Nasal Sounding 'inboard' Reeds

[adrian brown]

The 'inboard' mounted reeds on the Jeffries 38 system (and presumably the larger Anglos and duets too) have a slightly different tone to those mounted at the side of the reed pan. This problem is most acute on the left side thumb button reeds, where the tone is distinctly nasal in comparison to their 'outboard'cousins. Because of the position of this button's reed chamber, the air hole is drilled through both the pad board and the action plate and is thus twice the depth of all the other holes. In a sudden fit of activity last week, I thought I'd check what a reduction of the reed chamber size might do to the tone, since I'd seen a kind of 'double wall' on the respective chamber of another instrument, and also noticed that the air hole appears to be drilled more towards a position above the middle of the reed than on other notes.
Anyway, I made some experiments with modeling clay and found enough of an improvement to be worth mentioning here. The best position seemed to fill the chamber up to, and over the clamping bar of the reed as shown in this picture.

I have now replaced my son's modeling clay with a piece of blu-tac which I feel will less likely to stain or damage anything. It should also harden in time and be easier to remove. I've made a couple of very unscientific recordings both with and without the modification, which nevertheless seem to bear out my audible impressions. In both recordings, the thumb button is the middle note of each descending run of three notes and I've recorded both push and pull. Anybody else tried this, or found a different solution for this problem?

Adrian.

 

 

[Geoff Wooff]

Well, you certainly have made a difference Adrian. It would appear as if the chamber is too long for an accoustic balance with the other reeds due to the need for enough space to un-lock the inside reed from its tapered slot.

 

I remember finding old concertinas where a previous owner had 'packed' the chambers with Cork or tightly screwed up paper which I thought was possibly an attempt to increase the pressure in the chamber to either make the reed speak more quickly from start up or increase volume. This was in 'old' instruments that did not have well worked out reed chamber dimensions.

 

The Canted (or angled, tapered) Reedpan that one finds in the later EC's and Duets gives a good example of the lengths to which the makers went to provide an even tone and response throughout the range.

Edited February 17, 2013 by Geoff Wooff

[adrian brown]

Thanks Geoff - interesting to know this might have been done in the past. Here's a picture of a 'double wall' on one of those inboard reed chambers.

I'd assumed it was doubled only to shorten the 'normal' reed chamber below, but it appears the 'upper' wall is sitting on top of the screw holding down the reed!
As to why it should work, I would assume that the deeper chimney (to use a woodwind term) of this hole increases the chamber volume, which it perhaps doesn't need?
Thanks for not having a go at me vis a vis the blu-tac. I'm just taking steps to remedy that by resourcing some matière première into which I'll carve a more permanent solution tomorrow. Hic!

Adrian

[Geoff Wooff]

Alors Adrian,

vous avez trouvé bien le raison des murs double. Aprés un soir avec le medicament de bon qualité, coupé le bouchon et mettez en place pour une solution permanent, voila !

 

I feel sure this information will be helpfull to others, of a sensitive nature .

 

PS many many thanks Adrian, for help in that other matter which perplexes me.

 

Geoff.

Edited February 18, 2013 by Geoff Wooff

[Ann-p]

Geoff, Alors indeed!

[cplayer]

hello all

Interesting one this,has anyone ever checked the depth of the reed chambers?my jeffries is just over 3mm deep

having checked three other jeffries(more honkier than mine) they were 4mm deep,any conclusions?

[Geoff Wooff]

hello all

Interesting one this,has anyone ever checked the depth of the reed chambers?my jeffries is just over 3mm deep

having checked three other jeffries(more honkier than mine) they were 4mm deep,any conclusions?

3mm does not sound very much , are you sure about that ? I would have thought that 3mm would not give enough height for the valves to open sufficiently so as not to muffle the sound. The effect of this would be that the pull notes would be louder than the push notes.

 

Perhaps we are not talking about the same dimension.. so I would call the chamber depth that height from the face of the reedpan to the top of the chamber walls including the chamois gasket... thus the effective depth of the reed chamber.

 

The Chamber height of my 'honky' Wheatstone is 6.35mm ( a quarter of an inch) and another that is 18 years younger is the same.

