An Interesting Dulcimer Experiment

Just share stories or offer advice

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Tue Feb 16, 2016 9:22 am

Mountain Dulcimer – Back Bracing and Side Linings (Continued)

Sound Spectrograms

The copper thread single string plucks allowed reasonably comparable sound spectrograms to be made for each dulcimer. These show the number, strength and duration of the partials of the plucked notes and can give an idea of where sound energy might be concentrated or missing in the spectrum. The Owners Manual that comes with my PreSonus AudioBox USB microphone input device has some suggestions as to the effect of different parts of the sound spectrum on perception. I don’t know myself, but they seem to have reasonable experience in sound equalisation. Paraphrasing their suggestions (for acoustic guitar-like instruments):

Ø Strong overtones in the 250 – 500Hz region can add ambience and clarity to the sound, and maybe fatten it up.
Ø Too much energy in the 500 – 2000Hz region can sound thin or “tinny”
Ø Energy in the 2kHz – 4kHz region assists with “projection”
Ø Energy in the 4kHz – 6kHz region provides clarity and “presence”
Ø Energy above 6kHz improves “brilliance”

All the dulcimers tested have sound energy extending up to at least 10kHz. Here is a composite spectrogram for each string of each dulcimer.
Spectrogram_All Strings _Copper Thread Pull_93-94-95-54-Orthey.jpg

Top panel is 1st string, middle is middle string and bottom is bass string. There are differences in the spectrograms, and hidden in there must be the reasons for the differences in the perception of the different sounds of the test dulcimers. I leave it to you to scrutinise and conclude. For those who might like to see more detail I’ll insert the same three spectrograms but expanded in height.
Spectrogram_1st String _Copper Thread Pull_93-94-95-54-Orthey.jpg
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
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 9:25 am

Mountain Dulcimer – Back Bracing and Side Linings (Continued)
Spectrogram_2nd String _Copper Thread Pull_93-94-95-54-Orthey.jpg

Spectrogram_3rd String _Copper Thread Pull_93-94-95-54-Orthey.jpg

I think most people over 40 years of age won’t hear much of what is going on above about 5kHz.. It might partially explain some of the difference in assessing the sound of instruments.


Conclusion

The three dulcimers tested here were as identical as was reasonable without getting obsessive, except for the presence or absence of internal side linings and back braces.

They all sounded good (to me) but the one without braces or linings was my least favourite. However, the variation in sound between them is no more than I would expect if they were all constructed identically with or without braces and linings. So it isn’t obvious that the presence or absence of these components affects the sound significantly – the loudness, sustain or general spectral profile.. You choose.


Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Wed Aug 17, 2016 10:27 am

Mountain Dulcimer Acoustics – Effect of Opening a Hollow Fretboard Channel by Removing the Top Plate Wood Beneath the Channel.

I’m making two baritone dulcimers at the moment, tuned AEA, and would like them to be as tonally robust as they can be.

A problem arises in that the low A-note is 110Hz, the middle E-note is 164Hz, and the high A-note is 220Hz.

A typically normal sized dulcimer can’t produce sound much below 200Hz, or at least it does so very poorly. This means that the fundamental harmonic of the low and middle strings won’t contribute to the sound, and the fundamental of the melody string (A220) will only just start to contribute. We then rely on the higher harmonic series to carry all the tonal information, without the fundamentals, which might well be perfectly OK and sound good, but we’ve lost part sound.

It would be good to include the contribution of the string fundamentals to the sound if we can, so how might we do it? The problem is to get the dulcimer box to vibrate at a frequency down to 110Hz, or near it. Keeping in mind that the Helmholz resonance, the lowest air resonance of the box, the “rum jug” tone, is also the lowest resonance of the whole instrument, then three ways to do it spring to mind.

1. Make the dulcimer box very large - as the box size increases, the Helmholz frequency decreases. I’m not sure how large a box would be necessary, but I can’t do it anyway because the wood I am using is already cut to standard dulcimer size.

2. Reduce the size of the sound holes – the smaller the sound holes, the lower the Helmholz frequency. But even if the holes were really small, I know from experience that it would not get the Helmholz frequency anywhere near 100Hz. In any case, I'm using wooden rosettes, so the sound hole size is fixed.

3. Make the dulcimer box, and specifically the top plate/fretboard assembly very flexible. The Helmholz frequency in wooden musical instruments is not strictly the correct term to use because it refers to a totally rigid body. As the box gets more flexible, the “Helmholz:” frequency falls, and is better called the First Air Resonance. But a flexible enough dulcimer box to reduce the first air resonance to 110Hz would probably need very thin plates and no internal bracing – easily damaged.

