Description of the New Piano Action
by Clarence N. Hickman
Ampico Research Laboratory

Dr. Hickman probably wrote this 14-page technical paper shortly before he left Ampico in January 1930 to join the acoustical department of the Bell Telephone Laboratories.  The new grand piano action was his "baby", left largely in the care of the Knabe Piano Company who were planning to produce it.


    The present type grand action, when properly regulated and in first class condition, gives very good satisfaction. However, it is well known that the action will not remain in first class condition.

    The Research Laboratory made several attempts to cure the most objectionable features pertaining to this action.

    The pin joint (a) (See figure 2 showing old type action) is a source of much annoyance.  Under certain climatic conditions the joint becomes loose, and the shank rattles, producing a very objectionable noise. Under other climatic conditions it swells up and becomes so tight that in many instances the hammer will not return to its normal position of rest.

    An attempt was made to use a composition of Bakelite and Graphite for bushings, but met with failure due to the fact that when the hammer strikes the string, there is considerable vibration set up in the shank 10. This vibration causes a rattle at the pin joint "a", which produces noise enough to be heard over the entire room. Even with as small a clearance as .0002", there is an objectionable amount of noise.  This noise is very much like that produced when the hammer is loose on its shank.

    The swelling and packing of the pads (1)(2)(3)(4)(5)(6) etc., give no end of trouble. If pads are used which are sufficiently firm so that they do not change in dimension greatly, noises will result which are unbearable. Therefore, after several attempts the Research Laboratory decided that it could eliminate those difficulties, which occur in the old action, more easily by designing one which inherently is free from the objectionable feature found in the old.

    This paper will, therefore, deal principally with the mechanical superiority of the new action rather than with the artist's opinion of it.

    It is, of course, realized that unless the touch is favorable to the artist, it would be unwise to place it on the market. However, most all of the artists that have played on the new action are very much pleased with it, and some of them are thrilled with its response. The great value of the action, however, lies not in the fact that an artist notices that it is better, but rather in the fact that it may always be depended upon to remain in adjustment.

    For convenience in referring to the sketches and curves, they are bound separately.

oldAction.jpg (99 kb)
Figure 2.  Old action.

    Sketch No. 1 shows the type of action that was developed. It is believed that the description and comparisons contained in this paper will justify the claims which have been made; that the new action is greatly superior to the old.

    The mechanical features of both actions will be compared from several points of view.

    There is such a close connection between some of the sub-topics, that frequent repetitions may occur.


    In the old action (See Figure 2) friction is present at the following places:

    As the jack member "T" pushes against the nubbin (2), the nubbin scrapes on the end of the jack "T", producing a great deal of friction.

    When the surface "J" of the jack comes in contact with the surface "C" of the pad (3), the upper end of the jack "T" is pulled out from underneath the nubbin(2) with a great deal of friction.

    At B the capstan, as it pushes up against the pad (1), not only rolls but actually scrapes against this pad with an appreciable amount of friction.

    At the same time the jack "T" is being pulled out from underneath nubbin (2), the surface "J" rubs along the surface "C" an appreciable amount, producing more friction.

    When the pad (6) comes in contact with the head of the screw 2, there is some friction due to scraping at this point.

    In order to have the hammer firmly held in alignment, the pin joint "a" must be bushed reasonably snug. This joint produces considerable friction when the hammer is moved up or down. This friction is sometimes so great in humid weather that the weight of the hammer is not sufficient to overcome it, so that the hammer remains suspended in mid-air.

    There is also friction at the pin joints b, c, and d. These do not usually cause a great deal of trouble, with the exception of the joint c. Often the friction here becomes so great that the spring 8 will not return the jack to normal position, making it impossible to give the hammer sufficient velocity to reach the string.

newAction.jpg (109 kb)
Figure 1 (Sketch 1).  New action.

    In the new action (See figure 1) the substitution of pin joints for contact joints has eliminated most of the friction.  The type of construction (see figure 3) of the pin joints A1, A2, B1, B2 and C reduces the friction here to a very small quantity. The pin joint "C" is constructed of Bakelite impregnated with Graphite so that what was the worst joint on the old action, becomes the best on the new action. This type of joint cannot be used in the old action because the sting (vibration) of the shank would produce a noise. In the new action the sting of the shank is taken up by the cloth bushing A2 so that it is difficult to pass this vibration through the driving member "d" to the pin joint "C".  Of course, it is obvious that the type of construction for the pin joint "C" is such that climatic conditions have no effect on it whatsoever.

