Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[IanD]

Why are they only slightly tapered in cross-section?

From everything I've seen, the best shape for the cavity under the reed would be roughly triangular with the tip of the reed and the pallet at the broad end. This pushes the both the Helmholtz and cavity resonances (see Tom Tonon's paper, these tend to cause notes not to sound) up to the highest possible frequency, and for a given size chamber (which is already minimised) allows a larger pallet hole opening for less obstruction between the reed and the ear.

In addition, since the size of the reed frames decreases with higher pitch, the bottom can be "sliced off" at an angle so that both the height and width decrease as pitch increases; in all cases the reed cavity is now triangular and just tall enough to fit the reed up the diagonal side. This would also avoid the need to put filler blocks inside the higher pitched cavities to reduce the volume, it's done automatically.

So instead of conventional reed blocks which are maybe 40mm high/25mm thick at the base and 30mm high/15mm thick at the top you get ones which are maybe (depends on wall thicknesses):
-- at the low end, 40mm high, 35mm wide at the base and 5mm wide at the top
-- at the top end, 20mm high, 20mm wide at the base and 5mm wide at the top

I suppose somebody's now going to tell me why this either isn't possible or won't work or just plain isn't what has always been done...

Cheers

Ian

P.S. Obviously you couldn't get as many reed blocks in because they're wider, but most melodeons aren't wall-to-wall reed blocks anyway -- 2 reed 2 row have 2 blocks, 3 reed 2 rows and 2 reed 3 rows have 3 blocks so both would be OK -- 3 row 3 reed ones might have a problem.

[finnhorse]

I suppose somebody's now going to tell me why this either isn't possible or won't work or just plain isn't what has always been done...

I don't think anyone has done anything of the sort.  I think people have voiced their opinions and brought up relevant criticism to some ideas that look very attractive on paper.  That doesn't call for pessimism.  But since we're on the subject, you'd probably have to take this idea to Italy where reeds are made, and good luck with them.. perhaps the Czechs or Germans would be easier to deal with?  You're a one man revolution! 

[Lars]

An interesting perspective in this discussion would be the characteristics of the sound.

The most traditional sound is that of the "regular" shaped reed blocks. They allow easy access to the reeds for tuning and repair, and are probably not as expensive as other ways of mounting reeds.

When the reeds are mounted all flat, like a Casti Lilly or an old single row 2v pokerwork, the sounds sound much more clear to me. Some say louder, some say better, but we can all agree that the sound is different to that of a block-mounted reed setup. The reed is directy in the stream of air - the air stream is not affected by turning corners, or the pressure inside a wedget or square block.

Then there's the casotto system where the reeds are mounted inside a rather peculiar chamber. This softens the tone considerably - especially with low octave reeds.

If the reeds were to be placed "tip down" on the blocks, wouldn't that make it much harder to tune and maintain? I had the idea that placing the pit over the narrowst part of the block cavity would make the reed start quicker - exept the very highest reeds.

Triskel from this forum, Stephen, has a meloden where there's only one larger hole to feed air to two reeds (one block). The same is the case with Hibbs3, Ryan, and they both claim the boxes to be VERY very loud. Maybe that's an idea? I know they are much much harder to tune, since it's harder to isolate a reed for tuning, but still. The overall sound is the main goal here.

[IanD]

I suppose somebody's now going to tell me why this either isn't possible or won't work or just plain isn't what has always been done...

I don't think anyone has done anything of the sort.  I think people have voiced their opinions and brought up relevant criticism to some ideas that look very attractive on paper.  That doesn't call for pessimism.  But since we're on the subject, you'd probably have to take this idea to Italy where reeds are made, and good luck with them.. perhaps the Czechs or Germans would be easier to deal with?  You're a one man revolution! 

No change to the reeds is needed, just the construction of the reed blocks. I'm actually talking to someone in the UK right now who is proposing to start making meledeons next year, maybe I'll see if I can persuade them to try some of these ideas out.

Ian

P.S. Maybe I should have but a :-) after the " I suppose" comment -- I've just become rather disillusioned about the negative attitudes of many people to change :-) :-) :-)

[IanD]

An interesting perspective in this discussion would be the characteristics of the sound.

The most traditional sound is that of the "regular" shaped reed blocks. They allow easy access to the reeds for tuning and repair, and are probably not as expensive as other ways of mounting reeds.

When the reeds are mounted all flat, like a Casti Lilly or an old single row 2v pokerwork, the sounds sound much more clear to me. Some say louder, some say better, but we can all agree that the sound is different to that of a block-mounted reed setup. The reed is directy in the stream of air - the air stream is not affected by turning corners, or the pressure inside a wedget or square block.