So basically I’m stuffed – a standard sized dulcimer, with normal sized sound holes, and normally robust construction won’t reproduce the fundamentals of the strings if baritone-tuned. I’ll just have to live with whatever nature delivers, as Blue Lion do.

However, I was interested to see what a token gesture in the flexibility direction might achieve – specifically whether opening the hollow channel of the fretboard to the inside of the dulcimer would have any effect on the flexibility and vibrational behaviour of the top/fretboard assembly – before it's mounted on the sides. It isn’t known for mountain dulcimers how the resonances of the free top/fretboard might map onto the assembled instrument, but I suppose it’s reasonable to expect that generally lower resonances in the free top will translate to lower resonances in the assembled instrument.

Tests
For the two dulcimer top/fretboards in question I measured
- static deflection under 7.6kg (16.7lb) weight – a measure of flexibility
- spectral analysis of the tap sound using a rubber hammer (tapping the bridge area)
- and weight
before and after cutting out the top plate wood below the fretboard channels – as in the picture (click for animation):
FretboardChannel_Open_Closed.gif

Results
As is usual with nearly all my experiments results were not quite what I expected. I expected a modest reduction in resonant frequencies of the top/fretboard, and reduced stiffness (increased flexibility).
Top-Fretboard_Resonances-Open_Closed Channel.jpg

But for these two dulcimers, opening the fretboard channel had almost no effect on the stiffness of the dulcimer top, nor on the way it likes to vibrate. Changes in stiffness, resonant frequencies and weight were less than 1% or 2%.

Conclusion

The removal of the top-plate material below the channel of a hollow fretboard probably has very little effect on the tone of a mountain dulcimer. It doesn’t seem to affect top-assembly stiffness very much, or change the resonant behaviour of the free top. Weight is only reduced marginally. Open or closed – it seems OK either way.

There is a slight increase in box capacity with the open channel which does lower the first air resonance – but only by about 5Hz; not enough to notice an effect.

This all might not be very surprising, at least in my designs, because the fretboard channel only extends up to the strum-hollow area, leaving most of the lower bout unchanged with or without the fretboard opening. And most of the sound of the top of a mountain dulcimer comes from the lower bout.

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby kwl » Wed Aug 17, 2016 7:59 pm

Thanks for sharing this, Richard. Your tests confirm what I have maintained for years.

Ken
"The dulcimer sings a sweet song."
User avatar
kwl
Dulcified! (>2000 posts)
 
Posts: 17430
Joined: Mon Mar 18, 2002 5:25 pm
Location: Reedsville, PA

Re: An Interesting Dulcimer Experiment

Postby KenH » Thu Aug 18, 2016 7:05 am

Once again, a great experiment and results, Richard.

I have one question though. I see that in your design you have what I would call and overabundance of transverse braces in the top. I, and many other builders I know, don't brace the top transversely at all, only using the fretboard as an external longitudinal brace. I've wondering if -- in such a case -- if the opening of the top would give more than just the 1%-2% increase you recorded?
User avatar
KenH
Dulcified! (>2000 posts)
 
Posts: 10216
Joined: Fri Mar 15, 2002 7:16 pm
Location: Afloat in Fort Myers, FL

Re: An Interesting Dulcimer Experiment

Postby Robin the Busker » Fri Aug 19, 2016 9:33 am

Another great piece of work Richard 8),

Homer Ledford channelled his fretboards but did not slot the dulcimer top boards. Some of the early Kentucky makers (Thomas, Amburgey, Tignor) used solid fretboards, so Homer was to some extent breaking from his local predecessors. I think I read that Homer felt routing a channel in the fretboard didn't do anything for the tone but made the fretboard stiffer. Personally, I think that adding a channel does make a tonal difference. But I think that possibly the difference comes from enabling a better vibration transfer to the top (and rest of the instrument) rather than because of the added air chamber volume. This is because it is easier to get a good solid joint between the two thinner 'legs' of a channelled fretboard and the top than it is between the wide flat bottom of a solid fretboard and the top. That narrower contact down the sides of the fretboard will transfer vibration more effectively that the slab bottom of an non channelled fretboard.

I get Chas Hagen to rout out the fretboards for my Red Kite and Heritage models and he has noticed at the workshop the difference in tone this simple procedure has made compared to keeping the fretboard solid. We haven't gone for slotted top boards because we couldn't notice an advantage.

