Design and Development of the Composite Drone Reeds for Traditional pipes and Schoolpipes

(updated 22/07/13)

Description of the design history
One of the most difficult areas of the Schoolpipes design has been the provision of totally reliable replacement drone reeds. I have been working on this right from the start and have thought that the problem was solved several times.
The nub of the problem is that most of the common materials used for reed making are hydroscopic and thus change with the weather. One of the important requirements is that the materials used should be commonly available, reliable and not hard to use.
I started off following the accepted traditional composite drone route using a folded brass body fitted with a thin cane tongue. These can be made to work easily as is proved by the number that are fitted successfully to traditional sets. The major drawbacks are that they are rather labour intensive to make, require careful setting up, are rather fragile, susceptible to atmospheric conditions, require skilled intervention during a settling down period of use and would require complex tooling to make the bodies in sufficient quantities & in a repeatable way.
Quite early on I thought that one of the tricks that would solve the reliability problem would be to preset the curve in the tongue of the reed thus obviating the need for the flicking and tweaking that is part of the traditional "getting the reed to work" process. but this is a tricky thing to sort out as, not having a clue as to the physics of the reed, it would be a trial and error process with each error costing several days effort and some cash (I do this stuff for free). Here are some of the variables that I have been coping with:-
As you can see there are an enormous number of possible permutations. Some can be quickly dismissed but some are potentially viable and so must be tried. Nearly all of the body types worked to some extent. Of the different tongue materials tried the most promising early candidate was the 1/64" ply. This material seems to have the same tonal charactistics as the original Arondo Donax (and the same faults). The main benefit of using it is that it is cheap and of a consistant quality. This material was used for all of the drone reeds in the school sets up to the the middle of 2012. The principal limitation, and the reason that I have spent some 6 weeks of effort looking for a better material, is that it has the same problem as the Arondo Donax material in that even if it has been flicked and curved to work well and sound good it can't be relied on to maintain that state. A change in the weather or just the passage of time will cause the reed to stop working and it will need some manual intervention to get it working again. This is a "show stopper" for a reed that needs to be a drop-in replacment by a non-skilled (in the magic art of pipemaking) pipe teacher.
All of the likely combinations need to be tested by making at least 6 off and the trial conditions of these should be as controlled as possible. To make this less of a lottery I have built a test rig so that I can quickly try each reed. To see the test rig go here.
Drone reed requirements
I tried a number of different drone reed designs. I was looking for a reed design that achieved the following criteria:-
  1. Cheap to make
  2. Easy to set up
  3. Made from easy to source materials
  4. Reliable
  5. Able to be clamped in place
I decided early on that the traditional cane reed was too tricky to make and keep working so I tried a number of different ideas:-

In the picture above are a number of the designs I tried. The ones made from square section brass tube were very successful. The opening was milled and the end of the tube blanked off with a square of balsa held in with super glue. The top hat bush was glued on and the gap between the round and the square filled with the glue. I tried a number of different ways to hold the tongue to the body and probably the best from a production point of view was a narrow (1/4") strip of sticky tape. I used a tape made by Scotch called "magic tape" that is used for mending torn drawings etc. and has a much better and more permanent glue than Sellotape. I also found that a small O ring was a good bridle and easy to move about.
I started by making the tongues from the same material that we use for chanter reeds but the preparation of it is a bit tedious and hard to replicate with any accuracy so I tried a thin (1/64") plywood that I use for aeromodeling and I was pleasantly surprised to find that it was every bit the equal tonally to the reed cane tongues and so much cheaper and easier to use.
One of the influences was the new modern designs of Highland pipe reeds. These are extremely reliable and easy to set up but much too big for the smallpipes so I looked to see if there was any of the features that could be used on smallpipe reeds.

The Highland pipe drone reed shown above has a number of interesting features. The tongue is a carbon fibre item clamped with the grubscrew at the left side. The bridle is a cast metal item and is released and moved with the grubscrew second from left. At the right hand side is a rubber plunger that is used to adjust the internal volume of the reed. It wasn't clear at first what this did but I have found out that reducing the volume of the internal chamber raises the pitch of the reed and increasing it lowers the pitch. Unseen at the back of the body are 3 grubscrews that allow adjustment of the curve of the body thus adjusting the opening between the tongue and the face. This allows the reed to be set to play at the correct pressure. Click on the picture for an excellent set of composite highland pipe reed pictures
For the smallpipes most of the ideas require too much space to be adaptable to the smallpipe drone reed but I did plan to try some of them.

This reed body is machined from acetal and the intention was to tap a M3 thread in the end and use a nylon M3 screw to adjust the internal volume. I also planned to try a moulded carbon fibre tongue and I intended to mould the slight curve into the tongue. As this effort was likely to take longer than I had available(I was still working full time) it remained on the back burner.
The Schoolpipes project moved to the north and I was not needed to do any of the production reed making.

