# Marimba bar tuning and temperature

Marimba Australia
TL;DR Summary
What effect does temperature make on tuned marimba bars.
Hi all.
I make marimbas - www.marimbaaustralia.com - have a look.
I also invent new versions for schools.

I DO NOT KNOW algebra almost at all, so go easy on me please!

I try to tune to exact Fundamental note frequency (and the first overtone as well, if possible).
I also try to tune when it is 21 degrees Celsius, because that is close to what a school will heat/cool to.

I want to make an excel file that allows me to tune when the temperature is lower and higher.

I can only think of making a whole set, put them in a room I can control and heat and cool, then record the new pitch so I could tune sharp or flat as required and it will return to 0 cents deviation at 21 degrees.

My question - is there a file already made for this, or another way to get the results apart from doing as stated above??

I am aware that different bar sizes react differently - just hoping someone may have the details already for me please. Thanks.

Mentor
This article has a bit of Science in it along with what he (Youhass) does for tuning. Tuning is an issue for him as well. No math.
https://www.pas.org/Files/0708.52-57.pdf

Out of curiosity where is your A4 - 440Hz or slightly higher? I've played pearwood recorders and VERY old music with completely different scales and A4 as low as 408Hz. ...and it drove us/me crazy.

FWIW An A4 set to 410Hz was common in some German orchestras until recently, I'm told because early orchestras used it going forward.

You have to match A4 to what your client wants I assume.

Marimba Australia
A4 is always set to 440Hz. I don’t know of anyone tuning to anything else now. Thanks for the reply.

Marimba Australia
Just read the article. Interesting. I am amazed at how badly marimbas are tuned and made by handy-people. I am not making \$40,000 instruments but I hope mine are better tuned than what I see generally! :-)

I believe different species of timber, with varying moisture content will have different thermal coefficients. That will initially make it difficult to tune at non-standard temperatures.

You will need to experiment. Measure the frequency at one temperature, stabilise at another temperature, then measure the frequency again. Record the species of timber and the dimensions of the bar in your lab notes, along with temperatures and frequencies. Those accumulated records will allow you to extract the guidelines needed to refine your tuning predictions later.

If an axial hole was drilled in both ends of the bar, the mass would be reduced and the note should rise. Winding metal grub-screws equally into the holes should lower the note. You can then trim the frequency without needing to cut the timber.

I would consider lacquering the surface of the bar to reduce the moisture variation over time. I expect the weight of the lacquer would lower the frequency.

I am not sure of the mode of the harmonic oscillation, but it may be possible to independently fine tune the fundamental and the harmonic.

What criteria do you use to select timber for the bars? What species?

Marimba Australia
Thanks.
For schools the timber is always a species we call ’Tasmanian Oak’ (I am from Tassie) or ‘Victorian Ash’. They are Eucalyptus Gum. A very hard wood, endemic to Australia. I always try to get the heaviest piece I can (density) and with the grain going from face-to-face, not side-to-side, as that tunes better.
The bars are different sizes and are lacquered, for the reason you mentioned.
Tuning at the end will raise the pitch a little, I also first tune the torsional note - on diagonal corners of the underside at the lowest pitch side-to-side in the bar. That does what I term ‘set the bar free’, so it tries to sound one note and not many fighting each other. Much better sounds can be obtained then.

You have suggested what I now believe is the only way to get an excel file of each bar and the deviation tuning for particular degrees. That is, build a whole set at 21c, raise/lower the temp and record the variations once stabilised. That way I can have a reference sheet in front of me and tune sharp or flat accordingly so it will return to 0 cents deviation at 21c. (20c would also be fine).

[Mentor Note -- Minor Spam deleted from post]

Last edited by a moderator:
Staff Emeritus
Is it the block that needs to be tuned or the tube underneath?

Mentor
You are making assumptions about wood moisture gradients, grain direction and other things.

You want to make a good product. Based on Science.

So. You need to read: 'Understanding Wood' by Bruce Hoadley Period. No "ifs", "ands", or "buts" about it.
Nails moisture, humidity and wood reactions to them.

Maybe three pages of math - Young's modulus and such, you can ignore it. You might want to learn about creep. === There some numbers associated with that, but not needed to get the concept.

Your library can get it via interlibrary loan, and if you visit an engineering college they will probably have a copy.
I have owned one since it was first published.

Another source is the FPL. ( US Forest Products Library). All on the internet.

https://www.fpl.fs.fed.us/

jrmichler
Marimba Australia
Is it the block that needs to be tuned or the tube underneath?
The bars. Pipes are Ok.

Marimba Australia
You are making assumptions about wood moisture gradients, grain direction and other things.

You want to make a good product. Based on Science.

So. You need to read: 'Understanding Wood' by Bruce Hoadley Period. No "ifs", "ands", or "buts" about it.
Nails moisture, humidity and wood reactions to them.

Maybe three pages of math - Young's modulus and such, you can ignore it. You might want to learn about creep. === There some numbers associated with that, but not needed to get the concept.

Your library can get it via interlibrary loan, and if you visit an engineering college they will probably have a copy.
I have owned one since it was first published.

Another source is the FPL. ( US Forest Products Library). All on the internet.

https://www.fpl.fs.fed.us/
Thanks. I will check that out.