Robin
User avatar
Robin the Busker
Super Mbr (501-2000 posts)
 
Posts: 1763
Joined: Fri Apr 17, 2009 5:03 pm
Location: Snowdonia, Wales

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Fri Aug 19, 2016 11:06 am

Ken

Ken, the bracing on these two tops looks worse than it really is. The braces are ony 3mm high and are quite flexible - so they are low and wide rather than higher and narrow as I more usually do. It was another attempt to make the top assembly a little more flexible and lower the resonant frequencies, but still keeping it structurally stong across the grain, which I like to do. As it happens the resonances didn't really turn out any lower than for other tops with stiffer bracing. I haven't tested opening the fretboard channel on a top without braces, but I have measured my complete test dulcimer(which is a better test) with and without top and back cross braces. There was about a semi-tone reduction in frequency of the first two or three resonances when going from top and back braces to back-only bracing, and another semi-tone reduction when removing the back braces (no top or back braces). But for those three conditions, full; half; and no bracing, on the same dulcimer, I couldn't really tell much change in the sound when listening to it. It might well be because even though the first air resonance fell from about 200Hz (full braces) to 165Hz (no braces), that's still above the D147 or C130 for the bass string of a dulcimer, and so the fundamental part of the sound would still be missing. The presence or absence of top/back bracing alone doesn't contribute enough to get the first air resonance low enough to cover the bass string fundamental. A combination of no bracing, small sound holes and a large body might do the trick, but I haven't seen any reasonably sized dulcimer that does have a low enough first air resonance. I might make myself a Tennessee Music Box and see what that looks like - they look un-reasonably large to me and might represent the upper extreme in size, and hence a lower limit of box resonances.

Robin, I think Homer might have been mistaken in thinking a hollow fretboard might be stiffer than a solid one - in these two dulcimers it was slightly less stiff, about 2%, and that's what would be expected, but mass was reduced by a larger amount, maybe 30%. I think that does make a change in tone. I'm not a gluing expert, but I don't feel that contact between the top and fretboard is improved because of the smaller gluing area of a hollowed fretboard. A solid fretboard could be glued just as firmly - it depends on how well the joint is prepared and clamped. It's not really hard to make a good glue joint. If there is a change of tone attributable to the contact area between to fretboard and the top plate (how could you measure it?), I'd think it would be more likely caused by changes in energy transfer rates through smaller areas or something like that, rather than the integrity of the glue joint. If a fretboard was very light, with maybe a thin overlay of harder wood, I might leave it solid, but in general I wouldn't.

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby Daniel » Fri Aug 19, 2016 11:33 am

.
Last edited by Daniel on Fri Mar 23, 2018 10:48 pm, edited 1 time in total.
Daniel
 

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Sat Aug 20, 2016 9:24 am

Daniel, feel free to ask questions about any of this stuff I have to think about it when I re-read it myself. The one thing none of it will tell you is "the best way to ....." or "Do this and the dulcimer will be a good one". That's always up to you. Some of it might hint at what can make a bad dulcimer, and a lot of it will tell you that it doesn't really matter and you can then concentrate on something that does.

Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
Joined: Tue Jun 05, 2007 8:18 am
Location: Brogo near Bega, NSW, Australia

Re: An Interesting Dulcimer Experiment

Postby Jon » Sun Aug 21, 2016 4:23 pm

Richard, your conclusion is helpful for a different reason. If the top plate under the hollow fretboard is not required to be solid, then there is no need to accurately book match the two sides of the top plate.
Jon
Senior Mbr (101-500 posts)
 
Posts: 252
Joined: Mon Nov 09, 2009 4:32 pm

Re: An Interesting Dulcimer Experiment

Postby Robin the Busker » Sun Aug 21, 2016 6:20 pm

Robin, I think Homer might have been mistaken in thinking a hollow fretboard might be stiffer than a solid one - in these two dulcimers it was slightly less stiff, about 2%, and that's what would be expected, but mass was reduced by a larger amount, maybe 30%. I think that does make a change in tone. I'm not a gluing expert, but I don't feel that contact between the top and fretboard is improved because of the smaller gluing area of a hollowed fretboard. A solid fretboard could be glued just as firmly - it depends on how well the joint is prepared and clamped. It's not really hard to make a good glue joint. If there is a change of tone attributable to the contact area between to fretboard and the top plate (how could you measure it?), I'd think it would be more likely caused by changes in energy transfer rates through smaller areas or something like that, rather than the integrity of the glue joint. If a fretboard was very light, with maybe a thin overlay of harder wood, I might leave it solid, but in general I wouldn't.


Hi Richard - I don't think I explained that Homer Ledford example very well. He wasn't talking about hollowing out the fretboard to make the top stiffer but felt that a hollowed out fretboard was 'stiffer' in the sense it was less prone to warping than a solid fretboard. I don't know if this is the case?

