 years ago

Diatonic is when the note belongs to a scale, let's say C major scale. F# (sharp) doesn't belongs to C major scale, it isn't diatonic.

Chromatic is when notes moves half-step by half-step, from C to C# to D to D# to E and so on (Also B to Bb to A to Ab, etc). You have upward chromatism (sharps) and downward with flats.

I hope this helped :-)
1-40 of 102Next »