Edited February 21, 2013 by Geoff Wooff

[Robin Harrison]

Very interesting, Adrian.

I have a 38 key C/G Jeffries whose thumb button sounds SO bad I try to avoid using it, which sometimes obviates the point of a 38 keyer.

The reeds sounded like poor quality accordion reeds. Really unpleasant and made the chords dreadful.

So naturally I've tried your fix and it has improved it tremedously ! It's not perfect and given its position in the red pan it is always going to sound different. However it is so much better.

Because this was so successful, I thought I'd try it on my 38 keyed Jeffries F/C old pitch (actually F#/C# at the moment). The thumb button sounds OK but I thought it could only improve it.

Have a look at the pic. A double walled chamber ! May be an old after market improvement as no chamois.

What's old is new again ?

Thanks a lot for the sharing information...........I've never really liked the C/G partly because of the aweful thumb button, this improves it substantially.

[adrian brown]

I'm really glad it worked for you too Robin, and gladder still that it's not my ears playing tricks with me... My experience is the same as yours - this is more of an issue on my CG than on my GD, but I've no idea why this should be. Have your instruments got the deep, undercut hole on the underside too?

I'm going to try to fill this in next and see what happens. I'll report back if there's any further improvement.

Adrian

[MarkvN]

Thought I might chime in... Correct me if I'm wrong, but what struck me as the most interesting bit of Adrians remarks, is the insight that some of the inboard mounted reeds might sound dull because the hole in the action board is double as deep (especially the thump button): there are two layers of wood on top of each other, pad board and action board.

 

I recall from the research by Ttonon (http://www.concertina.org/ica/index.php/pica/subject-index/38-articles/87-reed-cavity-design-and-resonance) that the depth of the pad hole, e.g. where most of the sound comes out of the chamber (Tonon calls it t, the aperture length or action board thickness), acts as an obstruction to the higher frequencies: so, the deeper it is, the more reduced the overtones. This might also explain why a C/G is more troubled by this than a G/D, because of its higher overall tuning?

 

Reducing the deepness of that sound hole - chipping away part of the wood of the inside 'roof' the chamber, or enlarging the hole surrounding the pad - might be the true solution (e.g. resolving it in physical-theoretical terms), although obviously that's not something you might want to do with an old instrument. The interesting thing in this case might be, that it is actually possible to calculate what this should do to the sound spectrum, which
seems to be a rare situation in the concertina building business...

 

By the way, I noticed that in many 40+ button Jeffries, the sound holes for the extra reeds are usually located not underneath the action plate, but next to it.

 

Mark

[cplayer]

no pulling the wool over your eyes Geoff.It should have been mine 6.3mm, other three, 8mm deep

sorry about that

cplayer

[Geoff Wooff]

no pulling the wool over your eyes Geoff.It should have been mine 6.3mm, other three, 8mm deep

sorry about that

cplayer

OK Cplayer,

so the deeper chamber would normally suggest to me mellower tones... as in the variable chamber height instruments of the later Duet and EC instruments from Lachenal and Wheatstone... but with these it is more a case of balancing out the power and tone so that chords would be well matched , with out some notes sticking out or being lost. However in an Anglo with its equal height chamber walls , perhaps there is a 'best height' that gives the most pleasing tone throughout the instrument range.

 

It is also possible that you instrument was made to be at a different pitch or was made by a different person(s)... a Jeffries by a different member of the family... a Crabb made for Jeffries.. not measuring like for like ( perhaps the right hand ends have a different height to the lefts.. etc etc.... ??

Edited February 22, 2013 by Geoff Wooff

[Robin Harrison]

Have your instruments got the deep, undercut hole on the underside too?

Yes...........they do..............but so do a number of them on the treble side too where it only going through the reed pan. I can't yet see a rational for it..................anyone ?

[adrian brown]

I recall from the research by Ttonon (http://www.concertina.org/ica/index.php/pica/subject-index/38-articles/87-reed-cavity-design-and-resonance) that the depth of the pad hole, e.g. where most of the sound comes out of the chamber (Tonon calls it t, the aperture length or action board thickness), acts as an obstruction to the higher frequencies: so, the deeper it is, the more reduced the overtones. This might also explain why a C/G is more troubled by this than a G/D, because of its higher overall tuning?

