An Interesting Dulcimer Experiment

Just share stories or offer advice

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Thu Jan 07, 2016 9:21 am

Something a Bit Different – The Sound of Top, Table and Floor

Whilst I was fiddling with the piezo transducer for the previous post I could feel the table that the dulcimer was sitting on vibrating as I was playing. It occurred to me that this should be investigated.

Robin Clark proposes that mountain dulcimers fitted with certain tone enhancing components sound better when played on tables than when on the knee. As it happens, my test dulcimer is fitted with just such devices (three small wooden feet) and is therefore a good candidate to examine the effect.

It has also been suggested by some players that resting a mountain dulcimer on a table reflects the vibrations of the back plate off the table and outward, making the instrument sound louder. Others, including me, claim the effect is because vibrational energy is transferred from the dulcimer to the table, which then vibrates and makes sound in its own right. Here is an easy way to test the “vibrates the table” proposition.

A small piezo button transducer was stuck to the top plate of the dulcimer, then to the table the dulcimer was sitting on, and finally to the hard floor below another table. Short test tunes were recorded. The setup was like this:
Top_Table_Floor_Lo.jpg

The results sounded like this:

Top_Table_Floor Transducer.mp3
[ 939.63 KiB | Viewed 7075 times ]

Since the piezo transducer is designed to pick up mechanical vibrations, it is clear that the dulcimer is setting not only the table into vibration, but also the floor the table is sitting on.

The vibrations from the table were fairly large, about half as big as the dulcimer top vibrations. The floor vibrations were only about one tenth the amplitude of the table vibrations, so gain had to be turned up. Because of this mains hum was strong so I had to filter that recording to remove it.

In the process I learned that small weights on the piezo (say about the weight of a plastic pencil sharpener) have a large effect on the sound quality. No weight => lack of bass; too much weight => big boomy bass; just right weight => best balanced sound. I spent some time in matching weight to best sound, but it could have been even better if I’d wanted to make more effort. Something to consider when sticking these things on.

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby Truckjohn » Thu Jan 07, 2016 6:12 pm

rtroughear wrote:And TJ, thanks for the comments, but I'm at the stage now where I suspect I know hardly anything. So if you have any revelations, wouldja let me know ....


You know a lot more than you take credit for. The challenge is turning the results of the experiments (science) into real life results (engineering).

One of the most telling things you can do is to use some statistics and some intuition to rank the effects of changes. Turns out that the effects are not straight cumulative but more along the lines of sum of the squares. Say the effects you measure run like this (totally hypothetically made up).
Fretboard stiffness = 75
Fretboard weight = 55
Top weight = 35
Top stiffness = 30
Back stiffness = 10

The actual effect on the system works more like adding up the squares.
So say 90% of the sound = 5625 FS + 3025 FW + 1225 TW + 900 TS + 100 BS. (75x75 =5625)

In my hypothetical example which ignores the effect of tuning the air resonance - the contribution of the fretboard is like 72% of the sound. You can make adjustments all day long to everything else and the effect is noticeable but minor. (A great example is removing the entire top doesn't change the sound all that much....)

To me - the biggest "aha" moment is that the "top" includes the fretboard - which acts as the stiffest and heaviest brace.... And thus has a very large contribution. It effects the air and top resonances as well as the stiffness of the entire structure. This is very different from a guitar where the fretboard is more or less inert.

Thanks
Truckjohn
Junior Mbr (0-50 posts)
 
Posts: 15
Joined: Sun Dec 27, 2015 5:03 pm

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Fri Jan 08, 2016 9:50 am

Yes, changes are not linearly additive and we'll probably never precisely know the effect of multiple combinations of changes. That's a good thing, otherwise we all might as well go home now.

But the elephant in the room which I, and no-one else here has raised, but which has niggled at me since doing the first experiments in cutting the top off, is this:

With no top a dulcimer has no air resonances, and no interaction of those resonances with the wood, and no Helmholz sound radiation from the soundholes. But if you compare the bridge tap spectra of a normal dulcimer (wood vibrations) with the spectrum of air resonances excited by a small loudspeaker (air vibrations), nearly every peak in the tap spectrum has a corresponding peak in the air resonance spectrum. This indicates to me that air resonances play a central role in the sound of a mountain dulcimer, and there are wood and air cavity interactions over a a wide frequency range. If this is the case, a topless dulcimer must be producing sound differently than an enclosed instrument. The vibration modes (Chladni patterns) are certainly different from an almost identical dulcimer made about the same time. And yet here is a recording of those two dulcimers, one with a top and one without, that I made just now at the computer - can you tell which is which?