    The pin joint A1 is similar to the one used in the regular piano action, but may be freer. Therefore, there is less danger of friction at this point

    The pin joints B1 and B2 are quite free so that, friction at these joints is almost negligible.

    The pin joint A2 is bushed very tight in order to eliminate any noise that might develop due to the vibration or sting of the shank as the hammer hits the string.  This friction does no harm since it occurs only at the time the hammer is about 1/32" away from the string.  It is not felt in the fingers at all, since the driving member is stopped completely and the hammer swings on against this friction. It is apparent, of course, that when the hammer has only 1/3w" to move, before contacting with the string, that the motion at the pin joint A2 is almost negligible. So nearly negligible is this motion that if the bushing were cemented to the pin, the give of the bushing cloth alone would be sufficient to permit the hammer to strike the string firmly.

    It is, therefore, apparent that this pin joint, while it does not need to be bushed tightly, should be, in order to avoid any noise that might develop, such as frequently occurs with the old action.

RA1.jpg (46 kb)RB1.jpg (45 kb)
Curve 1.                                                           Curve 2.

   A comparison between the friction of the two types of actions may best be shown by the accompanying curves. These curves were obtained by an instrument designed and constructed in the Research Laboratory. The curves are in graphical form and the upper one represents the force on the key to depress it throughout the descent of the key. The lower curve in each case represents the force on the key as it rises to its normal position. The pressure is read in ounces to the left of the graph. The depression of the key is read in 64ths of an inch at the bottom of the graph.

    Curve #l is a record obtained with note 50 of a Chickering action and represents the best performance that it is possible to obtain with this type action. The action was in perfect regulation. By following the upper curve from the left we see that when the pressure reaches 2.2 ounces the key begins to descend without much increase in pressure.  When the key descends to 8/64" it begins to pick up the damper so that more pressure is required. When the pressure has been increased to 3.4 ounces the key again descends without much increase in pressure until it reaches 17/64" depression. At this point the repetition lever "U" contacts with stop screw 2 (See figure 2) so that more pressure is required to depress the key. The pressure rises on a rather steep slope to 5 ounces.  In the meantime, the key has been depressed to about 21/64". At this point the jack arm "J" contacts with the stop pad (3)(see figure 2). More pressure is now required to pull the jack "T" out from under the nubbin (2). The pressure curve now rises still more rapidly. Part of this increase in pressure is used to overcome the repetition spring which is still becoming stronger. When a pressure of 7 ounces has been reached and the key has been depressed 22/64", the jack jumps out from under the nubbin and the key quickly descends to the bottom, coming to rest on the felt punching at the front of the key. Any further pressure will not move the key further except for the compression of the felt punching.

    If the return curve below is followed (Curve 1) it will be seen that after the pressure has been reduced to 3 ounces, the key begins to rise with without further changes in pressure until it is only 18/64" depressed. At this point the repetition arm leaves the head of screw. (See figure 2). As the pressure is decreased the key rises to a point of 9/64" deflection. At this point the damper is gradually released so that it is necessary to reduce the pressure to about 1.5 ounces to get the key to rise. The pressure must eventually be reduced to zero before the key rises to its normal position.

    Now the distance between the upper and lower curves at any point represents the frictional forces existing at that point. The total area between the two curves represents the work wasted in overcoming these frictions. In other words, if there were no friction in the action, the two curves would lie on top of each other.

    Curve No. 2 represents a similar record for the new action. In this record the damper is picked up after the key is deflected 6/64". The screw 3 begins to break the knuckle after a deflection of l9/64" is reached. (See figure 1.) The stop screw 2 comes against pad "p" when a depression of 22/64" is reached.

    Comparing curves 1 and 2, it will readily be seen that not only are the forces for curve 2 smaller, but the frictions are much reduced. The total area between the curves #2 is not more than 1/3 that in curve #1. It should also be remembered that Record #l was obtained from an action under ideal conditions.