Then there's the casotto system where the reeds are mounted inside a rather peculiar chamber. This softens the tone considerably - especially with low octave reeds.

If the reeds were to be placed "tip down" on the blocks, wouldn't that make it much harder to tune and maintain? I had the idea that placing the pit over the narrowst part of the block cavity would make the reed start quicker - exept the very highest reeds.

Triskel from this forum, Stephen, has a meloden where there's only one larger hole to feed air to two reeds (one block). The same is the case with Hibbs3, Ryan, and they both claim the boxes to be VERY very loud. Maybe that's an idea? I know they are much much harder to tune, since it's harder to isolate a reed for tuning, but still. The overall sound is the main goal here.

I expect the sound would be much closer to the boxes with flat-mounted reeds, maybe even more so if the cavity volumes are even smaller -- also the tapered cavity shape (no flat end opposite pallet) would tend to reflect sound out the pallet instead of encouraging it to bounce around inside.

I'm not sure that having the tip over the narrowest part of the cavity will help the reed start quicker, in fact since the reed is started mainly by the air flow past the tip it might be slower. Having a smaller cavity should still help startup.

Is "tip down" harder to tune on a tuning table or just when mounted in the box? -- I don't know, maybe Theo or someone else with experience can comment. With modern tools (like Dirk's Accordion Tuner) you don't need to isolate reeds to tune them, you can see the pitch of up to 3 reeds while all of them are sounding. In addition this eliminates the need to do any actual tuning "on the box"; it can measure all the reeds in situ, then again when they're on the tuning table and record the differences, then tell you what you need to tune to on the table to get the wanted notes in the box.

It's interesting that you say the "one hole per two reeds" boxes are very loud, I think this confirms that a big hole opening is important for this. One reason for the Oakwood being so loud could be the size of the openings, these are rectangular not round and with very little gap between them, they're probably 1.5x the area of a round hole and 2x the size of a box with sliding stops.

Ian

[Dazbo]

Off the top of my head I'd think that they'd be more complicated to make, so more expensive, and would they sound much better in practice?  Interesting to find out.

[IanD]

Off the top of my head I'd think that they'd be more complicated to make, so more expensive, and would they sound much better in practice?  Interesting to find out.

I don't think they'd be any harder to make than a "normal" block which is tapered in both width and height anyway, it's just using much more slant on the sides than normal -- in fact if you made the block full height (like a Toblerone) and then sliced the face off before gluing it to the foot plate (the flat bit with holes in that sits against the soundboard) it might even be easier, especially since you then don't have to glue lots of little triangular filler pieces into the high reed chambers to reduce the volume.

Even with the increased face slope, ease of reed tuning with "tips down" could be an issue -- Theo, can you comment?

Ian

[Theo]

Even with the increased face slope, ease of reed tuning with "tips down" could be an issue -- Theo, can you comment?

Yes I'd agree with that.  The key question must be to test whether there are any real advantages to the system.  Maybe the changes in volume and tone will be worth the extra effort.

[IanD]

Even with the increased face slope, ease of reed tuning with "tips down" could be an issue -- Theo, can you comment?

Yes I'd agree with that.  The key question must be to test whether there are any real advantages to the system.  Maybe the changes in volume and tone will be worth the extra effort.

By the time the sound has had to pass through the base of the reed block and the soundboard you've still got quite a big mass of air in this resonator port, so maybe the sound won't be as close to flat-mounted reeds as I hoped.

But I'll see whether I can persuade someone to try it out ;-)

Another -- more radical -- idea...

For a 2-reed 2-row (or 2.5 row) it might be possible to get all the reeds flat-mounted without any Preciosa-style cantilevering. In this case there would be no separate reed blocks or chambers, the reeds could be mounted directly on a "soundboard" with the tapered reed chambers milled directly into it under the reeds (no built-up chambers needed) and the pallets landing directly on the other side of it.

This "reed pan" (like a concertina) could then be made removable for tuning (like in a concertina), and possibly held in place by screws sandwiching it between the bellows and the keyboard/action (like a concertina). If the whole thing was milled out a piece of (say) 10mm aircraft ply it would be very strong rigid and non-resonant, which exactly what Stormy Hyde was speculating is ideal for a reed block.

The fact that the internals would be a sort of concertina-melodeon hybrid construction need not be visible from the outside, except for the much shallower ends compared to a normal box. Tuning both inside and outside the box would be easy since all the reeds are flat and visible.

Comments anyone?

