Blowing machine

The Ultimate On-Line Whistle Community. If you find one more ultimater, let us know.
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: Blowing machine

Post by Tunborough »

Terry McGee wrote: Mon Feb 20, 2023 10:29 pm Looks much better to me, but I'll leave the experts to judge. I do note a little discrepancy where I change from the 5L/Min to the 20L/Min gauge but we were aware of that from an earlier experiment.

Are we there yet?
That is much better. A definite squared relationship, with just a bit of flattening around 8 L/min. The parabola is still reluctant to bottom-out at the origin; I need to give some thought to the implications of this. A more rigorous test of your dish soap would be to take measurements going up in flow, and then going back down again.

I don't think a bit of colour in the water would do any harm.
User avatar
jemtheflute
Posts: 6969
Joined: Tue May 23, 2006 6:47 am
antispam: No
Please enter the next number in sequence: 8
Location: N.E. Wales, G.B.
Contact:

Re: Blowing machine

Post by jemtheflute »

I see my flageolet efforts are being cited - thanks for the heads-up, stringbed.

I haven't read the whole thread, just the most recent 2 or 3 pages, and and not sure I'm particularly interested in the whole topic! Nor am I sure I have anything very much to contribute here, but I'll try to write a few things which may help...

First just to point out that there is a nice looking anonymous English-made (?Liddle factory style?) pillar mounted 6-key flageolet-piccolo set on eBay currently. Probably a bit over-priced IMO (and not selling and getting relisted because of that - though now open to Best Offer!), but it is complete, beak and all! www.ebay.co.uk/itm/165949035643. I'd be a little dubious about it being modern pitch - the vast majority of the English made ones are really Old Philharmonic Pitch A 452.4, though some, like my M Barr set, are just about viable tuned down to A440.

Some general responses to some of the points I've read up-thread:

I think you should take note that in historical terms flageolets do not have to have windcaps to be considered flageolets - and that is of course why Generation have always termed their whistles "flageolets". There's a very useful website on Flageolets in general, detailing the assorted types here: http://flageolets.com. It has a good structural explanation and history of the English flageolet. 1-keyed or keyless versions are not necessarily reversions from or simplifications of the 4/5/6-keyed examples - they seem to have existed throughout. I'd also note that, where an alternative piccolo head is provided, it is by no means necessarily a bad instrument as a piccolo.

My own favourite modern high D whistle is a boxwood one by Jon Swayne. His whistles are, I think, very much direct descendents of historical examples. They have a conical body bore, like the antique flageolets, though not necessarily with the same taper profile. IMO this bore conicity is probably what makes his whistles superior to my taste to most other wooden whistles with all-cylinder bores. I have plenty of videos of mine on my YouTube and Facebook. You can read Jon's explanation of them on his website: https://www.jonswayne.com/whistles. They are what I suspect would be described as "very free-blowing" in modern parlance - maybe even freer than Generations, but capable of much more power/volume and supportive of a large throughput of air to create volume without being air-hungry, harsh or hissy, but maybe not quite as free-blowing as a descant recorder, which of course has a significantly wider windway. Personally I very much dislike and in effect cannot play very strangled, resistant whistles like Goldies - even Colin's "free-blowing" models seem very constipated and overly resistant to me and I just make them squeak! (That's my deficiency, no doubt - so many great players love them and make them sound great - but they're not for me.)

Regarding "resistance" to blowing, none of the English flageolets with surviving beaks I have played has been subjectively significantly more resistant than a Generation whistle. The holes through the half dozen surviving bone and ivory beaks in my possession have bore diameters of between 4.05-4.73mm. With that information, Terry, you can make a few! I don't think length is a significant factor affecting air-flow, but they're all between c35-55mm long. Most are bone, not ivory. The narrowest of those is the Barr, which seems comfortably free-blowing to me, though I do notice it has a little more resistance than the bare head. I don't really notice any significant difference in resistance between blowing the whistle head directly or through the beak and windcap on most examples - nor would I expect to unless the bore of the beak was unserviceably narrow.

