News:

If you have difficulty registering for an account on the forum please email antespam@gmail.com. In the question regarding the composer use just the surname, not including forenames Charles-Marie.

Main Menu

Peter Collins on Tracker design, Tracker "pluck" force - BIOS journal issue 6

Started by David Pinnegar, January 26, 2012, 08:30:01 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

David Pinnegar

Hi!

There could be usefully on this forum a greater depth of hard information so that this forum is a resource to which people come. Recently a Surrey organ has been proposed to be converted from tracker to electric action and I am very far convinced that that is the appropriate or necessary thing to do as a matter of conservation of the integrity of instruments, even though at Vale Royal Tunbridge Wells I have experienced possibly such an instrument which is to be exported to Malta which is possibly one of the most exciting small to medium sized instruments I have ever come across.

The following article written by Peter Collins in 1982 for the BIOS Journal 6 is particularly relevant and possibly not so well known by today's organ consultants.

Apologies for any errors from the scanning process and I trust that in view of its academic importance, benefit to organ designers and advisers and rarity elsewhere, BIOS will not mind its propagation through this forum.

There is a diagram illustration which I will include if there is demand but uploading images is difficult. . .

Best wishes

David P

Quote
A QUESTION OF TOUCH
[/font]
The return to mechanical action started just over 50 years ago in Denmark when, in 1929, Marcussen & Son built their first 'new' tracker organ. They had of course built many in the 19th Century, but since 1939 this firm has to my knowledge (with the exception of two instruments) built entirely with mechanical action, although many firms in England have yet to build consistently in this true and proper manner for the control of a musical instrument. In passing it is worth recording that Marcussen have developed many new ideas for actions without general credit being given to them, their major achievement being the floating beam which allows actions to compensate in varying ambient conditions. Many mistakes were made of course in the 'early' examples, but reliable solutions were already appearing in the 1940's in Denmark (e.g. the remarkable organ at Jagersborg, 1944) and Holland, with Germany building many mechanical organs in the 1950's but in most instances, without the reliability and finesse of the former. Unfortunately Germany proceeded down the road of 'assembling' organs from 'standard' trade parts, which was naturally understandable, as this had been, and still is, precisely the format for building electric action organs, i.e. buying sub-contractor's precision machined parts and assembling them. Also the great demand for new organs in an expanding economy was probably a contributory factor. However, when building organs in this manner, the only skill lies in utilizing such parts to the best of one's ability, which may not always be to the best advantage of the instrument.

After a generation of mechanical action construction one now sees in the work of the Danish and Dutch masters a consistency of design, which means that the solution arrived at by a particular builder is satisfactory for him because of his skill and capacity to make an action which is reliable under various atmospheric conditions both natural and enforced. Customer approval is shown in good order books. Soundness of overall design was witnessed in this country in 1965 with the installation of the Frobenius brothers' instrument at Queen's College, Oxford, an instrument that has been consistently reliable since well before the return to mechanical action began here in any serious way (indeed even to-day for the majority of English organ builders a tracker organ is a 'special' and not the norm).
Native builders of my generation were never taught to build mechanical actions — we have had to learn from scratch, although even this seemingly difficult situation has had its advantages, since now (1982) after 15 years of uoiistantly working with the medium, we know some of the reasons why and not just how. Many books (even now that very word is an anathema to at least one highly respected organ builder!) have had to be studied. Mathematics, Architecture, Music, Structures and the use of materials are but some of the cas for the serious student of our great craft, although fortunately for us, never before has so much information been available to the seeker. It would indeed be strange if all this material, coupled with the facility of instant communication throughout the world, did not influence our thinking, but notwithstanding this, a profitable area for study is provided by the principles lormalised by Sir Isaac Newton (1642-1727) in the 17th Century. His laws of notion are as applicable to organ actions as they are to landing a man on the noon. Newton's second law is perhaps the most useful for action design, slating that the force required to move a body equals the mass multiplied by he acceleration of that body. (F = m x a).