The slightly difficult bit is getting all the reeds in without ending up with a very deep (front-to-back) box; normal treble reed frames (no octave reeds) are 25-45mm long and 16mm wide, 4 rows of these would mean a box at least 20cm deep. If 3 rows of reeds are spread out past the ends of the button rows I think there's enough space, but this means many of the pallets/reeds will be well out of line with their respective button -- at least 2cm lateral shift even assuming the reed plates are butted together (which means end-screw fixing not wax), more like 3cm if wax is used.

This must be normal anyway for some of the boxes (like the Elysium) where the keyboard is much shorter than the height of the box, but I don't know how the problem is solved because none of the boxes I've ever taken apart have been like this. What's the normal way of doing this (other than the Preciosa in-the-keyboard solution which gives a very thick keyboard and can only do a limited shift) -- crooked (metal) levers before the pivot, crooked pallet rods after the pivot, both, something else? Does anyone have any photos?

Cheers

Ian

[Dazbo]

I don't think they'd be any harder to make than a "normal" block which is tapered in both width and height anyway, it's just using much more slant on the sides than normal -- in fact if you made the block full height (like a Toblerone) and then sliced the face off before gluing it to the foot plate (the flat bit with holes in that sits against the soundboard) it might even be easier, especially since you then don't have to glue lots of little triangular filler pieces into the high reed chambers to reduce the volume.

I was just thinking it was easier to machine rectangular chambers than triangular.  Would the reed plate fit okay on a suitable sized triangular chamber - ie the base of the triangle is narrower than the reed plate?

[IanD]

I don't think they'd be any harder to make than a "normal" block which is tapered in both width and height anyway, it's just using much more slant on the sides than normal -- in fact if you made the block full height (like a Toblerone) and then sliced the face off before gluing it to the foot plate (the flat bit with holes in that sits against the soundboard) it might even be easier, especially since you then don't have to glue lots of little triangular filler pieces into the high reed chambers to reduce the volume.

I was just thinking it was easier to machine rectangular chambers than triangular.  Would the reed plate fit okay on a suitable sized triangular chamber - ie the base of the triangle is narrower than the reed plate?

Maybe some confusion here (at both ends) -- think Toblerone, with the reeds mounted on the sloping faces and the pallets on the solid chocolate bit that holds it all together, but all with a divider down the middle to separate the two sides. The face of the chamber that each reed mounts on is rectangular, looked at from the end of the block the chamber is triangular -- like a wedge of cheese with the reed on one flat size of the wedge and the sound hole where the rind would be :-)

How difficult it is to make depends on whether you build the reed blocks up or machine them out of a solid block, I don't know which is more common. Either way, one method would be to make a rectangular block first, then bandsaw it along the diagonal, flip one half over and glue it back-to-back to the other half, bandsaw off the tapering "tip" of the Toblerone, then glue this to the footplate with pallet holes in.

Cheers

Ian

[Falseknight]

"Proving the concept" might be expensive, but once you know what you want, it is a trivial job for a CNC router.

[IanD]

"Proving the concept" might be expensive, but once you know what you want, it is a trivial job for a CNC router.

Trivial indeed, but I don't now if many melodeon makers -- small or large -- have such a beast, in my experience they use non-CNC routers and jigs.

It wouldn't be hard to make a prototype by hand with a simple jig, but might take some time to do.

Ian

[Pete Dunk]

It wouldn't be hard to make a prototype by hand with a simple jig, but might take some time to do.

Ian

It's a shame we can't get Martyn Banks involved, he's a dab hand at figuring out how to reproduce things without the benefit of technology like CNC. There again he doesn't recognise the benefits of computer technology either.  Bit of a bugger these Luddites! 

[IanD]

It wouldn't be hard to make a prototype by hand with a simple jig, but might take some time to do.

Ian

It's a shame we can't get Martyn Banks involved, he's a dab hand at figuring out how to reproduce things without the benefit of technology like CNC. There again he doesn't recognise the benefits of computer technology either.  Bit of a bugger these Luddites! 

The reason the person I was talking about is looking at starting to build melodeons (which might be a bit like Oakwoods) is precisely because we *can't* persuade Martyn to start making boxes again -- we even offered to order a batch of 10 identical Model 4 like mine (and at an increased price!) but still without success :-(

Ian

[Pete Dunk]

'snip' we *can't* persuade Martyn to start making boxes again 'snip'
Ian

There are times when I look back and think 'if only'. I was hanging around Oakwood Instruments doing bits of outwork on hammered dulcimer accessories and so on at the time the melodeons were in development. Martyn wanted me to learn how to tune 'accordions' but I was a concertina purist and wasn't about to be involved in anything so vulgar.  Que sera.  :'(