Where I am today (26 June 2012)

I understand that the problem of fitting working reeds and keeping them working has proved a major obstacle to using the pipes in the teaching roll that they were designed for. The project has languished unloved and unwanted till today when a combination of enthusiasm from the new management of the Pipers Society, interest from Kathryn Tickell, my retirement from full time employment and some help from others in the NPS, we are once more trying to get things working. In 2011 I had 6 sets of un-made-up components to act as a prototype pool to develop the reeds especially the drone reeds. These sets were used for a trial session at the Cambridge Folk Festival and were very successful.
I was able to check the manufacturing quality for the first time and it was very poor. The company who did the machining (North East Assemblies) appears to have ignored the tolerances and in some cases even the dimensions. I have worked out a scheme for solving the manufacturing problems and I aim to have checked it out in the near future.
Early in 2012 I was issued with 10 sets from the "Working Pool" There were no working sets in this group. only 2 had chanter reeds and both were broken, 2 drones had reeds but they worked badly. I have now reworked all the mechanical problem with these sets and am in the process of setting them up with working reeds.
My experience with the ply tongues is that they require some hands on effort to keep them working - flicking the tongue open if it settles to a too closed state. This would be acceptable if the teachers using these sets had been able to do the reed tweaking but sadly this has not been the case. I have put considerable effort into creating G and d drone reeds that are completely reliable. My hope is that if a reed stops working for any reason then a replacement can be simply dropped in place in just a few seconds and will work without any intervention.

Design of the carbon fibre tongued drone reed

I have a design for the G drone that appears to be repeatable to make, reliable in use, very easy to play i.e. has a small pitch change to bag pressure ratio and sounds as good as a good cane reed. By adding a very small amount of weight to the tongue it appears that this same set of dimensions will work for the big D drone but I still have to do the development to prove that.
I have a design for the small d drone that looks very promising.
One extra change to the design has been to modify the interface to the drone by soft soldering a 1/8" o/d pipe to the main body and pushing the top hat interface piece onto that. By doing the fitting of the reed this way the same design will also fit the traditional set by just replacing the push on top hat with some waxed thread wrapping.
I have tested this on the G drone of a friends Nigel Barlow set that had dropped its traditional composite reed into the bag and the tongue was lost. I have also replaced the small d drone on the same set as the original reed was very unreliable and quite sensitive to pressure changes.
The Carbon fibre G & d drone reeds are the most stable that I have played and the pitch changes very little for a significant pressure change. This makes them very easy to play. The main feature of this reed design is that the reed body is bent rather than the tongue being curved. This has enabled the reed tongue to be moulded flat in carbon fibre thus giving a reliable action not affected by temperature, time, or humidity.
I have tried a few of the variables and have settled on a working set of dimensions that I am happy with (at present). Here are some to the options that I had to play with:-
I had the option of two thicknesses of Russian unidirectional carbon fibre cloth: 0.08 mm and 0.12 mm thick. I was lucky that I have a friend who uses this stuff for model aeroplane making so I was able to try some offcuts to see if we were in the ballpark. Only one of the layups worked sufficiently well for me to think that it was possible to use. This was a layup using 3 layers of the 0.12mm cloth having the top and bottom layers running across the tongue and the middle layer running along the tongue. I then persuaded my friend to do me a 4" square sheet of this layup and a similar sheet using the 0.08mm cloth. The resulting sheets are 0.39mm and 0.28mm thick. There is some slight variation in thickness across the sheets but that doesn't seem to have caused any problems.
I used 4mm wide strips of the 0.39mm sheet for the G drone reeds and 3.2mm wide strips of the 0.28mm sheet for the small d drone reeds

As part of the experimentation I used one of the first good G reeds to explore the effect of drone length on the pitch. Because all the drone parts of the Schoolpipes are interchangable this was easy, if a little time consuming, to do. The results were interesting. I was able to achieve the same pitch with a number of different lengths which seems to indicate that the large diameter part of the bore that results from the tenon of the sliding part is having a significant effect on the pitch.
To see the results go here

G drone reed

Carbon Fibre G drone Reed FrontCarbon Fibre G drone Reed Side

Click here for dimensioned drawings for the G drone reed

Small d drone & Large D drone reed
Small d Carbon Drone Reed
(the tongue hasn't been glued on yet and is held with the green silicon rubber rings as shown)
Large D Carbon Drone Reed
note the bluetac tip weight
Click here for dimensioned drawings for the small d drone reed

Reed making pictures
Cutting the carbon fibre strips for the tonguesMeasuring the actual gap at the tip of the reedAdjusting the bend to give the required gapReed bending jig
Slicing the bridles from silicon tubeSliced bridles ready to be usedClose up of the G drone reed tip

Some observations
  1. The bore of the reed (the soldered in tube bore) effects the volume of the drone
  2. The bridle has a significant effect for quite a small movement
  3. The position of the bend is very important - 1mm can make a difference of a tone in the sweet spot of the reeds operation
  4. If the drone has to be shortened too much by sliding the top moving part too close to the top of the fixed part to achieve the correct pitch, the quality of the sound suffers - It gets hard sounding