Earlier builders, who were conceivably not aware of the application of such useful information, nevertheless appear to have had an instinct about using the least effort to achieve work done, and perhaps this springs from the observation that most items were encouraged to move from their static state by animate muscle. There are exceptions to such observations, but broadly speaking all but the largest organs before 1800 had acceptable actions in terms of weight, since pressures were not excessive and the organs were in most cases of modest proportions; both these criteria assisting in obtaining a good touch, though without the builder necessarily understanding the reason why.

The feel of a mechanical organ is totally dependent upon the force required to unseat the pallet (an effect commonly known as 'pluck') and the mass of all the component parts and their accompanying frictional problems. Some of the larger organs were heavy whilst others were light, especially the French instruments with 'mécanique suspendu' and surprisingly, some of the South Germans organs with their detached consoles, such as the Holzhey at Rot an der Rot. Why were some of the larger organs at that time not excessively heavy to play compared with, say, organs in England of a hundred years ago? First, there are fewer eight-foot stops in earlier organs which therefore permits the use of smaller pallets. Another important point, though one that is often overlooked, is that with the addition of more eight foot stops and in some cases larger scaling, the windchest dimensions generally increased, in turn increasing the distance between the pallet centres and, without conscious thought, the Builders simply enlarged the width of the pallets to cover the vacant space. One still hears old — and not so old — organ-builders talking about giving the channel a 'good flush of wind' with a wide pallet; this is just not so, and the following calculation should convince all but the most obstinate.

A pallet 20 cm long with a movement of 0.7 cm allows 14 cm2 of air to pass, i.e. the area of a triangle is base x height:
Area = 20/2 x 0.7 + (20/2 x 0.7)(for 2nd side) = 14 cm2

This is true whether the pallet is wide or narrow, and for the above movement of 0.7 cm there is no point in making the pallet slot wider than 1.4 cm; suppose we did increase this slot to 2 cm then the increase in area for the flow of wind, keeping all other criteria, would only be 0.42 cm2. i.e. The depth of movement 0.7 cm takes up 1.4 cm of the slot width (0.7 + 0.7) cm which leaves only 0.6 cm x 0.7 cm (depth of touch) for extra wind at the front of the pallet which equals 0.42 cm2, which is an increase in the order of only 3%. Most important of all, the pluck is dependent upon the size of slot above the pallet and the wind pressure, and a simple formula for calculating the force required to unseat the pallet is as follows: Force = area x pressure. Assuming that the pallet and the slot are of identical dimensions, for reasons of simplicity, and using a pressure of 100 mm water gauge then together with the above figures:

Force = 20 x 1.4 x 10 cubic cms = 280 cm3

As 1 cubic cm (cc) of water (against which pressure is measured) weighs 1 gram the pluck would therefore be 280 grams, but since the pallet is advantageously a simple lever in itself the actual pluck is therefore reduced by one half, which demonstrates that a force of 140 gms would be necessary to unseat such a pallet. Using the dimensions for the second size of pallet above we obtain the following:

Force = 20 x 2 x 10

Pluck = 1/2 Force = 200 gms

This means that for 3% enlargement in area for the wind to pass an increase of 43% in effort is required, a point that even the great Victorian builders apparently did not comprehend. The pallet, slot and bar must be properly related to the demands of the pipework of each particular channel to achieve efficient results in terms of touch weight and sensitivity.

From the above simple calculations it will be observed that the wind pressure is a vital factor, and the force required to open the pallet is directly proportional to it. Thus using the dimensions from the first example above, but changing the pressure to 5 cms water gauge, a quite different force factor is obtained:

Force = 20 x 1.4 x 5 = 140 cm

Pluck = 1/2 Force = 70 gms.

The choice of pressure by the voicer, therefore, becomes a critical factor for the characteristic of the mechanism, and should not be decided upon without realisation of its effect.

A great advance in reducing pluck came with the introduction of the pallet board which enables modern builders to use small pallets and openings and yet keep large enough bar spaces proportionate to the requirements of air flow and pipe spacing. The insistence of some builders in constructing bar spaces as large as possible is another fallacy sometimes encountered (although this is as inexcusable as making them too small), they must be optimal for the maximum windflow necessary. When they are too large an undesirable 'burp' sound may sometimes be heard from the pipe due to shock waves in the bar; control of this airflow is as much part of the mechanism as any other component and must be as carefully computed and made as for all other parts.

