One objective in testing this method has been to try to determine the physical size limits of the test samples and still obtain consistent results in
measurement of the speed of sound. This method has been tried out recently on another thread 'Vegetable oil on soundboard' where the opportunity was
taken - by successively reducing the cross section of a test sample - to audio test much smaller section samples. Here are the results for speed of
sound measured longitudinally for spruce:
#1 - sample 399 mm X 17 mm X 8.2 mm - average SOS = 5430 m/s
#2 - sample 399 mm X 17 mm X 2.5 mm - average SOS = 5385 m/s
#3 - sample 399 mm X 17 mm X 1.5 mm - average SOS = 5366 m/s
#4 - sample 399 mm X 12 mm X 1.5 mm - average SOS = 5352 m/s
#5 - sample 399 mm X 04 mm X 1.5 mm - average SOS = 5384 m/s
The average values for speed of sound in each case were determined from the first four harmonic resonance peaks shown in the Audacity frequency
spectrum analysis - there being four clear frequency spikes in each case. As can be seen all of the results are within a maximum value 1.4% of each
other. Good enough for comparative tests between different sound board materials if not for determination of absolute values of SOS, stiffness and
Sound Radiation Coefficient.
The slight variation in speed of sound between samples may be partly due to localised physical differences such as grain direction, grain width.
The significance of this is that even relatively small samples cut from thin sound board blanks can yield reasonable results as long as the sample is
rod like - long and narrow. A length of 40 cm may be about the minimum length for best results.
From a practical perspective, the small size of sample requires that it must be prevented from being driven towards the microphone by successive
impacts of the tiny hammer. I used a piece of masking tape folded in half (acting like double sided adhesive tape) to lightly restrain each sample on
one of its rubber support pads.
![[*]](images/xmbprobrown/default_icon.gif){kind=icon}