Two Dulcimers - One With No Top.mp3
[ 480.53 KiB | Viewed 7057 times ]

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby Jono » Fri Jan 08, 2016 12:23 pm

I'm not a builder, I'm just a player, but the question must be asked----does all this experimenting and technical analysis end up producing a superior sounding dulcimer?
Jono
Junior Mbr (0-50 posts)
 
Posts: 16
Joined: Sat Dec 05, 2015 8:25 pm

Re: An Interesting Dulcimer Experiment

Postby Truckjohn » Fri Jan 08, 2016 2:22 pm

But it DOES have a top - just a much narrower one..... Aka the fretboard.

What you have done is threefold:
make the sound hole area very large.
Reduce the weight of the top significantly but not the stiffness very much
Reduce the area of the top by somewhere in the in the viscinity of 66%

So..... Making the sound hole area very large raises the Helmholtz because the air inside can move much faster. It also reduces the "spring" effect that effectively stiffens your enclosure against vibration. This tends to speed up the attack a lot and can also make things louder.

Reducing the weight of the top significantly but not the stiffness much tends to "brighten" things a lot without adding much bass. The reduction in weight will increase the volume significantly over the same structure just heavier.

Reducing the top area significantly should reduce the overall radiation of sound - but it's going to be balanced out by the much faster attack and the significantly lighter and responsive "top" that remains.

Thus there is a "change" but it's not nearly as large as you might think it would be.

Thanks
Truckjohn
Junior Mbr (0-50 posts)
 
Posts: 15
Joined: Sun Dec 27, 2015 5:03 pm

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Sat Jan 09, 2016 7:41 am

Jono

It's a reasonable question, but unfortunately there's no reasonable answer.

For me personally, the answer is no, experimenting doesn't lead to consistently superior instruments. By now I have a fair idea of what I like in a mountain dulcimer, and I think I can recognise a superior one when I hear one. But I never know until I string up a new instrument what it will sound like, except in a very general way. And I know that if an instrument turns out to have that extra "something" about the sound, it will not clearly show up in the crude measurements that I can make (loudness, attack, sustain, spectral content etc).

For the rest of the world, I have no idea what any one individual might consider a superior instrument - no-one does. That's one of the problems with mountain dulcimer comparisons - there is no concensus about good, bad or excellent. So making measurements and doing experiments can't help a lot when there's no target to aim for - other than the most general of descriptors such as bright or mellow - and even those very general terms are unreliable, one persons' "bright" can be another persons' "mellow".

So the experments I do can't provide a recipe for making superior instruments. But by finding out something about how a mountain dulcimer produces sound, the scene is better set for makers to concentrate more on matters that might affect the sound and worry less about things that have been shown to affect the sound less. For example, before I started experimenting, I didn't know that in one vibration mode dulcimers vibrate like a xylophone bar - but they do, and it informs my thinking about where to put the three feet on the bottom so I don't damp out that vibration when played on a table. Also I no longer stress out about top bracing, string break angle, top wood species, hollow or arched fretboard types, sound ports and sound posts etc etc. I know the answer to these things now, to my satisfaction, and unfortunatelty didn't before because of the variety of different answers from a variety of luthiers.

In a general sense, there's no right answers, and no wrong answers, just different answers. Same for dulcimer sounds, no better or worse, just different. You'll like one different sound better, and I'll like a different different sound better.

But the main reason to do these experiments is because if I wasn't doing them, I'd have to be mowing the grass.


Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby KenH » Sat Jan 09, 2016 9:14 am

Jono;

Richard's experiments, no matter how he downplays their importance, are the ONLY scientific, quantitative, information-gathering ever undertaken on the immensly complicated subjects of How A Dulcimer Produces Sound, and What Effects That Sound (and to what degree). The guitar, by comparison has something like 500 years of quantitative as well as quantitative information which builders can apply

Since Richard began his experiments back in 2008, various builders around the world have been reading of his verifiable results (not just qualitative impressions) and applying some or all of that information to their own dulcimer building. So, it can be said that YES, Richard's experiments are leading to a better (for values of better) instrument by each dulcimer builder who chooses to apply those lessons.

