Blowing machine

[Terry McGee]

. . . we're talking dancing beads . . . .

If I may ask: do they dance all the time ? Or only at the top of the range ?

I just watched both flow meters as I wound up from 0 to 40L/Min. Up to 10L/Min each, they are pretty stable, becoming increasingly unstable as you then head up to 20L/Min each. But it doesn't seem to be a simple uniform increase in dancing. At some flow rates, one will be dancing while the other is steady. Or one will be steady for a while and then start to shimmy, and maybe then to bob up and down a little. Or you can have both beads vacillating, but at differing rates.

I understand that the bore inside is vaguely conical, so as we go up, the bead has more room to wander about in and dance if it feels like it. And as we saw, the scales are not linear, suggesting that the conical bore is not linear. You can imagine with all that air rushing past the bead and junctions in the bore, it would be pretty amazing if the bead just sat there!

[Terry McGee]

OK, the shorter tube....

Notes:
Remember I mentioned that the tuning meter seemed unsure of the frequency at the lowest flow rates? See here with the shorter tube, I think we must be in an Aeolian mode for those first three readings. I'm hearing the same as the tuner, so it's real, not just a measurement quirk. I can revisit the longer tube in this region if we need clarification.

I've ended up taking more measurements than we had talked about. Even once you have identified the top and bottom of a regime, it's not easy picking halfway, quarter way etc points. It's easier just to pick the next round number on the flow gauge. Hope that's OK.

Note when I get to 20L on the LH meter and bring in the second flow meter we see an anomaly, so I backtracked there in case it's important. (That noise off-stage is trill unsuccessfully trying to muffle an "I told you so"... )

Again a chaotic gap between Regimes 1 and 2, but hysteresis between 2 and 3.

Only 3 regimes available here. I cranked the flow up way beyond 40L/M until the overloaded flow meters were complaining, but couldn't get to R4.

Code: Select all

Feadog Mk1 with 146.08 mm blade to end of tube, 8 April 2023			
Regime	Flow	Press	Hz
Aeol?	2	3	973
	4	4	1005
	5.5	5	1034
			
One	5.8	4.5	622
	6.2	6	664
	6.6	7	729
	8	10	870
	9	13.5	947
	10	17	975
	11	20.5	984
	12	24	992
	13	28	999.5
	14	33	1007
	15	38.5	1014
	16	42.5	1018
	17	48	1022.3
	17.4	52	1024
			
Two	18.3	55.5	1930
	19	58.5	1940
	20	68	1962
	20	60	1946
	21.2	68	1962
	22	72	1970
	24	85	1980
	26	105	1997.3
	28	124.5	2012
	30	145.5	2021.3
	32	159.5	2030
	34	181.5	2040
	35	193.5	2047
			
Three	30	139.5	2924
	32	159.5	2939
	34	181.5	2951
	36	200.5	2960
	38	230	2973
	40	267	2990
	>>40	580	3099

[trill]

. . . it would be pretty amazing if the bead just sat there!

Indeed, especially at higher flows.

Thank you for checking and for giving such a thorough explanation.

[trill]

. . . That noise off-stage is trill . . .

It's true. I have a track record of generating noise . . . like a loose BB gun !

Honestly, though, it gave me a good laugh !

[Terry McGee]

Now, while we wait for Tunborough to finish hunting Easter Eggs and get back to analyse the short tube results, I got restless. Remember I mentioned having made a 14mm headbore tapered-bore piccolo some years back, but decided at the time it got too twitchy up around the top of the 2nd octave to be able to release upon an unsuspecting public. So, I got to thinking, I wonder how its tapered body would go with a whistle head? So I thought to play...

The only whistle head I had in that region of bore was an ancient Bb Generation - from the era with visible ridges fore and aft, and a peculiar oval on the back, perhaps the injection point for the molding process? Still stuck now very firmly to its brass tube. Getting it off was quite an endeavour. And, once cooled, sanding out its socket to allow the tube back in ditto. We have to keep in mind that these antiquities are shrunken relics of their former glory. (Glory? I don't remember that much glory! In fact, on its Bb body, there are a few slightly scritchy notes.)