RH.jpg (46 kb)RC.jpg (46 kb)
Curve 7.                                                                  Curve 8.

    Looking at curves 7 and 8 we see how climatic changes may increase the friction of an action. Curve 7 is a record of note 52 on a Chickering piano which was in the home of President Urquhart on Long Island. This piano was brought to the Research Laboratory on May 23rd, 1928 and record 7 obtained the following day. The piano had been left completely closed from the time it arrived until the record was taken. Curve 7 shows that, had there been 1/2 ounce more friction, the hammer would not have dropped at all. Observing the top of the curve it will be seen that 10 oz. was not sufficient to pull the jack out from under the nubbin. A one-ounce weight was placed on the key and then at 10-3/4 ounces the jack was released. The area between the top and bottom curve represents the work to overcome friction, and this is easily 20 times that of curve 2.

    Another record was taken of the same action 24 hours later. This record is shown in graph 8. The piano was left open for 24 hours and the parts had dried sufficiently to make the improvements shown in curve 8.

    Some of the action units in this piano had been worse than the ones shown in curve 7. For when the piano was in Mr. Urquhart's home some of the keys would not return to their normal position.

    It may be said that the conditions noted here are not peculiar to a Chickering action. They all do this same thing under similar climatic conditions. The Repair Department is called upon to fix all makes of actions. The number of complaints received each summer due to sticking actions is very great. It is well known that only a small number of such cases are reported.


    In the old action a change in the dimension of the thickness of the pad (1) affects the position of the hammer by a factor of 10 to 1.  That is, if this pad packs 1/64" the hammer will drop 10/64".

    If the nubbin(2) packs, the position of the hammer will be changed by a factor of 5 to 1. That is, if the nubbin becomes 1/32" thinner, the hammer will drop 5/32".

    As a matter of fact, due to changes from those two pads alone, the position of the hammer, in actions that have been used for some time, is such that it is resting on the rail 7, instead of on its nubbin (2). In some actions these pads have become so packed that it is impossible to bring the hammer closer than 1/2" from the string as the key is slowly depressed.  This, of course, means that the artist has no control over the velocity of the hammer, or the intensity of its blow after it reaches a point 1/2" from the string.

    Under such conditions, it would be impossible for him to play on this type of action with any degree of precision.

    The pads (3)(4)(5) and(6) may swell or pack causing serious results in the operation of the action.

    For instance, if the pad 3 packs sufficiently, it will be impossible to remove the jack "T" from underneath the nubbin(2). Consequently, the hammer will block against the string. On the other hand if it should swell the let-off would occur too early.

    If the pad (4) packs, the jack "T" will rest so far under the nubbin 2 that there will be a great deal of energy used in rolling it out from underneath the nubbin. On the other hand, if the pad (4) swells, as it may in damp weather, the jack "T" may not get sufficiently under the nubbin(2) to deliver any power to the hammer, but will slip out on fortissimo blows without moving the hammer at all. This trouble is quite common to actions.

    If the pad(5) packs, the repetition bar "U" will push the nubbin 2 away from the jack "T" so that there is lost motion and noise will result. On the other hand, if this pad (5) swells, the repetition arm "U" will recede from the nubbin(2) leaving the jack "T" projecting out beyond the repetition arm "U". Under these conditions, the jack "T" will not return underneath the nubbin (2) and consequently no force will be delivered to the hammer.

    The swelling and shrinking of the pad ("6") affects only the drop-back of the hammer after the let-off. No great difficulty is experienced at this point. The pad usually packs slightly and the second or third regulations correct this packing so that further difficulty at this point is usually quite small.

    In the new action (See figure 1) if the pad "k" packs, this affects the position of the hammer by about 4 to 1.  That is, if the pad "k" packs 1/64", the hammer will drop 4/64". This is the only pad which affects the position of the hammer when at rest, and it will be seen that by comparing this effect with those that may occur in the old action, that this is almost negligible.

    If the pad "p" swells or packs it changes the position at which the driving member is stopped and consequently the position of the hammer at the time the driving member stops by a factor of 3 to 1. If the pad "p" swells 1/64", the hammer will stop 3/64" further away from the string than when it was first regulated. This change is less than the normal distance of regulation with the old action, so that the effect is not serious.