I have in the past experimented with placing (natural) sponge in the windcap chamber in various ways, and IMO it simply doesn't work - I actually seriously doubt how authentic or common a practice it ever was and suspect it is largely a myth. Even a small piece somehow lodged so it can't blow onto the windway opening can cause serious problems with resistance and control of airflow to overblow etc. and once saturated (which happens quickly) offers no advantage whatever - you just have to keep opening up the windcap to wring it out or swap in a dry piece, which simply isn't worth bothering with.

The main thing I notice when playing a fully assembled windcap flageolet is that, to the player, it seems rather quiet and distant. It took me a while and some experimentation to realise that that is not an experience shared by an in-front auditor. If you take the windcap off and simply blow through the main windway to play just like a normal whistle, the sound will seem normally loud and present to you and comparable to a familiar whistle, but an audience won't notice any significant difference between windcap on or off. It's just to do with the location of the sound emission from the whistle window relative to the player's ears! Holding the fully assembled toot far sideways so it approaches one ear also demonstrates this to the player. Most of the flageolets I have played in good playing condition have been at least as loud as, usually louder than, a good Generation whistle in terms of projection to an audience or recording device.

There are several videos using the M Barr flageolet on my Facebook, including one with a direct comparison between the flageolet and piccolo heads. This link will take you to those and some other flageolet materials I have posted.

There's a full demo of another anonymous English flageolet I restored some years ago on YouTube here (cf the video blurb for info): https://youtu.be/-2jQHShRSfw

Some other bits (photos, videos etc.) on flageolets I have posted on Facebook:
Alexander Liddle Flageolet/Piccolo set photo album - this one is unrestored but has its beak. There's a link to a video in the album blurb.
Starck whistle-flageolet - posts including videos and photos of a simple 1-key flageolet, which we would now no doubt simply call a wooden whistle. I have a several other examples, unrestored, of whistle headed keyed flageolet/whistles by French and German makers as well as English. My old mate benhall.1 has or had an antique 6-key all-metal whistle which I seem to recall played quite nicely.
Last edited by jemtheflute on Wed Feb 22, 2023 7:37 am, edited 2 times in total.
I respect people's privilege to hold their beliefs, whatever those may be (within reason), but respect the beliefs themselves? You gotta be kidding!

My YouTube channel
My FB photo albums
Low Bb flute: 2 reels (audio)
Flute & Music Resources - helpsheet downloads
david_h
Posts: 1735
Joined: Mon Aug 13, 2007 2:04 am
Please enter the next number in sequence: 1
Location: Mercia

Re: Blowing machine

Post by david_h »

A fitted trendline (I am just using a chart in Excel, nothing fancy) of Flow vs root P, not including the zero point, does go through the origin. (sqrt(P)=0.4129*flow + 0.0096.)

There is some scatter and it looks like the values from 22.7 l/min and upwards and the 6.3 - 10 l/min groups may be on steeper (higher Resistance) straight lines across the overall trend. So unless there is a flow model that would do that (or it is just chance) it looks like there is still a measurement issue. Is the overall trend good enough for modelling? If the 'cross trends' are because the flow meter was lagging and jumping to catch up then the overall trend, going through zero, might be good. Resistance of the calibrator is measured at 0.41. However, if the measurement is good over what look like straight sections but something changes between them then the resistance is greater (0.45 - 0.49) but we have a problem with a predicted flow at zero pressure.

I am not looking at pressure against flow because non-linear relationships are harder to handle. That we keep seeing straight lines of flow against root P as basic theory would suggest makes me think it really is that relationship and the problem is measuring it well enough for Tunborough to model.

If the actual relationship is the same at low flows can we dispense with the low flow-rate meter and not measure that low? That would remove a potential complication and any problems measuring the manometer at low pressures. Though if Terry needs that range for a real whistle head measuring it needs to be understood.