I think you are right about the reduced contact area of a routed fretboard (rather than easier gluing) attributing to the tone - and the reduced mass must have some effect. Overall, a stiff, light build does seem to perform better than a heavier build. There's certainly something good going on when you rout out the underside of the fretboard. I'm reminded of the times that a loose brace has killed the tone and volume on some of the old dulcimers I've had through my hands and how a tiny drop of superglue has bought the instrument back to life! Which is why I was wondering about the better contact possible from a routed fretboard than a solid one?

Wow, all this information. I'm new here, so I've now got about 10 years worth of technical reading ahead of me.


Daniel - Richard's right in that it doesn't tell you here how to build a good dulcimer, just how not to make a bad one!!! The instrument itself has fundamental problems for contemporary (DAd chord melody) playing, because essentially it was never originally designed for that job so there's an element of compromise - so feel free to think outside the box (literally) if that's the playing style you are building for. However, for traditional playing styles the instrument in its classical form does a great job. I think the Glen dulcimers are about as good as you'll get as a template for a general traditional workhorse instrument (I regularly use a couple of old Glen dulcimers for gigs) because they are loud and have great melody string clarity against the drones. Their features are - long VSL - hollow fretboard - shallow, wide body, 3 string, wood pegs (light headstock), light overall build with very limited bracing, one piece back and two piece top gapped under the hollow fretboard (like a McSpadden), bridge near the tail of the instrument over the tail block.



There is no 'perfect' design because each instrument will have a voicing that will work for some things but not others. As a rule of thumb however a light build does seem to generally produce more resonance - so a channelled fretboard is generally going to produce more resonance than a solid one. A channelled and scalloped fretboard may work even better. Possibly the loudest 'standard' (in terms of shape and size) dulcimer I own is an old Orthey. It is actually much shallower than a standard McSpadden but a much lighter build with a channelled and scalloped fretboard - despite its smaller internal air chamber it is a deeper toned and more resonant instrument as well as being much louder. Here's a very unscientific side by side recording of a McSpadden and Orthey:


McSpadden v Orthey.mp3
[ 1.02 MiB | Viewed 3408 times ]


The Orthey I feel points the way towards making a nice responsive instrument for DAd chord melody playing - you don't have to have a big, deep body to get a big tone - a very light but shallow build seems to do the job. Interestingly, the bridge is right near at the tail of the instrument over the end block. My personal, and very unscientific, impression from the instruments I have played is that this bridge position brings out both the highs and the lows more than when the bridge is placed further in-board where more of a mid-range 'honk' is emphasised.

Dulcimers are very much 'horses for courses' so you do need to experiment to find what's going to suit your playing style. Even really simple changes like fitting heavier or lighter strings or tuning up or down a tone or raising/lowering the action can have as big an effect on tone and volume as a more major design change. Personally, I run heavier strings and a higher action than most players (often pitched up higher too) - because that's how old dulcimers were set-up - and just that simple change dramatically improves the volume and voicing on old dulcimers. The bottom line is that to get the most out of our wonderful instruments you've got to put some energy into the box in the first place (or use a pick-up!!!!).

Robin
User avatar
Robin the Busker
Super Mbr (501-2000 posts)
 
Posts: 1763
Joined: Fri Apr 17, 2009 5:03 pm
Location: Snowdonia, Wales

Re: An Interesting Dulcimer Experiment

Postby rtroughear » Mon Aug 22, 2016 1:04 pm

Jon, I don't spend a lot of time making a perfect top-plate joint because I know it will either be cut out, or hidden. Although it's not that hard to make a quite good match just by sanding the edges of the two plates on a flat surface - the Titebond glue then does a good job of pulling the edges together as the moisture dries out..

Robin, I think a lightweight approach is generally a good one too, but I just built myself a Tennessee Music Box, and it's hardly in that category. It weighs three times as much as my standard dulcimer. In fact I might put a handle on it so I can carry the thing like a suitcase.
TMB_01.jpg
And it has a solid fretboard (but of lowish density). So there are circumstances where norms can be transgressed and the instrument still work. This TMB sounds really nice to me, it has a smooth resonant bass, with sweet treble and mid-range, and moderately loud. Don't quote me, but I have a feeling in my big toe that "lightweight" is less critical to the back and sides of a mountain dulcimer than to the top/fretboard assembly. And also that the width of the top plate in the lower bout contributes more than the height of the sides - something I haven't tested because I use a standard template for my own dulcimers.

The Orthey design would be a very good one for a new maker to start with (with appropriate acknowledgement of the design source). It should be relatively easy to make with limited machine tools. I think the secret to it is the hollow/scalloped fretboard, and that could be made up of three pieces of wood.
Orthey Replilca.jpg


Richard T
rtroughear
Super Mbr (501-2000 posts)
 
Posts: 612
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: No registered users and 1 guest