 

 

By the way, I noticed that in many 40+ button Jeffries, the sound holes for the extra reeds are usually located not underneath the action plate, but next to it.

 

Mark

 

This would presumably also explain why the higher of the two notes is also consistently the worst sounding - I thought it might be because these are push, not pull notes, but I've never swapped them around to check.

 

The 38 layout also has second 'inboard' reed on the left, which although different sounding from the 'normal, outboard' reeds, does not tend to have this nasal quality. I feel pretty sure if the action plate was removed around this hole, and the pad seated directly on the pad board, these notes would sound better, but there's not a lot of room there for the lever post. Like you say, it's not something I'd want to try on an old instrument.

 

 

 

Have your instruments got the deep, undercut hole on the underside too?

Yes...........they do..............but so do a number of them on the treble side too where it only going through the reed pan. I can't yet see a rational for it..................anyone ?

 

Mark had the idea that instruments with the rubber ink stamp 'R. Whitten' on their pad boards often seem to have this (random?) undercutting, which might indicate it was a later (dealer , or repair) modification.

My CG has it on both the left (6 holes in addition to the thumb button) and right side (3 holes), and some are undercut towards the centre of the pan - others towards the rim.

However all of the 38 Jeffries I've seen apart, have this undercutting on the thumb button hole and this may perhaps be more of an original feature? It would be interesting to hear what modern makers have to say about this, but with the exception of Jürgen Suttner, most of them seem to concentrate on 30 or 31 button instruments, and so I suppose they wouldn't necessarily have come across this? With the 31 button Jeffries layout, the thumb button has its reeds 'outboard', and so doesn't suffer from this problem.

 

Adrian

 

[ttonon]

Hi Adrian, in the .doc attachment, Figure 1 shows the raw data traces of the sound files you provided, but with all traces organized so that each “Without” trace is followed immediately by its corresponding “With” trace. I think you’d agree that the amplitude (volume) of the middle note in each case – the notes made by the cavity you experimented with – does not show appreciable alteration by the insertion of Blu-tac. In fact, the last note shows that the “With” sound has lower volume than the “Without” trace, but I don’t think it’s very significant. It’s thus interesting to me that, although you feel your modification improves the sound of the two notes (both push and pull), such improvement does not show up as an increased volume.

 

But your sound files I think do belie a significant change in sound due to the cavity modification. The .wav attachment is a sound file extracted unmodified from yours, containing just the two notes made by the cavity in question, with the “Without” trace followed immediately by the “With” trace, for both push and pull, for both playings of the notes. I think making both sounds immediately next to each other might make comparison easier, and to me, the second sound, which is from the “With” file, reliably has a more treble sound to it, a sound some might describe as “more bright.”

 

Taking the Fast Fourier Transform of these sounds gives you the frequency spectrums shown in Figures 2 and 3 of the .doc attachment, comparing the notes before and after cavity modification. In Figure 2, which is for the C#3 note, the x’s (with modification) show a spectrum that has reduced low-end partials (frequency range up to about 2,000 Hz), and increased upper partials, from about 2,500 to 6.000 Hz. The frequency range about 6,000 to 8,000 Hz shows an interesting reversal, but for the highest end of the spectrum, the modified chamber again shows larger amplitudes. For me, this comparison is adequate evidence for a more trebly sound, as my subjective hearing concludes. I have no explanation for the 6,000 to 8,000 Hz spectrum, but more on that later.

 

In Figure 3, which is for the G3 note, the x’s do not show much drop in overtone amplitude at the lowest end, but do show the increased amplitudes in the range 2,000 – 6,000 Hz shown by those for the C#3 note. The range 6,000 to 8,000 Hz shows the same counter trend. The highest end of this spectrum does not appear to show much difference between the modified and unmodified sounds. But I think the comparison in the region 2,500 to 6,000 Hz provides a reasonable verification of the more trebly sound that I hear for the sound with the modified cavity. Such is reasonable simply because of the SPL amplitudes themselves, which our hearing emphasizes more in the 1,000 Hz area, and are also vastly numerically larger than the amplitudes further up the spectrum. Remember, we are dealing with a log scale. Further, one’s hearing at the highest frequencies can often be diminishing, and so they contribute less to the sound, especially for one in my age category.