So I then needed something to join the head and the tapered body section of the piccolo body. I ended up turning a 40mm-long tuning-slide section, and making up a delrin barrel to fill in the rest of the length. I made the bore of the barrel 14mm, as that was the bore of the piccolo head. And the tuning slide section bore about 14.8mm, the same as the Bb whistle body. There could be an argument that I should have made the slide bore 14mm, same as the old pic head bore, or that I should have made the barrel bore 14.8mm, same as the Bb body bore. To be determined....

I will concede it all looks a little odd - bright faded pink head, glimpse of brass tube, black delrin barrel and rich brown oiled timber body. But hey, nobody needs to see it, do they. Unless it becomes my favourite....

And it plays OK, pretty well even. Although I was immediately struck that the low D was flat. Interesting, I hadn't been aware of that in the piccolo. Anyway, here's what I found when I measured both:

Code: Select all

14mm Whistle Head vs Pic Head, on tapered body		
Note	Whistle	Pic
D5	-21	-5
E5	-5	7
F#5	-15	-5
G5	-6	-4
A5	8	-2
B5	14	-3
C#6	1	-24
D6	1	-3
E6	0	5
F#6	-11	-1
G6	-11	3
A6	-17	9
B6	-13	11

More below, as I'm struggling with one of those Internal Server Errors...

[Terry McGee]

Continued....

(I should mention I'm using Flutini to do the measurements, just running up and down scales, arpeggios, passages, etc and letting it accumulate averaged results.)

Looking at the Pic results first, we notice the tuning isn't bad, especially remembering that this was a prototype which I didn't bother to progress further. The worst note is the flat C#6, and that could probably be tweaked in a later version if needed.

The whistle tuning is not so good - a bit lumpy, with the low notes and high notes on the flat side, and the middle notes sharp. Note that the flat C#6 is no longer a problem! It's worst note is the flat low D which I had been immediately aware of in playing. And that note is a little weak. But again, we need to keep in mind it's essentially a found object, not a properly developed instrument. But certainly a possible starting point for further development.

It does suggest I'm right in wondering about the simple interchangeability of pic and whistle heads, as was common with fife/flageolet sets back in the period. Given that on the flute, we tighten our embouchure and possibly aim the jet a bit higher to go up an octave, whereas on the whistle all we can do is blow harder, I can't really see you could have one body and two heads and achieve the best results from both. But maybe good enough....

So, this is my first recent experience with tapered bore whistles (my ancient Clarke C long having disappeared), and I'm liking it. Apart perhaps from the flat low D. Tempted to give it a tweak....

[Terry McGee]

So, you'd think with all that adventuring and butchery, I might give it a rest for a while. Nah, not me. So a bit more tweaking...

I decided to try increasing the bore diameter of the new delrin barrel. I had made it 14mm to match the old pic head, but then tried 14.5 and subsequently 15mm. That seemed to iron out some of the tuning bumps (see below), but left the bottom note still on the flat side, noticeably so.

So I then backreamed the foot, which had the usual effect of raising the bottom note without messing significantly with its octave. Gave the bottom note a bit more volume and authority too. But I noticed that left the F# notes a little flatter than I would like, so I opened that hole out a bit by drilling and further by undercutting. Those results in the Twk'd column below.

Further fiddling around would be possible, but moving some holes a bit would probably bring bigger improvements. Mk 2, if we get that far!

Code: Select all

Whistle Head vs Pic Head, tweaked					
Note	Pic	Wh 14	Wh 14.5	Wh 15	Twk’d
D5	-5	-21	-21	-16	-2
E5	7	-5	-9	-9	-6
F#5	-5	-15	-14	-11	-11
G5	-4	-6	-8	-6	-4
A5	-2	8	5	3	0
B5	-3	14	8	-1	-4
C#6	-24	1	0	-10	-18
D6	-3	1	2	0	8
E6	5	0	6	8	7
F#6	-1	-11	-9	-1	-1
G6	3	-11	-3	6	0
A6	9	-17	2	13	7
B6	11	-13	9	27	21