There may never be a scientific description of what makes a "mellow" vesus "bright" dulcimer, but if there ever is, it will only come from people like Richard who are willing, and able, to look at the thousands of variables and classify their interactions.

Anyone can take a 2x4 and some tools and make a dulcimer that is playable. But only a handful will be able to make a superior instrument. ANd those are the rare few who seriously look at the work of others and apply the knowledge gained to their own building efforts.
User avatar
KenH
Dulcified! (>2000 posts)
 
Posts: 10148
Joined: Fri Mar 15, 2002 7:16 pm
Location: Afloat in Fort Myers, FL

Re: An Interesting Dulcimer Experiment

Postby Robin the Busker » Mon Jan 11, 2016 6:47 pm

Hi Richard,

I've just had the chance to read through all your recent posts and sit down with a cup of tea to write a reply. Thanks for doing all the work you do - it is so enlightening and moves a lot of ideas from concept to practical experimentation.

The results of the dulcimer with three wooden feet experiment does not surprise me. I was playing a 3 wooden footed dulcimer on a solid wood bar room table at a session last week - the reason being that the bar table made the dulcimer LOUDER than when sitting it on my lap. I know the table was 'in play' because if I put felt pads under the feet to avoid scratching the table the dulcimer is muted somewhat - so, it is not just the back of the dulcimer vibrating freely (rather than being damped by my lap) that is creating the additional volume, it is the table (and floor!) vibrating and becoming a part of the instrument. And this is where many possum boards fall short. A lot of them have felt pads to protect the dulcimer and so this type of modern board only frees the back from leg damping, causing a tonal change and perhaps some volume increase. However, the old school method of making 3 wooden feet for a dulcimer (Thomas, Prichard, Mawhee et al) so it can be table played does have a considerable volume boost potential. Not a new idea even in the 19th Century as the American dulcimer's European ancestors were table played with direct contact feet to make them loud enough for dancing.

We have discussed before how this factor of direct vibration transfer around the dulcimer (and outside of it in this case) has a significant effect on tone and volume. As much of an effect perhaps as the sound box itself (judging by your dulcimer with and without top experiment). I've never been convinced that the Helmholz effect has any part to play in dulcimer making - in fact a tonally specific resonance (sound holes matching the internal air volume) is the last thing I want in a dulcimer!!!! I have had a couple of instruments where specific frequencies have been 'resonant' and they are a PITAS to play - I've ended up changing the string gauges and pitch and taping up sound holes.

Your observation about piezo pick-ups is something I've played around with. I spent a long time a couple of years back trying various manufactures for pick-ups for my resonator guitar business. The 'problem' with them all was the lack of consistency. Each said they had a good system but there were too many variables (position, adhesive compound, pressure) for me to recommend any to my customers. During my experiments I noticed how pressure in particular affected performance - which is why undersaddle piezo systems produce more consistently good results. For a quick fix for dulcimer I use the Cherub Guitar p/u WCP 60-G. They are less than $10, you can quickly clip them into a sound hole and you can easily wrap an elastic band around the clip to adjust the down-pressure as much or as little as you want 8) But don't tell anyone or all those expensive specialist stick on piezo manufacturers will be after me!

Regarding your work enabling the building of better instruments - well that's a no-brainer - the answer is 'of course'. Although I'm not a dulcimer maker, your work published in this thread has been really useful for me in analysing old instruments and also for the design work I did to help Chas Hagen produce his Heritage model for my shop. For the Heritage we took the design features I liked from a number of old dulcimers and built a new instrument for old style playing. Because of the work you've published Richard I could analyse why I liked certain features and decide on the ingredients - then Chas made the instrument pragmatic, repeatable for a production run and I must say beautiful!

http://www.dulcimers.co.uk/14.html

Robin
User avatar
Robin the Busker
Super Mbr (501-2000 posts)
 
Posts: 1768
Joined: Fri Apr 17, 2009 5:03 pm
Location: Snowdonia, Wales

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Tue Jan 12, 2016 9:18 am

Robin, well I'm pleased someone is making use of some of this, I'll have to go back and re-read it myself.