 

Okay, I’m convinced that your modification does change the sound of the note, and not really it’s playing amplitude. Providing a clear physical explanation for this is very difficult and beyond anything I can do here. It is true that your modification will most likely modify the resonance characteristics of the cavity/aperture system, and I would definitely expect it to change the amplitudes of at least some overtones, but an explaination why it would modify so many overtones would take a good deal of theoretical investigation, once having obtained accurate dimensions of the geometries involved.

 

Mark, thanks for bringing my attention to this thread. (For some reason, my notification feature was turned off.) Concerning your query about the effect of the aperture, the situation is a bit more complex. I myself, and probably professional acousticians, can apply only the simplest and basic concepts in trying to understand how the free reed makes the peculiar sounds it does, without a serious research effort. My approach is to apply the well-understood models of resonators: Helmholtz, quarter-wave, half-wave, etc. and try to draw conclusions. Thus, if you’re applying the Helmholtz resonator model, you would consider the aperture in conjunction with the cavity volume, and perhaps also the aperture created by the reed slot itself. If you invoke a quarter wave model, you look at key linear dimensions in the cavity and in the aperture, taken separately. For instance, the peculiar response enjoyed by the 6,000 to 8,000 Hz range in the above frequency spectrums indicates that the cavity volume did not affect whatever it is that is causing this. One possibility is perhaps that the aperture is behaving as a half-wave resonator for these overtones.

 

Note: I'm editing out an incorrect interpretation here. In this frequency range, it appears that the Blu-tac lump does effectively eliminate much of the enhanced overtones, for both notes. Perhaps there is some kind of half wave effect of the aperture coupled with the cavity volume that is eliminated when the cavity volume is changed.

 

Before even the simplest models can be applied, as described here, one would have to first make accurate measurements of the relevant geometries, then apply some clever detective work, starting with the simplest and building from there.

 

Best regards,

Tom

www.bluesbox.biz

Aybee, Blu-Tac Cure.doc

Aybee, thumb note only, without bt, then with.wav

Edited March 15, 2013 by ttonon

[MarkvN]

Hi Ttonon,

Thank you for the detailled comments.

The depth of the aperture in the action board (your aperture length, sorry, can't get used to view it as length) would actually be quite easy to vary in a physical model: one would only have to add a sort of temporal ringlets underneath the pads, e.g. around the sound holes on top of the action board, to make the holes 'deeper'.

 

I've never understood why the reeds in the middle of the reed pan of a 38-k Jeffries should sound any different from the ones that are lined up along the edges (and the difference is often really obvious). From your theoretical perspective - reed chambers as resonators - there seems to be only one detail that differs consistently for at least some of the inner reeds: the depth of the sound hole, and maybe the positioning of the pad above it. If a substantial part of the sound is transmitted through the wood itself, that would add another variable, but as far as I understand it that's not what you would expect to be the case. Maybe someone will pick up the glove... ?

Thanks,

Mark

[MarkvN]

One more remark for ttonon,

Seeing the formulas I had expected that it would be possible to predict whether a change in the depth of the sound hole from say 2 to 4 mm could have a noticeable effect on the higher partials, even if the exact changes would be difficult to assess. But maybe I'm being naive here.

[adrian brown]

Thanks a lot Tom for resurrecting this post; it's great to have a technical explanation for what for me is a musical issue. It's very kind of you to take the time to analyse my rather makeshift recording and I'm glad that my feelings (and my ears) seem more or less in line with your explanation. I don't think the amplitude as such was the problem for me, rather that these notes tended to stick out like a sore thumb when played in the typical sequence I recorded (also in chords). I don't know if it might be useful to therefore compare them more in relation to the notes in either side of the sequence with and without the blu-tack, to see if their harmonic architecture is more similar with than without it? I'll try to make new recordings next week of another instrument with the same modification, where the effect is less marked. I'll also try to be a little more tidy about the recording (It's difficult to record something, take the instrument apart and then play the same thing with exactly the same bellows pressure - I could almost do with connecting a manometer...
Thanks again for your help,

Adrian