And remember I mentioned a few scritchy notes, especially with the head on its original Bb tube? Looking down the windway showed a bit more light through the body than seems desirable, so I resolved to try "lowering" the blade. I have some nice 0.2mm annealed copper shim material, so I thought that would be a good start. I cut out a tiny patch the width of the window and about 4.5mm long - not so long as would be at risk of being bumped if the slide were pushed fully home. I balanced that on the end of a paddlepop stick, put a drop of rubberised contact cement on it, and offered it into the head socket, raising it up to stick to the underside of the ramp platform with its leading edge protruding every so slightly into the windway. (There might be a better well-recognised way of dealing with this - do let me know!) It seemed to do the trick - the scritchies have been expunged. It's possible that I should repeat the medicine and drop another 0.2mm, but maybe I should be looking at making a new wooden head instead! The family are pleading with me to try dyeing the lolly-pink head! It does look awful against the other materials.

I think at some point it would be worth documenting what actually happens when you lower the ramp base. Even if it's just in terms of detecting and listing issues before hand, and reviewing how far you have to go to resolve those issues. We should think about all that. (Again, let me know if someone has already done that!)

That's probably enough excitement for one day. I might just go and play some whistle....

[trill]

. . . end of a paddlepop stick, put a drop of rubberised contact cement on it, and offered it into the head socket . . .

My own excursions into whistle-making showed them to be *very* sensitive to teeny-tiny changes. "one slip of the file . . .".

Any chance for a picture ?

[Terry McGee]

. . . end of a paddlepop stick, put a drop of rubberised contact cement on it, and offered it into the head socket . . .

My own excursions into whistle-making showed them to be *very* sensitive to teeny-tiny changes. "one slip of the file . . .".

And really hard to measure in to see what level of manipulations should be done. We need a breakthrough here.

Any chance for a picture ?

Urk! What of? You've seen paddlepop sticks before I imagine...

[david_h]

I think at some point it would be worth documenting what actually happens when you lower the ramp base. Even if it's just in terms of detecting and listing issues before hand, and reviewing how far you have to go to resolve those issues. We should think about all that. (Again, let me know if someone has already done that!)

If I understand what you have done then the jet is being split in a different place. So I think we expect the harmonics to change as descibed in the paper I linked above (too far above to find, so here: https://newt.phys.unsw.edu.au/music/peo ... al1980.pdf )

Terry - next time you get the system error check if you have any multiple full stops at the end of a paragraph and try deleting them. Doing that allowed me to make the post I was failing with last week. But I also managed top make posts that have them

[Terry McGee]

Thanks, david_h, I'll try that next time!

Now, something else to ponder upon. I mentioned above that I'd snitched the head off my ancient Bb Generation flute for experimenting with the tapered-bore high D whistle. And that I added a scrap of shim copper to lower the underside of the ramp. While I was doing all that, I noticed something I thought strange. Strange even by my standards. And I'm a flute maker!

You might remember previous discussions on how sometimes the block, which forms the floor of the windway in some designs, protrudes a little into the window space, often accompanied by a chamfer atop its protruding end. Most commonly seen on lower-pitched recorders and low pitch whistles. On this intermediate pitch whistle head, we see exactly the opposite. The floor of the windway stops 0.4mm short of the window, the opening of the window being defined on the entry side by the top of the windway. And no chamfers, neither top nor bottom. So what's that about?

The window length (measured from end of top of windway to blade) is currently 5.87mm. It would have been a smidge longer before I added my mighty 0.2mm thick copper slab under the ramp. But the difference wouldn't be anything like the 0.4mm shortfall in the floor of the windway.

[david_h]

The floor of the windway stops 0.4mm short of the window, the opening of the window being defined on the entry side by the top of the windway. And no chamfers, neither top nor bottom.

On my Bb head of the same general description ('seam' and oval on the bottom) the floor seems to be same length at the sides but shorter in the middle. It's a curve but most of the offset is within the first 2mm or so.

Mine is blue on a plated tube though. It feels like I could get it back on the tube but it now does service on my WARBL (taped over) so I am not going to risk splitting it.

