I was asked to have a go at fixing an old Paquet metronome.
am loath to hack into something that may be an antique but this device
had obviously been dropped as the cover had a glued split and the lower
front panel had also been re glued.
The complaint was that the beat was uneven at low settings.
using Audacity (see below) showed that set to 60 the rate was 60.72 but
the ratio of the short beat to the next was 1 to 1.35.
adjustment and the "corrections" shown below the result was much
better. I timed 180 beat periods with a stop watch at 180.4
seconds. This compared favourably with a cheap AU$17 electronic unit
that gave 178.61 seconds in a similar test.
beat now seemed perfectly even to me. The average ratio of the shorter
beat in a pair to the next from 4
random measurements was 1 to 1.025 and the worst 1 to 1.043. By
comparison the electronic unit when measured had perfectly even beats
and also a superior ease of setting and range of rythms, pity its
beat rate was not more accurate.
As received, with the weight near the top the swing was obviously off to one side.
spring was let to run down and I found that with the weight near the
top, the pendulum came to rest about 15 degrees off centre.
This is a picture of the metronome in an unwound condition after I worked on it.
Note that the pendulum is unlatched with the adjustment weight at the top and it is now perfectly vertical.
Any unbalance will be exaggerated the further up the rod the weight is.
As the unit had evidence that it had been dropped I supposed that the lower weight had somehow shifted.
To balance the pendulum some solder was added to one side of the weight.
Increasing the weight increases the frequency so an amount of the weight was rasped off the other side.
The aim was to get a balanced pendulum of the correct weight for the rhythm to match the printed scale.
A level platform was constructed and a microphone attached to my computer.
any tilt will effect the evenness of the beats. It is possible to
"adjust" a slightly uneven metronome by packing one side up.
Audacity is a free programme that records sound, lets you view the wave form and make accurate measurements of time intervals.
Audacity being used to measure time interval between beats.
The clack noise is produced by the escapement wheel peg hitting one detent or the other.
The impulse is transmitted to the case via a metal insert which the pivot shaft end rests against.
Even with a balanced pendulum the beat was uneven.
My aim was to get the escape to occur symmetrically in the swing.
pivot shaft can be adjusted via the three screws for equal
got the escape timing adjusted to within 3% at a rate above 60 beats per minute
but at slower speeds the beat was uneven and when the spring was more fully
wound the pendulum slowed to the point where it stopped.
At this point I removed the mechanism from its case and went hunting for the source of friction.
The nose of the shaft had worn a dent in the anvil that it pounds against.
A section of a snap off knife blade was glued on to the anvil.
The shaft nose had worn flat so it was rounded off.
This is the heart of the escapement mechanism and the likely source of the friction,
There is a chip out of one face which from the wear marks is perhaps not involved in the sliding path.
the pendulum shaft rotates the escapement pin is pressed against
the face of one detent or the other.The friction of this contact
increases when the spring is more tightly wound and has a bigger effect
when the pendulum is slowly moving.
The detent faces get a hard pounding as the escapement pins hit them.
To make a complete repair some resurfacing of these worn faces would be necessary.
I decided to reassemble the unit, oil it and see if my efforts so far had improved things.
while giving satisfactory results (within about 3%) above about 60 BPM
at a slower tempo performance deteriorates.
However it is still usable at 40 BPM provided it is not tightly wound.
Note I have no expertise in this area and the above is just a report of my attempts at repair.
The following is the advice I gave to a reader who had a metronome with irregular beat.
Imagine a perfectly working metronome. The
click would occur at the centre of the swing when going in either
direction and the beat would be even for all positions of the
Consider the conditions necessary for a this to occur. A
free swinging simple pendulum will swing side to side an equal
distance and time around the position that it comes to rest. A metronome
beat is adjustable by sliding a weight and for the beat to be even
at all positions of the weight the rest position must not change when
the adjustment weight is moved. The only way for this to occur is for the rest
position of the pendulum to be vertical for all positions of the weight.
Checking the rest position. The
pendulum is driven (pushed) by from side to side by spring force acting
via the angled face of the escapement. To find its free rest position
this force must be removed. One can let the spring completely run down
or remove the spring tension by applying winding force to the key. The
pendulum will be almost certainly be near vertical with the adjustable
weight at the bottom but it may not be with the weight at the top.
that the pendulum rest position is vertical regardless of the
adjustable weights position (which means that the amount of free swing
will be equal about the vertical) all that is necessary is for the
"click" to be centred. The "click" may occur in different positions
depending on the direction of the swing but should be adjusted to be
symmetrical about the centre. This adjustment is made by moving the
pendulum support shaft side to side while keeping the shaft parallel to
the escapement wheel (adjust front and back support brackets).
adjustments to the evenness of beat perhaps can be made by adjusting
the pendulum support shaft so that it is not parallel to the escapement
wheel. If the shaft is not parallel to the escapement wheel the amount
of "push" that is given to the pendulum will be different in each
direction of swing. This adjustment can correct for slight errors in
the angle of the escapement faces and may do something to correct for
an unbalanced pendulum.
Balancing the pendulum. To balance
the pendulum it is necessary to move some weight around without
changing the tempo of the metronome. To do this the mass must be kept
the same and the centre of mass must be kept the same distance from the
Note. On my metronome I could see no evidence that the
lead bob was adjusted by filling or adding lead in the factory and the
slow tempo pendulum beat was way uneven. This leads me to speculate
that the unit was dropped and the shaft was bent but that was not
obvious. Bending the shaft to correct the balance has the advantage of
keeping the weight the same and not changing the distance from the
pivot but I did not attempt it.