Whilst entering the absorbing world of misunderstanding, some mention of the pallet shape and its effect upon flow and mechanical performance must be made. With adequate spacing its shape actually has very little effect on flow and none on pluck and it is, in fact, difficult to achieve any noticeable improvements when trying to streamline it to permit its efficient movement through the pressurized air, as one must remember that a wind pressure of 100 mm w.g. only represents an increase of 1% above atmospheric pressure.

The other important consideration for the feel of an instrument is the mechanism, which is the actual part that will convey the nuances of rhythm and timing of the most sensitive of musicians from his finger tips to the pipes, and should therefore be of such a quality that the player may solely concentrate on producing musical interpretation from the sounding parts of the instrument, without being conscious of the manipulation of a mechanism.

To obtain a lively action the component parts must be as light as possible, and this consistently proves to be an area for the individual genius of each organ builder to indulge in. Metal actions constructed mainly of aluminium were widely used in Germany in the last decades, in the belief that this material, with its low density to strength ratio and its predictable dimensional behaviour, was the way forward. However, very few builders use this medium now and wooden actions are again to be seen in abundance; I am sure that most thoughtful and progressive builders have arrived at this decision independently, from their own learning and observations and not simply just because everyone else is using that form of construction.


David Pinnegar

#1
Quote
The return to wooden actions has improved mechanical performance, since it is both light and strong as a material and, perhaps just as important, because it is wood, the builder has the tools readily at hand to fashion the medium without any sophisticated and expensive tooling. The importance of the builder being in control of as many parts of the instrument as possible cannot be overstated, particularly because he is not then subject to the whims of other industries. For example, where will all the specialised spare-parts come from in the future if the sophisticated units at present in commonplace use are superceded by different components? Valves, modern electronics and fibre optics have all been developed virtually between the cleanings of a well made mechanical action instrument.

A normal mechanical action has the following components: a key, backfall or two squares, a roller with two arms, a pallet and the necessary connecting tracker, to make up the mass of the moving parts, and each piece should be carefully designed by the Builder to be as strong as required but as light as possible to keep the inertia and kinetic energy factors as low as possible. One still sees actions made of Oak and Magogany by famous names, and can only assume that they are either satisfied by the results they produce, or are unaware of everyday information which could easily be used to advance their mechanisms. Compare the figures given below for a typical Victorian action and a modern one doing the same work; all dimensions are in cms.

Victorian action
Key 55 x 2.2 x 1.2 = 145.2 cm3 x density of Pine 0.4 = 58.08 units of mass
Backfall   35 x 1.8 x 0.5 = 31.5 cm3 x density of Mahogany 0.55 =   17.32 units of mass
Iron Roller   50 x iØ@ 0.15 thick = 23.56 cm3 x density of Iron 7.7 =   181.43 units of mass
2 Iron Roller Arms 5 x 0.40= 1 .25 cm3 x density of Iron 7.7   = 9.67 units of mass
Pallet 25 x 2.5 x 1 .8 = 112.5 cm3 x density of Pine 0.4 = 45.00 units of mass
Tracker   200 x 1.0 x 0.3 = 60 cm3 x density of Pine 0.4   = 24.00 units of mass
NB   Equivalent area/shape   Total:   335.50 units of mass

Modern Action

Key   50 x 1.5 x 1.2 = 90 cm3 x density of Pine 0.4   = 36.00 units of mass
2 Squares   2 [(5 x 0.8 x 0.4)+(4 x 0.8 x 0.4)] = 5.76 cm3 x density of Pine 0.4 =   2.30 units of mass
Al. Roller   50 x 0.8Ø@ 0.15 thick = 18.84 cm3 x density of Aluminium 2.7 =   50.89 units of mass
2 Al. Roller arms 4.5 x 0.3Ø(solid) = 0.63 cm3 x density of Aluminium 2.7 =   1.72 units of mass
Pallet 25 x 1.5 >< 1.5 = 56.25 cm3 x density of Pine 0.4 = 22.5 units of mass
Tracker   200 x 0.6 x 0.1 = 12 cm3 x density of Pine 0.4   = 4.8 units of mass
NB   Equivalent area/shape   Total:   118.21 units of mass