    If the pad "q" swells or packs, it slightly affects the tine at which the knuckle breaks. This is so negligible that it can do no possible harm.

    If the pad "n" swells or shrinks, it merely changes the position of the pin joint A1 with respect to the pin joints B1 and B2. As long as it does not swell sufficiently to cause the pin joint A1 to come in line, or to the right of a line connecting pin joints B1 and B2, no harm whatsoever has been done. The factor of safety called for in the manufacture locates the pin joint A1 to the left of the line joining B1 and B2, a sufficient distance to guarantee that the swelling of the pad "m" can never produce any trouble.

    The swelling of the pad 1 has no effect whatsoever on the action. It is merely placed there to set as a bumper and eliminate any possible noise that might develop.

    In view of the small changes that may occur with the new action, it is believed that one regulation will be sufficient instead of the many made on each old action.


    In the old action, adjustments are made at the following points:

    Capstan 1
    Dropback screw 2,
    Let-off screw 3,
    Repetition spring 4
    Jack stop screw 5
    Repetition lever screw 6,
    Repetition lever screw 7,
    Fly spring 9

    The most important adjustments are 1, 2, 3 and 4. It is found necessary to readjust these frequently due to swelling and packing of the pads enumerated in the preceding discussion.

    In the new action, we have adjustments of

    Capstan 1
    Driving member stop 2
    Knuckle breaker 3,
    Repetition spring 4.

    Here we have 4 adjustments in contrast to 8 on the old action and it should be remembered that the driving stop 2 is an extra adjustment which has been added to obtain a better quality of tone, and to obviate the necessity of a drop back of the hammer. For best method of making these adjustments see the article on Adjusting New Action at close of this paper.

    These adjustments when once made will remain sufficiently correct almost indefinitely.  It is only the filing of the hammer or the packing of the felt "k" that will affect any of these adjustments.

    As a matter of fact a life test was made with this action and with the old action over a period of time sufficient to correspond to 35 years of use. A disinterested regulator inspected both actions from once to twice each week, and made such changes in the adjustment as were necessary to keep the actions in perfect adjustment, and at the close of the test period the following facts were revealed: FIRST: on practically every visit of the regulating man at least some adjustment was changed on the old action.  Not a single adjustment was made on the new action, and at the end of the test it was in perfect adjustment. The capstan on the old action had been screwed up considerably; the let-off screw had been changed; the repetition spring had been changed; the jack stop had been changed, and the repetition lever screw had been changed, and even then, at the end of the test, the hammer was resting on the rail; the let-off was about 3/32" late, and the drop-back was only about l/64".


    In the case of the old action, the capstan adjustment No. 1 will have to be changed from  time to time depending upon the swelling and shrinking of the pad 1 and nubbin (2).

    It will be necessary to change adjustment 3 due to swelling or packing of pad(3).

    It will be necessary to readjust screw 2 due to swelling or packing of pad (6)

    It will be necessary to readjust the repetition spring 4, due to the fact that the pin joint "a" may become tighter or looser.

    It may be necessary to readjust screw 5 due to swelling or packing of pad (4).

    It may be necessary to readjust screw 6 due to swelling or packing of pad (5).

    Screw 7, if properly set at the factory, seldom requires readjustment.

    Spring o may have to be readjusted if the pin joint "c" changes in tightness.

    Those difficulties have already been discussed at length.

    In the new action the stop screw 2 will only have to be readjusted in case of filing the hammer or packing or swelling of the pad "p". Any change due to the latter is almost negligible.

    The knuckle breaking screw 3 will only have to be readjusted in case the pad "q" packs or swells.  Any change in the thickness of this pad is so small that it is almost negligible.

    Any readjustment of the spring 4 will be due to change in the friction in the pin joint "C", and the type of construction of this pin joint insures that this change will be almost negligible.


    Dry or humid weather affects the old action in the following manner:

    The swelling or packing of pads (1,2,3,4,5,6) and (15).  Also the Swelling and shrinking of pin joints "a", "b", "o" and "d".

    The effect of change in dimension of pads (1) and (2) is often quite serious
as has already been shown.

    It has also been explained that the swelling of pin joints "a" and "c"
cause a great deal of trouble.