By the way, I am still thinking about the issue of air density in the flow metres. If Terry's valve is on the compressor side the density will be going up steadily with flow rate as the pressure goes up. Do we have an angle on the scale of that? It is related to trill's point about what gas the meters designed to measure. The effect on measurement is probably some Bernoulli-like relationship. If the valve is on the manometer side then the density will drop until the compressor kicks in, increase, then start dropping again as it cuts out. Or is there a pressure regulator after the reservoir?
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: Blowing machine

Post by Tunborough »

david_h wrote: Tue Feb 21, 2023 12:15 pm A fitted trendline (I am just using a chart in Excel, nothing fancy) of Flow vs root P, not including the zero point, does go through the origin. (sqrt(P)=0.4129*flow + 0.0096.)
Excel does support polynomial trendlines on charts. (That's what I'm using, too.) Depending on how much of the data set you pick, most of the time the parabolic trendline bottoms out at a positive, non-zero flow and pressure. Does this mean our flow meter and manometer are both reading a little bit high? This isn't consistent with the lowest-flow measurements. Or is the relationship between flow and pressure more complicated than a simple squared relationship? Not sure yet.
david_h
Posts: 1735
Joined: Mon Aug 13, 2007 2:04 am
Please enter the next number in sequence: 1
Location: Mercia

Re: Blowing machine

Post by david_h »

I prefer to keep things simple until we are sure we have got rid of, or can live with, the noise in the data. It can be very hard to interpret the meaning, if any, of lines fitted to non-linear relationships once they go beyond the data. Ask the pandemic modellers.

We can be fairly sure that the dominant physical process here would, on its own, give a 'straight line through zero' relationship between one variable and the square root of the other. Overall it is looks very close to that. The departures from that straight line then give us indications as to what else may be happening in terms of the things we a measuring, rather than some complex function of them. If, for example, the points that theory says should be a line look like a curve then we might doubt to theoretical model. It is more complicated to get to that point if we plotted and fitted a curve.

By 'nothing fancy' I meant I was not in a position (or intending (or able off-hand to remember how)) to compare the statistical significance of different interpretations, or justify selecting from the data. I suspect the departures from linearity are something to do with the instrumentation. One could just accept them as measure experimental uncertainty and move on to whistles, knowing that the data may be too noisy to pick up some fine details.

Theoretical question. Is the pressure at a windway exit, partially nestled in to the bottom and sides of the head and facing the blade, going to be atmospheric pressure for the purpose of the flow down the windway?
User avatar
Terry McGee
Posts: 3337
Joined: Sun Dec 12, 2004 4:12 pm
Please enter the next number in sequence: 1
Location: Malua Bay, on the NSW Nature Coast
Contact:

Re: Blowing machine

Post by Terry McGee »

Woah, lots there to catch up on! But just thought I'd throw in these new tidbits for anyone contemplating making up a system.

You'll remember mine goes Compressor feeds Workshop via a Regulator. Then I take a line from the workshop into the Lab which goes via a lever Valve to another Pressure Regulator, to give me convenient local control of pressure, and increased safety from accidently overdoing things. Then via a Flow Regulator, on to the Flow Meter or Meters, to a T-junction to feed the Manometer and through to feed the item-under-test.

I wondered if I really needed the Flow Regulator, so to find out, I bypassed it. I closed down the Regulator and the Needle Valve on the Flow Meter, and carefully opened the air feed Valve. Good, no flow. I gingerly opened the Needle Valve a few turns - very good, still no flow. Then started cranking the Pressure Regulator up and the flow came up. I can control the flow via either the Pressure Regulator or the Needle Valve or both. So the Flow Regulator is probably not an essential part of such a rig, although I'll probably put it back in as 1) I'm used to using it, 2) it's very smooth, and 3) it offers a bit more safety.

