Paper vs Kevlar


This is just an experiment to study the differences between two similar size speakers with different cone materials, one being paper and the other woven Kevlar, both shown below and humorously referred to as the usual suspects.  The Heppner is a general purpose use speaker intended for use in small portable record players, table radios and portable radios when portable radios were large enough to accommodate a 5" speaker.  Today we use earbuds. The Focal was originally intended for car use, hence the weather resistant Kevlar cone but serves extremely well as a mid-range.  


The Usual Suspects



The T/S parameters were determined, for all they may be worth and are listed here.  Xmax was determined by driving the speaker at 100hz and measuring the response with a microphone 1m away.  The mic was connected to an oscilloscope and a THD analyzer and Xmax was determined when the waveform showed the slightest deformation at its peak which was shown to be 1%.  The SPL for the Heppner and Focal were 81dB and 87dB, resp.

The tests were made with the speakers in a 300 in^3 sealed box stuffed with pure Australian wool, washed but otherwise untreated. 

78C219-2 575919 Heppner


Fs 61.52 Hz Re 7.50 Ω Sd 86.5901 cm²

Qms 1.72 Qes 0.37 Qts 0.30

Cms 2.9742 mm/N Mms 2.2500 g Rms 0.51 Ωm

Bl 4.20 N/A dBspl 94.94 VAS 31.1395 L

Zmin 8.36 Ω L1kHz 0.26 mH L10kHz 0.22 mH

Xmax 1.5mm


5K013L Kevlar cone Focal


Fs 55.66 Hz Re 6.90 Ω Sd 83.3229 cm²

Qms 3.21 Qes 0.39 Qts 0.35

Cms 1.2319 mm/N Mms 6.6359 g Rms 0.72 Ωm

Bl 6.41 N/A dBspl 89.25 VAS 11.9431 L

Zmin 7.51 Ω L1kHz 0.59 mH L10kHz 0.34 mH

Xmax 3mm




The impedances of the speakers are shown here in free are (left) and in the box (right).  The red traces are the Focal and the green traces are the Heppner.  In the following tests, both speakers were driven with the same voltage applied at 900hz, where the impedances are equal and thus, the power transfer.  At other frequencies, the power transfer would differ between each speaker.

FIG. 2


FIG. 3

Frequency response measured at 1w1m, red=Focal; green=Heppner.  The peak centered at 5500hz is quite prominent and very audible in a subsequent listening test.  The tonal quality is also quite different even after equalizing the peak.  This will be shown later in FFT analyses and spectrographs.



Figs. 4, 5 and 6 FFT (fast Fourier transform) analyses of pink noise.  The power level is unimportant but it was around 1 watt and measured at a distance of about 5 feet in my living room.

The speakers are 5 1/2 feet apart and tilted inward about 15 degrees to allow their points of intersection to meet slightly left and right of the mic 64" distant. While this arrangement hardly reflects an ideal listening position, it's intended for comparison testing only. The mic, therefore, is not directly on axis which will cause a drop in high frequency response. Also, keep in mind that neither of these speakers spec out at performing above 5khz to 6khz.


The left photo was taken from behind the mic.  The speakers are bi-amped and the pair onto which they are seated are disconnected. The woofers under the end tables are active.



This shows the pink noise FFT of (Focal) green/right channel and (Heppner) red/left channel.  Somehow the colors got reversed.



This figure has the same green and red traces of Fig. 4 with the addition of a grey(black) one which is both the Focal and Heppner active, both channels.  The grey trace is slightly different than either of the other two due to interferences between the radiation patterns of each speaker.  Note the exceptionally flatter part between 4khz and 10khz.  If  I'm correct, this shows an almost 180 degree phase difference between the two.  Another of smaller degree can be seen at 2khz..  Now, look at the peak at 80hz.  the red and green seem to be in phase and sum to the higher grey. In areas where the grey is between the red and green, it is thought that the phase shift can be anywhere between 0 and 90 degrees.

The above conclusions can also be affected by the location of the speakers, despite their equal distances to the microphone.  Consider the Focal on the right.  It is closer to the drapes and is opposite the wall behind the electronics.  The Heppner on the left has no wall on the left; it's an archway opening into the dining room and it's looking straight into the hallway. 



Both channels active and driven with pink noise.  Grey, without equalization and light green with equalization.  While the resemblance above 250hz is much the same within 4dB, the sonic difference is quite large.  The 4dB difference is not an oversight in using the output level controls on the equalizers.  It was left as such due to the similarity of the traces which would have been super-imposed and difficult to differentiate.  Increasing or decreasing the level of one of a pair of similar traces by as much as 10dB is common for clarification.  This, if done, is stated in the graph.






Unfortunately, the spectrographs had to be resized to fit across the screen which renders the horizontal and vertical scales unreadable.

The horizontal scale is marked from 0hz to 25khz in 5khz increments.  The fundamental, the highest peak is on the left of each graph.

The vertical scale, starting at the top is marked in arbitrary dB beginning at 20dB and decrementing in 11dB increments.  This puts the baseline (instrument noise floor), that band of hash at the bottom just above -68dB.  This puts the peaks of most of the sidebands around -46dB relative to the fundamental.

The total harmonic distortion can be calculated from any of these graphs but a good THD analyzer would be much easier if one can spare several hundred dollars or more for such an instrument.  Such an analyzer may be available as an APP for a schmardt phone.   1/3rd octave analyzers are available that will show a spectrum of what it hears by its mic.  Media players have such a device and are more than adequate for general purposes.  However, if one wants to check THD of an amplifier or speaker, one will need a signal generator, also available on them thar schmardt phones and with Bluetooth, one doesn't need cables.

Believe it or not, despite all the tech equipment here, a schmardt phone isn't one of them.

Eventually, I'll get around to it.   





Just a bunch of spectrographs not intended to indicate which speaker sounds better but to show the harmonics of each which accounts for their difference in sound.  The fundamental is the tallest peak on the left.




























These two spectrographs may have been better if run at 5500hz., the center of the resonant peak.  The Heppner (left) may have looked worse.


This paper cone shows strong 2nd and 3rd harmonics, probably due to resonances in the cone material, such as cone break-up.


Most of the harmonics are around -40dB or lower. The woven Kevlar diaphragm seems to dampen or suppress the effects of cone break-up.



Back to the loudspeaker main page