    The curves 7 and 8 showing the touch of a Chickering piano action, which had been subjected to the humid weather of Long Island, shows the manner in which the friction of the pin joint "a" affects the action.  (These curves are discussed on pages 5, 6 and 7.)

    It will be seen by observing curve #1 that if the friction had been one-half ounce greater, the hammer would not have returned to its position of rest but would have remained suspended near the string.  This often happens and accounts for the fact that the keys do not come back to their normal position in damp weather.

    If pad (3) swells, the let-off will occur too early. This difficulty is not great on actions using printers cloth for pad (3), but often causes trouble where felt is used.

    Pad (4) often swells and pushes the jack "T" out from under the nubbin (2). If printers cloth could be used this difficulty would be less, but it is difficult to prevent noise with this cloth.

    Out of about 8 pianos at the Research Laboratory and Editing Department the one having the new action was the only one that did not show trouble of this nature during the past summer.

    Pad (5) frequently swells enough to prevent the jack "T" from getting back under the nubbin (2). This difficulty has been experienced on the recording piano.

    The swelling of pad (6) may slightly change the drop-back, but this is not serious.

    The swelling of pad (15) has the same effect as that of pad (5).

    The swelling of pin joints "b" and "d" seldom are enough to cause serious trouble, however, the Chickering factory has shown that, next to joint "a", joint "b" introduces the most friction.  There have been cases where this joint was so tight that the weight of the wippen was not sufficient to overcome the friction at this point.

    In the new action, climatic changes may slightly affect pads "k", "l" "m", "p" and "q".  However, none of these are appreciable and it would take a delicate measuring device to know that there really was a change.

    Climatic effects on pin joints A1 A2 B1 and B2 are quite small.

    B1 and B2 are quite loosely pinned so that no difficulty is experienced here.

    A2 is quite tight, and if it becomes tighter still, no harm results, as explained on page 4.

    In the case of pin joint A1, this cannot be too loose, otherwise the result
will be noise.  Therefore, this joint is the only one that might logically be expected to cause trouble.  It, however, may be pinned considerably looser than the hammer shank in the old action so that it should never give any trouble.

    The seriousness of swelling joints, etc. may be realized when it is learned that in some localities the piano is considered useless during certain months of the year.  Pin joints and action parts are often treated to make them moisture-proof. It is claimed by some action-makers that the treatment does more harm than good.

    The Research Laboratory has constructed a testing tank for the purpose of measuring the friction of action joints under all humidities, from zero up to 100%. All types of joints, treated and untreated will be tested.  It is hoped that these tests will clear up some of the uncertainty regarding treated joints. It will also show the effect of the most severe climatic conditions on the new action.


    The new action contains only about half as many pieces as the old. Simplicity of construction usually is an indication of freedom from trouble.


A great deal of the inertia existing in the old action has been eliminated.  Therefore, noise is greatly reduced, especially in the return of the hammer and the wippen.

    The usual thud which occurs as the key returns to its normal position is practically eliminated with the new action.

    The nubbin on the old action, which has been a source of difficulty due to noise in case it was too hard, and lack of power if it was too soft - has been completely eliminated.

    The noise of the jack coming against its stop and of being jerked out from underneath the nubbin, has been eliminated.

    The noise of the jack coming against the let-off stop has also been eliminated.

    The only place for noise in the new action is the stop screw 2 coming against the pad "p"; the stop screw 3 coming against pad "q", and the jaws of the knuckle at "m" coming together when the key is released.

    On the whole, the noise of the new action is greatly reduced over that of the old action.

    Attention is now called to a bad noise noticeable in the Ampico which
has been completely eliminated by the new action.

    The noise is often thought to be caused by lost motion in the Ampico action. However, it is really in the old action.  It is a nubbin noise present in damp weather.  It occurs only on repeated notes, i.e. the repetition of notes that have been sustained by extended perforations.  It is caused by the swelling of pin joint "a" which makes the return of the hammer from the back-check position more sluggish than the return of the wippen.  If the note is quickly repeated, the hammer will not be all the way down so that there is lost motion between the jack and the nubbin.

This noise was quite pronounced in Mr. Herman Spain's piano during the past summer.