However, looking at the D Generation heads the floor on a blue 'old' one is maybe slightly shorter but on a blue 'new' one distinctly longer than the top. A 'new' red one is the same at one side but a tiny bit longer at the other and not very well formed. I wonder if this is why nickel sounds different to brass.

Sorry for the qualitative descriptions I will try to find a way of probing them.

[Terry McGee]

Hmmm, I'm not sure that we could expect to see a systematic difference between blue and red. I don't imagine they have and have always had separate casting lines for the two colours (but happy to be proved wrong!). I imagine they have blue weeks and red weeks, as supplies demand. And now green weeks! It would be fascinating to know stuff like how many lines they run and how long each head stays in the mold, but we'd probably need to get one of the workers really drunk to get that kind of information...

I rounded up the motley pile of old Generations I have here. I'm assuming there is no difference in heads of the same bore, eg Eb and D, and I couldn't guarantee that they are still on their original tubes anyway, so I've lumped them together.

Eb/D:
Faded Pink, ridged, 2, flush, OK. (One of these is Old Gen in my data)
Bright Red, ridged, 1, flush, plays badly but ramp is visually malformed

Faded Pink, smooth, 2, protruding ~ 0.25 - 0.3mm*, both very poor players
Bright Red, smooth, 1, flush, OK
Bright Red, smooth, 1, protruding ~0.33mm*, good in top octave, poor notes in bottom

C:
Blue, ridged, 1, protruding just 0.18mm, good

Bb (now on experimental tapered bore D):
Faded Pink, ridged, 1, short by 0.4mm, very good, but needed ramp base lowered by at least 0.2mm

* I noticed that in the several protruding cases I came across, the degree of protrusion varies across the width. It rather looks like someone (and of course it might have been me!) has got in there and pared it back from it's out-of-mold state. Anyone else see such signs of tampering? Could yours be that, david_h?

And the protruding ones seemed largely to be the ones with the poorer performance, particularly down low. But of course that could be due to other factors I haven't investigated. It could be interesting to sacrifice one to experiment...

And what's with the difference between faded pink and bright red? Just exposure to sunlight perhaps? I do remember a time when I kept my whistles in a jam jar at the window end of my desk.... The Bright Red ones having been left in the drawer as they were not great players!

[Terry McGee]

Now, trill, assuming you are familiar with the paddlepop stick and weren't hoping for an image of that, is this what you were hoping for?



You were warned - it's ugly! From left to right:
- the really old Generation Bb whistle head in faded pink,
- a brief glimpse of the brass tuning slide I turned up (not having supplies of that diameter tube), then
- the barrel in delrin, and finally
- the prototype tapered piccolo body I'd made years ago but decided not to progress, as the top few notes in the second octave were pretty hard to control.

A few things to note:
- even in this image, you can see that you can't see the end of the floor of the windway. Is "undertruding" a word?
- a hairline crack in the head socket. (a very thin dark line running just below halfway at the right hand end of the pink head.) I see that in quite a few old Generations. I guess it's further evidence of shrinkage over time. Fortunately, it doesn't seem to progress to leaking.
- I didn't waste much time on finishing the visible parts of the Delrin barrel. It's job is just to hold the ends together until I work out if this is worth progressing.
- not even a terminal metal ring on the wooden body. Again, it's part of the journey, not the destination. The timber is an Australian acacia, possibly gidgee or lancewood.
- pretty big holes, but not hard to cover as the outside diameter is considerably greater than on a brass tube whistle.
- holes not perfectly round and crisp. Being a prototype, I would have walked them a bit as well as enlarging them during the tuning process.

You can see why the family are pleading with me to dye the head. I image that that's not so easy! Better I think to make a new head, slide and barrel in matching timber when I'm convinced it's got a future.

I decided to lash out and order a Clarke Sweetone to further update my experience of more recent tapered flutes. Should be here in less than a week. Will be interesting to compare them.

[trill]

. . . You were warned - it's ugly! . . .

Nonsense ! It's magnificent ! Exactly what research-whistle should look like !

. . Is "undertruding" a word?

Couldn't find a definition for that or pretruding.

Also, I hope you post some reactions to the Sweetone when it arrives.