I also reintroduced a small amount of colour to the water in the manometer to help reading the meniscus. A couple of spot checks gave the same readings as yesterday, so I think we can safely say we haven't introduced any new errors. And the manometer looks responsive - the slightest change in flow produces a change in the manometer level, whereas before the detergent, you got the sense it was reluctant to respond. And it doesn't seem to care if I'm going up or down.
Tunborough
Posts: 1419
Joined: Sun Dec 05, 2010 2:59 pm
antispam: No
Please enter the next number in sequence: 10
Location: Southwestern Ontario

Re: Blowing machine

Post by Tunborough »

david_h wrote: Tue Feb 21, 2023 3:01 pm Theoretical question. Is the pressure at a windway exit, partially nestled in to the bottom and sides of the head and facing the blade, going to be atmospheric pressure for the purpose of the flow down the windway?
That has been my assumption. I'm willing to revisit it, but I'm hoping that isn't necessary.
Terry McGee wrote: Tue Feb 21, 2023 7:38 pm And the manometer looks responsive - the slightest change in flow produces a change in the manometer level, whereas before the detergent, you got the sense it was reluctant to respond. And it doesn't seem to care if I'm going up or down.
Good to know. Nothing left but to peer at the latest numbers until they reveal their secrets.
User avatar
Terry McGee
Posts: 3337
Joined: Sun Dec 12, 2004 4:12 pm
Please enter the next number in sequence: 1
Location: Malua Bay, on the NSW Nature Coast
Contact:

Re: Blowing machine

Post by Terry McGee »

Indeed. But we have to face facts - all our previous measurements are likely to be dodgy, particularly down low. So feel free to ask me to repeat any that will help you work out what's going on and will convince me we have the technology under control. I think we do have to keep in mind the inherent lack of precision of the measuring tools we have.
trill
Posts: 687
Joined: Thu Aug 17, 2006 8:44 pm

Re: Blowing machine

Post by trill »

Terry McGee wrote: Tue Feb 21, 2023 3:05 am. . . Mine look the same but are 20L/Min. . .
Note how the scale varies.

Image

I was worried it might complicate my check of Nitrogen-scaling.

I ordered a flowmeter. I think I can run a test comparing oxygen + air here.

More when I get it.
trill
Posts: 687
Joined: Thu Aug 17, 2006 8:44 pm

Re: Blowing machine

Post by trill »

Terry McGee wrote: Tue Feb 21, 2023 2:59 am . . . what interests trill and david_h have in this question. . . .
For me, whistles have been a gateway to the joy of music !

I wanted to play flute. Got a whistle as a stepping stone. Found it plenty of work but also plenty fulfilling, and decided to stay.

My dad was a machinist. I grew up having tools around + working on things. My natural urge was to try making whistles.

What could go wrong ? High-school physics right ? WRONG !!

So, I started looking at books (Benade, Hopkin, anything online). Learned terms like "bore pertubations" and "cutoff frequency". Found that when making a windway, 1 slip of the file can spell disaster. Talk about sensitive !

Anyway, my work involved lots of programming, making graphs, and comparing predictions to measurements. So, that carried over to my interest in whistles. And here I am.

As an aside, my impression is that the bulk of this thread has been all about pressure + flow. Very little mention of sound + musicality. Personally, I find both fascinating and hope learn and illuminate as much as possible.

My own experience as a player has shown *huge* variations in voice among whistles. There are variations in volume, windy/pure, rich/screechy, lots of air, little air. I simply like to understand things. As I've continued reading, I've found things like: slits alone can result in vortices all by themselves. Add an edge and you get "edge tones". Add a tube and you've got an instrument. The "wafty tones" are a totally new phenomena to me.

For example, this blows my mind . . .
Image

Been enjoying this thread immensely. Hope to contribute where I can.
User avatar
Terry McGee
Posts: 3337
Joined: Sun Dec 12, 2004 4:12 pm
Please enter the next number in sequence: 1
Location: Malua Bay, on the NSW Nature Coast
Contact:

Re: Blowing machine

Post by Terry McGee »

Interesting. I wonder why it's so compressed in the middle? Finally found an image of my 20L/Min one, for some reason it shows two!