    The tone of the piano is improved by using the new action.  This is brought about by the fact that --

    The driving member "d" is stopped completely, so that the hammer flies on and attacks the string without any other member having any effect upon it:

    The length of the shank is less, so that there is less wobble to the hammer; this prevents bad tones;

    The decrease in the length of the shank makes it possible to reduce the weight of the shank, thereby bringing the center of percussion closer to the center of the hammer.  This insures a quicker rebound of the hammer and a clear bell-like tone.  This improvement in tone is especially noticeable in pianissimo passages.

    The tone is further improved by the fact that there is less noise present.

    Another improvement in tone, obtained by an artist, is due to the fact that he has better control of dynamic.  This better control is due to the fact that he is permitted to keep control of the hammer velocity until it practically touches the string. It is impossible to adjust the old action for this degree of control without greatly increasing the probability of the hammers blocking or being forced against the string - thus killing the tone.


    The touch of the action is delightful to most artists. There is complete absence of sluggishness or sticky feeling.

    The actual weight of the action may be adjusted to suit the artists taste.

    Without a single exception every artist who has tried the new action agrees that he has much better control of the intensity with which he plays and is able to play with precision in pianissimo passages.  The repetition is decidedly improved; some of our artists remarking that it was uncanny the repetition that could be obtained with such a light action.  (In this case the action was adjusted for a very light touch.)

    All of the keys are weighted exactly alike; the difference in weight between the heavy hammers and the light ones being compensated by the manner in which the repetition spring is applied to the driving member.

    The down pressure of the repetition spring on the key remains approximately the same from one end of the piano to the other.

    In the old action it was necessary to increase the tension of the repetition spring in order to compensate for the heavy hammers.  This increases the pressure required by the hand in order to depress the key.

    The accompanying curves 3, 4, 5 and 6 show the effect of changing the hammer on each type of action.  Curves 3 and 5 are for the old action with hammers 58 and 1 respectively.  Curves 4 and 6 are for the new action with hammers 57 and [??] respectively.  Here it is easily seen that increasing the hammer weight for the action raises the maximum force from 5.7 ounces to 7.4 ounces; whereas for the new action the force is changed from only 3.7 ounces to 4.2 ounces.

    The direct manner in which the power is delivered to the hammer by the key, insures a big fortissimo with little effort. There is very little wasted energy in the transmitting members.  In the old action it is necessary to go around through a double set of levers and apply the power through scraping joints so that a great deal of the energy is lost.

    Without exception, all of the artists who have played on the new action have remarked about the ease with which fortissimo passages are executed.


    As indicated in a preceding paragraph, a life test was made on the two types of actions, extending over a period of time equivalent to 35 years of use in the average home.

    This test consisted in striking a note ten (10) times per second, nine (9) hours per day, for a period of a month.  A very conservative estimate indicated that this was equivalent to 35 years use in the average home.

    When this test was ended, the same parts of the new action were subjected to the same test using new parts for the old action.  In both tests the pin joint "a" of the old action was worn so badly that the hammer had a wobble of over 3/16".  This, however, was the only seriously worn pin joint in the old action. In one case the buckskin of nubbin(2)was worn through; in the other it held up reasonably well.  The hammer, in all cases, was badly compressed and worn and the back checks seriously worn.

    All of the pin joints in the new action stood up perfectly.
    The pads (1), (2), (3), (4), (5), (6) and (15) were badly compressed, so that it was necessary to frequently change the adjustments in order to keep the action in proper condition; whereas with the new action not a single adjustment was changed, and at the end of the test the action was still in perfect adjustment.


    As has already been explained, there is very little readjustment of the new action to be expected. There are some distinct advantages with the new action in case of other repairs. For example, if it is necessary to replace an ivory that has broken, this may be done without removing the entire action, as is necessary with the old one.  It is only necessary to remove the member "c" (which is held in place 5 screws) disengage the clip "g" from the capstan, and remove the individual key which is to be repaired.

    Also, in case it is necessary to re-level a set of keys, this may be done without removing the action.  Remove the key rail and lift the keys up clear of the balance rail pin and change the punching as desired.  To do this, it is only necessary to break the knuckle joint A1 and the key may be lifted clear of the balance rail pin.

    In the old action, it is necessary to either remove the complete action, or at least remove the jack stop rail.


19 February 2002