But regarding Helmholz resonances. We shouldn't get too carried away with it, but it is a natural part of any rigid enclosed structure with a hole in it. The Helmholz resonance is the lowest resonance of a mountain dulcimer (and a guitar) and interacts with the dulcimer wood to make it vibrate. It's easy to see its effect by strumming and covering/uncovering one or more sound holes with some cardboard. There's likely to be a modest tonal change - maybe for the better, maybe not. It's just one of the ten to fifteen resonances of the instrument below about 1000Hz, and it does affect the tone. It's not even technically the Helmholz resonance because the body of a dulcimer is not perfectly rigid - the frequencey is a bit lower than the true Helmholz, but that's neither here nor there. I'd rather call it the 1st Air Resonance.

What is not sensible is the notion of "tuning" the dulcimer to the Helmholz resonance and in the process making the sound better overall. I'm not sure where this notion arose. Positioning one of many resonances at a particular frequency might be beneficial in some cases, but moving one resonance is not going to "open up" an instrument. It's more likely the density of resonances, their even spread over the spectrum, and maybe the frequency ratios of some of them. All largely outside the direct control of the maker.

If I had my way I'd generally like to the 1st Air Resonance fall at about 170 to 180Hz for a standard sized dulcimer. That should get it well below the first bar resonance to smooth the bass (not boomy with superimposed resonances), and closer to the fundamental of the lowest D note. But because of the size and shape of my dulcimers, it generally falls about 220 - 230Hz, so I lose the fundamental harmonic of the bass string, and the tone is a little less mellow. Thin plates, large box and small sound holes will lower the Helmholz resonance, (if that's what you want to do).

The Helmholz is just one of many resonances - it plays its part but doesn't dominate all the others.

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby Robin the Busker » Tue Jan 12, 2016 12:12 pm

Thanks for the explanation Richard 8)
User avatar
Robin the Busker
Super Mbr (501-2000 posts)
 
Posts: 1768
Joined: Fri Apr 17, 2009 5:03 pm
Location: Snowdonia, Wales

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Sun Jan 17, 2016 8:52 am

But it DOES have a top - just a much narrower one..... Aka the fretboard.


TJ, I suspect you are technically correct in your comments, but in a practical sense, I wonder how much air the fretboard on its own can push around. A guitar neck is also flaiilng around, but I can't recall hearing that it contributes much to the total sound.
mode4.gif
(Click to see animation)
The topless dulcimer is just as loud as when it had a top, in terms of sound pressure level. I had assumed this was because the back was then the new top - it does vibrate strongly. I had also assumed that all the air resonances were lost, but maybe you are right, maybe some have just moved up the spectrum.
I did some quick tests with a small loudspeaker inside the dulcimer body near the nut, and recorded the microphone response inside the body along the length of the instrument. It's not a full mapping of any standing waves, but the frequency spectra give an idea if there's any resonant air activity still going on. The speaker swept in frequency from 100Hz to 2000Hz.
Air Rersonances_No Top Dulcimer.jpg

The bottom left is the response outside the dulcimer - it's surprisingly flat given the cheapo mic and the 1" speaker. The bottom right is another dulcimer of the same pattern, with top intact - speaker at the upper sound hole, mic at a lower sound hole. It's typical of the air rersponses I normally see. The two main peaks are the first and second air resonances - about 230Hz and 360Hz.

The other panels represent what's going on in the body of the dulcimer with no top, at different distances from the nut. And there's something going on - the spectra were quite repeatable. Between 1000 and 2000Hz there's clearly some variation in SPL at different frequencies. How much of that gets radiated out, I can't say, but it indicates that all air resonance activity is not lost and may contribute to the sound after all.

Richard
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Tue Feb 16, 2016 9:15 am

Mountain Dulcimer – Back Bracing and Side Linings

In previous posts I’ve looked at the effect of top side linings and also the effect of adding and removing back braces on my Orthey replica dulcimer.

Recently I acquired some wood that was resawn into a number of homogeneous-looking plates, and since I had to make few instruments for a festival I took the opportunity to experiment with the construction and see what resulted. As is usual, nothing spectacular.

