An Experiment with a low-cost Hearing Aid - page 1
I was looking at the cost of hearing aids recently. Many people don't need a tiny unit that is hidden in the ear (for a price of thousands of dollars). I tried a few of the cheap hearing boosters that are advertised on t.v., but they were just junk. I decided to try making one that would work well, had low noise, and a long battery life (with low-cost replacement batteries). The big problem was that I don't have the facilities to manufacture a sub-miniature unit that fits in the ear. I decided to make one that was similar to the early hearing aids - about the size of a pack of cigarettes. This would allow larger batteries, electronic components large enough to be assembled easily by anyone with a little knowledge of electronics, and give more of a directional pattern. It could be placed in a pocket, hang from a chain around the neck, or sit on a table. The major problem is noise generated by the aid, brushing against clothing while a person walks. I don't see an easy way around that, so it would be something that just has to be worked around. There is a big advantage to putting distance between the mike and speaker. The feedback that so many aids are prone to, is virtually eliminated.
The parts I used are commonly available through mail-order electronics houses. The unit consists of a directional mike, a pre-amp, a headphone driver, and a headphone. I also decided that the aid should run off 5 volts, or less, and draw 5 ma. maximum. I only found one suitable headphone driver chip in an 8-pin DIP (sockets are readily available), the LM4881. All the other ones I looked at, came in very tiny packages. After designing the first working model, I found that the chip had been discontinued. I went to an op. amp., the TLV2460, as a substitute. It is rated at 2.7 v. to 5.5 v., and draws under 1 ma. The secret of an aid that works well, especially on the low end, is to have a good seal around the ear. Foam headphones, or even earbuds, will not work as good as an earpiece that has a soft rubber insert that fits into the ear, or an around-the-ear headphone. If you can find an in-the-ear type, with a small channel cut into the earpiece, to allow some outside air to come in, so much the better. According to what I've read, this can help to reduce an unwanted boost below 500 Hz, that can be a problem with completely sealed units of that type. The earpiece I used, has a small channel cut into it. It also comes with interchangeable rubber inserts, to fit different size ears. It allows for easy cleaning, too. The choice of an electret mike with a high S/N ratio is best, to reduce amplifier requirements. The more you need to to amplify a signal, the more noise you wind up with.
Here is a schematic of the basic hearing aid that I ended up with.. Note that high gain is not required in the amplifier. The amplifier only has a gain of about 100 times (about 33 X for the pre-amp, and 3 X for the headphone driver). While a 3 volt supply is shown, the circuit will work from 2.7 v. to 5.5 v.
ENHANCEMENTS:
1. Adding a second mike (of the same type) in parallel with the first one will
almost double the signal to the pre-amp. The noise will increase also, but since
the noise is random, and the signal output from the mikes is in phase, the
Signal/ Noise ratio will increase. I've tried up to 3 mikes in parallel. Above
that, there may be a point of diminishing returns with this simple setup.
2. Using tuned circuits instead of just a 2.2K resistor for the mike load,
allows a boost in the area between 2 Khz and 3Khz, for better voice clarity. See
figure on left below. The values shown give peaks at 1.5 Khz, and 2.7 Khz.
NOTES:
The finished unit above includes the second mike and the tuned
circuit option (instead of just a 2.2 K mike load resistor). I also used a 3.6
volt cell phone ni-cad battery pack, instead of alkaline batteries. This made
the unit a little bigger (the case is 2.25" x 3.6" x 1.5"). I also added a
high/low switch to cut the low end response for reducing low frequency
background noise (adds a 0.01 capacitor in series with the 0.01 cap to U1 pin
3). Note the holes in the side of the top section for sound cancellation, and
the aluminum foil used for shielding. I also used some thin self-stick felt
padding on the back side of the case, to make the unit less prone to moving
around, if worn around the neck. There are other methods of gain control than
the one I chose. It's just the one I decided to use.
Parts were from Digi-key. Note that the case and 100 mhy coils are being
discontinued as stock items there.
Electrolytics are Panasonic 6.3 v units.
Amplifier gain (as shown) is about 100 times with a 3 volt supply.
The Unidirectional mike I used has holes on the back side a well as the front.
This allows sound that enters both ends to cancel. The case should have a solid
flat surface on the front side, and there should be holes drilled into the sides
of the case, to allow side noises to enter the back of the mike. This makes the
mike respond best in the direction that it's pointing.
The headphone impedance is 32 ohms. Two headphones in parallel give 16 ohms. The
unit probably shouldn't be operated with lower resistance than that. It isn't
designed as a headphone amp. I used a low-cost Fellowes "Jabra" earset I got at
a dollar store. I cut off the connector plug and hooked the headphone wires to a
standard 1/8" miniature stereo plug. I didn't use the integrated earset mike
wire.
A standard 1/8" miniature stereo headphone jack was used for the output, to
allow different headphones to be plugged into the aid.
Why the transistor pre-amp.? Using the 2N5089 pre-amp., instead of just using the op.amp.
at higher gain, drops the amplifier output noise by about 50%.
Response: about 350 - 5500 hz (50% down)
For wider response on the high end, decrease the 470 pf capacitor value (say, to
220 pf).
For wider response on the low end, increase the .022 mfd cap to the base of Q1
(say to, .047 mfd), and the .01 mfd cap to pin of U1 (say to, .022, or
.033 mfd).
Output noise: < 0.75 mv p-p with mike(s) removed. Almost inaudible with the
specified earpiece.
Current draw: 2 ma., or less (no signal), at a 3 volt supply voltage.
The unit will work a long time on a pair of alkaline "AAA" cells. I ran a test
using a couple fresh alkaline "AAA" cells and got over 100 hours. The 2.7 volt
lower spec. is the limiting factor. If you use three 1.5 volt batteries in
series (4.5 volts), you should get well over that. You could also use a 3.6 volt
(3 cell) rechargeable battery pack. That's what I did in my test prototype unit.
Another possibility is to use a miniature 6 volt camera battery. You need to put
a 1N4148 (or similar) silicon diode in series with it, in that case, to drop the supply
voltage down below 5.5 volts. Keep in mind that the transistor pre-amp. gain
(and noise)
also increases with supply voltage.
Max. output @ 3vdc supply.: 2.5 v. p-p with no 1N4148 diodes from U1 pin 6 to
pin 2 (this op. amp. has nearly rail-to-rail performance).
Max. output @ 3vdc supply.: about 1.5 v. p-p, with 1N4148 limiting diodes. The
diodes give a "soft" limiting action, rather that just clipping the waveform.
If the hearing aid gain is too high, with the gain pot turned up all the way,
the 22 ohm Q1 emitter resistor can be increased to 47 ohms (or up to 100 ohms,
if needed). You don't want to operate the electret mike(s) too much below 1.5 volts
in normal operation,
if you can help it. Increasing the emitter resistor will make the gain of the
pre-amp. stage decrease, and also the amplifier noise (they track together). It
also gives a little boost to the input resistance (less mike loading).
2009 Update - Lately, I've been seeing some really cheap,
imported hearing aids for under $25. You can't build a hearing aid for that
price. Some are in-the-ear, and others are behind-the-ear designs. Some are
rechargeable - no batteries to replace. Obviously, they aren't digital, or with
frequency-selective filters, and they don't have molded, custom-fit shapes like
the expensive models, but for the average person, who doesn't want to spend a
thousand dollars (or more), they might be worth trying. I tried out a few, and
they worked pretty well for me, but like a lot of people, I just have a gradual
loss of sensitivity, increasing at the higher frequencies. You can find them on
E-bay, and other places (search "hearing aids").