Image

You'll note it doesn't seem quite so compressed. Except at the bottom, which is unfortunate for us. And it's why I had hoped to be able to use the 5L/min gauge to give better resolution, but that gets a bit confused by the seeming disagreement between the 5 and the 20. I have wondered if I should get a 10L/min gauge for the low work.

Hmmm, sudden thought. Should I be considering making a new manometer, combining a vertical section from say 100mm upwards, but inclined between zero and 100mm? Calibrating it might be fun...
User avatar
Terry McGee
Posts: 3337
Joined: Sun Dec 12, 2004 4:12 pm
Please enter the next number in sequence: 1
Location: Malua Bay, on the NSW Nature Coast
Contact:

Re: Blowing machine

Post by Terry McGee »

And I wondered again whether my Digital Manometer has anything to offer. It looks like this:

Image

So I hooked it up alongside the analog manometer and got these comparisons:

Code: Select all

Analog	Digital
0	-6.5
20	14
40	36
60	56
80	77
100	97
120	117.5
140	138
160	159.5
180	180
200	200
220	221
240	242
260	263
280	282
300	301
320	321
340	341
360	360
You can see it starts off with a significant offset (it's reading -6.5mm at zero pressure), but is pretty much in tune from 180mm on. I guess I could build a correction table, but it seems to be a klutzy approach.
david_h
Posts: 1735
Joined: Mon Aug 13, 2007 2:04 am
Please enter the next number in sequence: 1
Location: Mercia

Re: Blowing machine

Post by david_h »

Terry McGee wrote: Wed Feb 22, 2023 12:16 am I guess I could build a correction table, but it seems to be a klutzy approach.
Analog = Digital*0.9785 + 4.9175 is within 3mm throughout. However, it starts off -1.5mm at zero and with a few wobbles increases to +2.3mm at 260mm then goes steadily down to -2.8 at 360. Is +/- 3mm near enough to understand whistles?

Has that manometer got a zero set function?

For studying windway geometry, especially 'artificial' ones, how about putting another piece of pipe on the exit side, with a T-junction to the low pressure side of the manometer (analogue or digital) and then the flow meter before the final exit to atmosphere? That way the flow meter would be emptying to atmospheric pressure throughout.

Calibrating the sloping manometer shouldn't be hard. If the low pressure section had a 1 in 5 slope 1 mm up would be 5mm along the tube. Or draw horizontal lines on a backing board. As a check could progressively top it up and compare with the level on the other side. Would need to decide how to read a diagonal meniscus.
User avatar
stringbed
Posts: 189
Joined: Tue Apr 12, 2022 9:36 am
antispam: No
Please enter the next number in sequence: 8
Tell us something.: Playing woodwind instruments for over 70 years and deeply interested in their history, manufacture, technology, and performance practices.
Location: Sweden
Contact:

Re: Blowing machine

Post by stringbed »

Terry McGee wrote: Wed Feb 22, 2023 12:16 am And I wondered again whether my Digital Manometer has anything to offer.

You can see it starts off with a significant offset (it's reading -6.5mm at zero pressure)
According the user’s manual, the differential reading on this instrument does not automatically reset and can be zeroed by pressing the HOLD button for about five seconds.
david_h
Posts: 1735
Joined: Mon Aug 13, 2007 2:04 am
Please enter the next number in sequence: 1
Location: Mercia

Re: Blowing machine

Post by david_h »

stringbed wrote: Wed Feb 22, 2023 5:19 am According the...
Yes, I'd had a look at that as well but wasn't sure Terry hadn't done that. If it performs as specified that manometer would save a lot of messing around with water and peering at rulers. Simpler measurement may allow more measurements which is usually a good thing. I think the sealing wax and string approach may be more traditional though.
Post Reply