I normally install both back braces and internal side linings in my dulcimers. Being an enthusiastic amateur I can afford the time and effort to do it. But for those makers who do it for a living the extra day or two per instrument to put these components in decreases their throughput and is a commercial reason to leave them out.

So does it make any practical difference to the sound – in or out? The short answer is – maybe, maybe not. The difference in the final sound of the three dulcimers in this experiment can’t really be pinned down to the presence or absence of braces or side linings. The differences might be because of those, but is just as likely to be a result of the natural sound differences produced by wood sample variation between the instruments (not wood “species” variation). So if put braces and linings in or not you are probably safe to assume it won’t be a sound disaster either way because of them.


The Experiment

In this experiment I made three “identical” dulcimers with back and sides of an oak-like wood called Dillenia Papuana (<5% weight variation between the three back/sides sets) with Sitka Spruce tops and Dillenia fretboards and end blocks(<2% top-assembly weight and stiffness variation). Final weights varied by less than 5% and most of that was the presence of linings and braces in one instrument. Final weight was ~2.4lb(1100gm)

One instrument (#93) had no back braces and no side linings; one (#94) had side linings but no back braces; and one (#95) had both side linings and back braces. All had a centre join support strip of Dillenia.
Dulcimers_93_94_95_Internal_External.jpg

All three had the same top configuration: hollow fretboard; five cross braces. There was a conscious attempt to make these dulcimers sound more on the mellow side than the bright side. The top was thinned to 2mm (a little thinner than normal for me); total sound hole area was smaller than usual (for me) Both of these factors might tend towards more mellow. But due to circumstances, other factors that might be important for mellow/bright were basically at the mid-point of the mellow/bright divide according to my previous study. Strum hollow height was 9mm; side height was 44mm; completed top deflection was 120/1000”; bridge position was 70mm from internal end block. So it could go either way.

For most tests I also tested my own dulcimer (#54)and my Orthey replica under the same conditions. I consider my own dulcimer a superior sounding instrument (to my mind) and it’s the one against which I compare all dulcimers I make. It’s fairly stiffly braced top and back, with side linings. The Orthey is a very nice sounding dulcimer which sounds better without back braces than with, and has no side linings.

Results of Testing and Listening


Perceived Sound Differences
These are my perceptions only – yours might be quite different. It’s a minefield.

#54 – Very bright sound with a “sting” to it, strong upper treble cutting power and lower treble “punch”; solid bass but not exceptionally resonant. Good dynamic range. Middle string a bit reduced in clarity.

Orthey relica – A loud instrument, and with good clarity on all strings over the whole fretboard. Slightly reduced dynamic range and punch compared to #54. Nice resonant bass.

#93 (No linings or back braces) – most muted treble of the three test dulcimers; slightly muffled middle string but nice rounded bass. Reduced “punch” and attack. The least preferred of the three (by me).

#94 (Linings but no back braces) – Characterised by “clarity” on all strings. Reasonably “punchy” but reduced cutting “ring” on upper treble. Marginally the most preferred of the three test instruments.

#95 (Linings and back braces) – Seems loudest, and punchy. Good cutting power on high treble. Slightly muffled middle string. Best dynamic range.

All three test dulcimers seemed to have the same sensitive response to soft playing, none were overly mellow or bright - about middle of the road for me.

Measurements

Sound Level Output
Recordings were made of single strings using the copper thread pull method for repeatability. (look back at the second String Break Angle tests for the method). Three tests per string; nine string plucks per dulcimer. The first three seconds of each pluck was measured for average sound intensity using the PRAAT speech analysis software.

Results were:
Copper Thread Pull.jpg

Despite my perception that #95 was the loudest of the three, it was technically the quietest, so other factors (attack, frequency bias) are colouring the perception of loudness. The Orthey did clearly measure louder than the others but only by about 3dB which is not that noticeable. Overall there was not much between them in terms of absolute sound pressure levels. (The numbers themselves don’t mean much but the test conditions were constant for all dulcimers so relative comparisons should be OK).
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 602
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

PreviousNext

Return to Making Dulcimers

Who is online

Users browsing this forum: KenH and 4 guests