The Trouble With Ceramic Earpieces made by Eagle
or Altai and Others

These are
alternatively called crystal earpiece, piezo earpiece, crystal
earphone, piezoelectric crystal earpiece, and sometimes crystal
receiver.
Background
A while ago I bought a few hundred of these little earpieces
from a UK retail distributor to sell with crystal radio
kits. They are not the finest, most hi-fidelity piece of
listening equipment available, but they are quite sensitive at
low signal levels and have a fairly high impedance of about 27nF
capacitance and very high d.c. resistance. I was a bit
puzzled by strange intermittent behaviour on one that I was
using and received a couple of returns from customers.
Time for some investigation.
Of my remaining stock of 100, I found that about 30 were
dead. They were open-circuit when connected to a low level
transformer coupled crystal set output and some of them would
wake-up when plugged into a transistor radio earphone
socket. A few would be intermittent and respond to
touching of the diaphragm or squeezing the plastic body
itself. With a capacitance meter connected it would show
low capacitance when not working and when squeezed in the right
place would buzz and show about 27nF. This was a puzzle,
but I wrote those off and bought another 100 from another UK
retail distributor. The results on testing were the same,
and the distributor tested their stock on a radio earphone
socket and admitted that there was a quality problem and removed
the item from sale.
After another couple of months they were back on sale with fresh
stock. Unfortunately the construction remains exactly the
same and so the problem remains. This time 38 out of 100
units tested dead on the crystal radio, and about 25 still
showed completely dead on a powered radio output earphone
socket. Let's look first at the manufacturing problem
inside the earpiece.
Inside the Ceramic Earpiece
After unscrewing the rear cap you can see the soldered
connections at the back. So far, so good. Then after
splitting the plastic case in two along the seam you can see the
front of the aluminum diaphragm.


After carefully cutting through the glue and easing the
diaphragm up, you can see the rear side where the ceramic
element is visible. The ceramic element has a metal
plating on its underside. There is a small copper strip
soldered to this plating at one end and to the wire contacts at
the other. The other wire contact has another copper strip
and this goes to the outer edge of the diaphragm. This is
the second contact to the piezo element. It seems that
when undergoing assembly, the strip is folded round to the front
face of the diaphragm, the diaphragm glued into place then the
strip folded down, blobbed over a bit with more glue before the
unit is sealed around the split line, either with cement or
maybe ultrasonic welding.



There is a problem. The contact to the aluminum diaphragm
is being made with a copper strip being held against it just by
having non-conductive glue blobbed over it, and by whatever
force might be pressing upon it from the glued seam. This
will often be zero force. The units may work after
assembly but after passage of time and a flight in an aircraft
it is inevitable that the connection will be poor. There
are reasons that switch contacts aren't made out of aluminium,
and one of those is that the hard aluminium oxide coating that
forms on it in air is non-conductive. You need real
contact force to make an electrical contact. Whatever the
case, you can't make a reliable electrical contact with zero
contact force and non-conductive impact adhesive. You
could probably cure this problem cheaply by blobbing a tiny
amount of conductive paint glue under the copper strip during
assembly, even if that was also a bit of a nasty fiddle.
You really need to make a solder connection, and that would mean
changing to a solderable material for the diaphragm.
The Reaction From the UK Retail
Distributor
The distributor replaced my duds and said that they had quite a
discussion with the manufacturer. The manufacturer's reply
was that the returned stock tested out OK because the measured
impedance was greater than the minimum specification of 2M
Ohms. "Sigh." Yes; Open circuit connection
problems will tend to test out OK if a high d.c. resistance is
your specification for a good item. Sometimes people don't
want to listen and will give any stupid answer. I'm afraid
that when dealing with far eastern factories, you have to be a
bit more persistent after the first series of attempted
brush-offs. Despite knowing about the problem, the UK
distributor intends to continue supplying these to retail
customers untested.
As the entire production is probably made in one small group of
factories, the problem is likely to be worldwide. I'm
hoping that, "By the Power of Google," this article will
eventually get to the attention of the actual factory, the named
manufacturer Eagle or Altai and the UK importer who I believe to
be Electrovision. I won't single out any retail supplier
as everyone in the whole world is probably doing the same
thing.
The retail supplier made some defense of their attitude by
pointing out that customers are not generally slow in coming
forward with problems where they are encountered, and that I'm
the only person seeing a problem. I have an explanation
for this: The vast majority of these earpieces are going
to be connected to crystal set radios which will be put together
by kids during the summer holidays or a few days after Boxing
Day. If, as a kid, you can't get a crystal set working you
tend to assume that your aerial isn't big enough, you're doing
something wrong, or that crystal radios just don't work, it's
boring and that throwing it into a corner is the best course of
action.
Connecting your piezo earpiece to a standard mp3 player or
similar will usually show if it is dead or not, though in about
20% of cases the higher level signal from a powered device can
wake up an intermittent earpiece.
Conclusion
So let's see what happens in the fullness of time. Until
the manufacturer makes a production change, I won't be selling
these as single items. I continue to throw plenty straight
into the bin, and only supply rigorously tested ones in my
crystal radio kit.
Navigate Up
Recent Edit History
05-NOV-2011: Page created
07-JAN-2024: Removed reference to 'the other place,' changed
nocache directive, viewport