[CW] A note on headphones

[Richard Knoppow via CW]

     Perhaps O.T. here but I don't know where else to post 
it.  You are all aware that a great many companies made 
headphones over the years, somehow or other I began to 
collect them and have a small collection. Most are bi-polar 
magnetic but I also have a couple of pairs of Baldwin 
balanced armature phones and some moving coil phones.  I 
mesured the impedance of most of them. I have a PDF but need 
to ammend and update it.  Generally, phones fall into three 
classes of impedance: high, medium, and low. In the early 
days of wireless phones were usually used on detectors like 
coherers and later crystals. These have rather high 
impedance output so require phones of as high an impedance 
as possible to avoid loading them. Later phones were used on 
vacuum tube amplifiers and other sources where a low or 
medium impedance is desirable. Its very common to find the 
DC resistance of a headphone specified with no reference to 
its impedance. There is no fixed relationship between them 
but for bi-polar magnetic types the ratio of impedance to 
resistance is somewhere between 3 and about 7 times so, for 
instance, a pair of phones with 2K ohms DC resistance may 
have an impedance of something like 10K to 15K.
    Another characteristic never specified is sensitivity. 
This is completely independent of impedance and has to do 
with the efficiency of the mechanism in converting 
electrical power to acoustical power. In turn this is 
dependent on the strength and effeciency of the magnetic 
circuit and other factors such as the effeciency of the 
coupling of the diaphragm to the ear.
    Now, I decided to make a couple of very crude tests as a 
start. This is simply to drive a few pair of phones through 
a high resistance and see which ones were loudest. I judged 
this by ear but have a simple sound survey meter that I can 
use to keep from being fooled.  I chose a 100K resistance to 
put in series and drive it with a General Radio audio 
oscillator. I also checked for uncoupled resonance (it is 
slightly different when the phone is coupled to an ear). The 
most sensitive, i.e. loudest phones were a couple of pair of 
Western Electric 509Ws, about even with these were a pair of 
Trimm Featherweight specials (24K impedance). Other phones 
such as a pair of Federal Radio and Telegraph 53-W were 
noticably less sensitive. These have about 15K impedance. My 
Baldwin phones test at about 18K and are somewhat less 
sensitive than the WE phones. So, it would appear, at least 
from these simple tests, that the effective sensitivity of 
these phones is proportional to their impedance when driven 
from an impedance substantially higher.
   Most of the magnetic phones have a strong primary 
resonance at about 900 hz. It probably changes slightly with 
acoustical loading.
   Impedances were measured at 1Khz using a General Radio 
1650A impedance bridge. I have means for measuring at other 
frequencies but it takes many separate measurements so I 
have not done it.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com