This demonstrates that by careful selection of materials and attention to dimensions the action weight may be drastically reduced, thus resulting in lighter touch, faster repetition and a better feel.
Victorian actions mostly worked on a one to one basis, i.e. tile pallet moved exactly the same distance as the head of the key, but in moderii act ions this is often not the case, for example the action at Hexham Abbey has a mechanical advantage in the order of 10:3 and is then levered up at the last moment to give the pallet sufficient opening. Such an action needs very careful construction and is more sensitive to ambient changes than those of a 1:
action, but there are significant advantages for the extra design constraints.

The following two examples will serve to demonstrate the substantial gains that are now obtainable almost without loss in other directions, for which we return to Newton and his simple formula of Force = Mass x Acceleration and our comparison of an action of a hundred years ago and one of to-day.

In terms of mass the Victorian action had 335.5 units and the modern action had 118.21 units which is a ratio of approximately 2.84:1 and in terms of distance travelled the former moves 1 cm and the latter 0.5 cm.

Victorian action Force = 2.84 x 1 = 2.84 units
Modern action Force = 1 x 0.5 = 0.5 units

An effective (but not actual) gain of this order has tremendous advantages to the player of a crisp and obedient mechanism, not to mention the relative and actual drop in kinetic energy, which gives many actions both old and new that disconcerting 'kick back' when you land in the key bed. It is worth noting that a tracker travelling a distance of 1 cm in a fast trill of 20 semiquavers per second has a speed of 360 metres per hour in either direction, and although this is not a vast speed any reduction improves the performance of the mechanism.

Once these simple principles are understood by the organ-builder, he will find that they will govern the whole approach to action design. With this awareness, the choice of a backfall, or two squares and a tracker, or a roller, to transmit the motion, becomes an exercise of totalling all the masses of the moving parts, together with their relative inertias, and deciding which is the most efficient system to permit the musician to control the attack, hold and release of each note.

After applying all the refinements and possibilities mentioned above, it is sometimes felt in the case of larger instruments (say 50 stops and over) that the action is still not light or responsive enough, although this depends upon individual taste — one organist will criticise an organ for being too heavy whilst another will say it is too light.
Various leverage devices are described in Audsley's 'The Art of Organ Building' Volume II, page 229 onwards, and study of these is greatly recommended to the modern builder. But beware! The pneumatic balancer Fig.CLXXVI, only works in one direction - once the note is on, it will remain that way until the wind is off rather inconvenient! Many modern builders use this type of pneumatic motor but with the difference that the motor is tubed to the windchannel of the given note, so that immediately equalisation of pressure between the well and note channel occurs, the pneumatic becomes ineffective. Other forms of assistor have been devised using pressurized pneumatics (Cavaillé-Coll for example) to pull open the pallet and again Audsley describes the main principles behind these mechanisms, but I feel that a study of these matters is outside this paper in which I have wanted to remain within the bounds of pure mechanical action which is efficient enough for the majority of organs.

Improved actions, beyond anything achieved in previous centuries, are now possible for the musician's use. It is high time, therefore, that a meeting was called by those in the teaching establishments to discuss, not whether an instrument should have electric or mechanical action, but what type of tracker action should be used. Everyone wants 'tracker feel' even with direct electric actions, but what is this tracker feel? Is it the type of action that 'fires' like a harpsichord, i.e. a pluck up to 100 gms then a very low holding weight in the region of 30 gms, or is it the type with low pluck feel and a sensation of squeezing a spring device, or a combination of all these criteria, or none of them at all? Where do organists require the pleno to be fully winded — 1 mm from the top, halfway down, or when the key reaches the bed? The key is not merely a make or break device; in this short distance of the depth of touch the organist must learn to express every mite of musical interpretation he possesses.

The future for the mechanical action organ in this country promises great instruments. The ground work has been prepared by the organs of the last decade and by the end of the next decade, organists should have as much subtle control over the sound producing parts of their instrument as a flautist.

[/size]