Evaluating the Performance of the Reeds
I have found a way of gathering accurate data reflecting the reeds performance. I am looking to see how the design of the reed compares with other commercial and traditional designs. My intention is to measure the change of pitch in cents as the pressure supplied to the reed changes. This should give a good guide to how the reed will feel to the player and how easily it stays in tune whilst being played.
My equipment used to do this testing is the same basic setup as I used to set up the schoolpipes drone and chanter reeds. There is a description of the set up here
Here is a picture of the test rig being used to gather data on the drone reeds.
The tool I am using to gather the pitch data is AP Tuner 3 which detects the actual note being sounded and gives a readout of the tuning error in cents. The resolution is to 0.1 cents but realistically it is only really feasable to work to 1/2 cent and this is more than the ear can distinguish.
The data below is up to date as of September 2012. For an interesting discussion on this topic go to the NPS discussion forum
Results up to the end of September 2012

The testing of these reeds was done in my test rig as described above. The reeds and the tongues were as close to identical as I can manage. All the reeds were G reeds made to the drawings above and fitted with 4mm wide x 0.39mm thick carbon fibre tongues also as described above.
The testing was done in a concert G drone, made for use in the Schoolpipes, by adjusting the drone to play as close to concert G as possible when the pressure was 15"WG (15 inches water gauge). The pressure was then changed by 0.5"WG and a reading taken of the change in pitch. This was repeated for pressures above 15"WG up to 17.5"WG, the maximum the test rig provides) and down until the reed stopped sounding (or when I felt that any lower would add no value to the data). All this data was recorded and loaded into a spreadsheet to create the graphs.
This set of reeds is from some that I made during my search for a really reliable reed for the Schoolpipe Project. I had noticed that the results I was getting were quite insensitive to pressure changes and this is shown in the graphs. This design has also proved totally reliable and continues to work without the need for any intervention. All of the reeds in this set were stable to better than 5cents for 2"WG pressure deviation. One of the reeds (021) is different to the rest and that reed has a narrower carbon blade (about 3.3 wide) compared to the 4mm wide blades of all the others.
I have done a similar data gathering exercise for small d drone reeds. Again the design is as described above using 3.2mm wide strips of the 0.28mm thick carbon fibre. I will probably have to look at reed #022 to see why is is so different to the others.
This set of reeds is also from some that I made during reed development for the Schoolpipe Project. This design has also proved totally reliable and continues to work without the need for any intervention.

Results from traditional sets
One important set of data that I need is from other designs of reed so that I have a base line for comparison. I decided to do this using a more traditional drone design. The drone I used was from my first keyed set and was made to the dimensions in Cocks & Bryant revised 1975 edition plates 8 and 9. This set of drones, when played using traditional all cane reeds sounds and feels just like any other set. I started by measuring the traditional all cane reed fitted to this drone. The actual playing length of the drone was 346mm from the tip of the reed tongue to the centre of the top side hole in the drone sliding part when playing the traditional all cane reed. I have also fitted composite reeds to this drone so that I can get a set of comparative pitch and pressure measurements. When looking at the graph it is important to remember that this set of data is taken from only 1 traditional, 1 machined and 1 bent reed design. The yellow line was so good that I went over the data several times as it seemed to be too good to be true, but I was able to repeat it several times. Here is the graph of the pitch/pressure.

This is the set of reeds that the data was taken using. I also swaped the plastic and the carbon tongues to see what the effect was. The middle drone reed is a "new design" Schoolpipe G reed and the bottom reed is a machine body F reed from Richard Evans (a rather old one now).
This is the test rig being used to get data from the traditional drone. The interface stock is held in a vice to give me 2 free hands to adjust the needle valve that controls the pressure and to write up the data.

Trying the effect of changing the internal volume of a drone reed

I have wondered for some time if the internal volume of a drone reed has any effect on the performance of the reed an what if any that effect is. I have now had a chance to make a prototype and test the result. I used a reverse design where the pivot of the reed blade is below the vibrating tip of the reed blade rather than above as in the traditional design. I had tried things this way round much earlier during my efforts to fine a good reliable reed design but this time I used a solid machined version based on the design by Richard Evans . This enabled me to tap out the internal bore of the reed body and, by using screws of different lengths, change the internal volume of the reed. The results showed that, although a small difference was measurable, the effect of reducing the internal volume was minimal. reducing the volume by 6mm (changing the 6mm long screw for a 12mm long screw) raised the pitch by 20 cents which was the same as moving the sliding part 2.5mm.

What is next?
There is a lot more work to be done before the fog clears. I have high hopes that the tongue can be made from 0.020" plastic as that is very cheaply available. I will publish more info what I have it. I need to do a lot more testing of the small d drone reed and gather some comparison data from traditional all cane d reeds. Having done that I need to do the same process for the big D and the small g reeds for traditional G & F sets. So a lot more to do!

Back to Introduction page
Back to Index page

© Mike Nelson. All rights reserved.

This manual has been put together by Mike Nelson for his own amusement :)

Go back to Mike Nelsons Home Page

This site is created and maintained by Mike Nelson for his own pleasure

1997 Mike Nelson