Simple Hearing Amplifier

Hearing aids are probably more high-tech than you think. They are tiny. They have to go through a lot of trouble to prevent feedback. They need a long battery life. The good ones match their amplification to the inverse of your hearing loss (amplifying only the bands where you don’t hear as well).

[NotionSunday] put together a hearing amplifier project that probably doesn’t hit many of those design criteria. However, thanks to a 3D printed case, it looks pretty good. The device uses a dual opamp to boost the output from two microphones and feeds it to a conventional headphone.

The device is wired point to point, and is, perhaps, pocket-sized. The opamp circuit is simple. We might have considered an LM386 or some other integrated audio amplifier block to get better performance without blowing up the parts count.

Technically, devices like this are not really hearing aids, they are “personal sound amplifiers” and the FDA warns people not to substitute them for hearing aids. If you want a proper hearing aid, you might check this out, but  you are going to need big ears since it has an Arduino inside.

25 thoughts on “Simple Hearing Amplifier

  1. Probably not for me. My hearing loss is linear and starts at 95dB on low frequency end to 125dB at 3KHz and it’s off scale. My last hearing aid pair was $3200 -_- If it was as simple as using a $5 chip I would have gone that way.

    1. Right. It can be quite a complicated problem! I have a daughter with complex hearing issues (asymmetric and fluctuating hearing loss combined with severe hyperacusis and what not) and advanced devices don’t really cut it.

      I could easily add high end MEMS microphones with a DSP, amplifier and all — no problem! The real problem is that DSP options are very limited. Either you need some really advanced knowledge in this “real” DSP stuff (that often means BGA parts) combined with knowledge from the audiology field, or you’re stuck messing around with a simple but really limited DSP (the kind you program with a drag and drop GUI, like SigmaDSP parts).

      The best option *by far* is the specialized DSPs, like the RHYTHM series from ON Semi. I could design something with them no problem, the actual problem is buying them. You can’t buy ’em on digikey/mouser/whatever, and you can’t find them at any distributor either (nor findchips/octopart). And the company will only sell to them with a MOQ of 250 or more, at like $50+ a piece… Programming/controlling the devices might be an issue too (often software and datasheets require an NDA or whatever)

      1. you don’t really need a lot of horsepower or special-purpose hardware anymore to do some digital processing on audio signals. Heck, look at what a Raspberry Pi can do in gnuradio..

        so what you need is basically an equalizer with high gains ? Stereo in, stereo out, 44kHz, maybe even change the curve on your smartphone over wifi, piece of pie for the Pi.

        1. You should use Pi as pizza slicer too…

          A dsPIC has enough horsepower to do all the DSP necessary for equalization, noise suppression and even frequency shifting. If you need big, power-hungry ARM board running Linux for that, then you must be not a very good programmer. Also hearing aids are generally small, portable and running on primary batteries for weeks. RPi is none of that. And really, RPi is NOT A GOOD solution to every problem. It’s a solution for bad programmers who need a quad-core CPU to make some blinkenlights…

          1. hey relax guy, the context was that he had no experience with DSP and that the commercial products were very expensive. I just reached out to him with an easy an inexpensive DIY solution, implying this makes me a bad programmer is just arrogant nonsense. I have been designed many commercial products, with some parts still in space today. Just because i participate on a DIY forum does not make me an amateur like yourself.

      2. You could probably do something with an ARM Cortex M4 too. I have no real experience with audio DSP processing, but there are some cheap eval boards, from ST for example, to get you started and check if the controller could keep up with your needs.
        I would look for something like the STM32F407G-DISC1 board. Comes with the STM32F407 controller, MP45DT02 digital microphone, CS43L22 audio DAC and amplifier. So it’s basically ready to pick up sound, process it in some way and drive some headphones with the included 3.5mm jack. Avaliable for less than 21$ at mouser and most other distributors. There are some more boards from ST with bigger F4xx controllers and included mems microphones and audio DACs, like the STM32F411E-DISCO, STM32F412G-DISCO with a little display and SD-slot, or an even bigger board with the new M7 controllers like the STM32F746G-DISCO, so just look at what you think you need and give it a shot…

        1. I already use ARM Cortex M4(F) extensively, and M7(f) would be just as easy. The hardware is a non-issue in this case. It’s all the DSP stuff that requires a lot of math and specialized field knowledge. Here, having a Raspberry Pi CPU or a 16 core Xeon on the board wouldn’t help me either.

          It’s about compressing and shifting some frequency bands, having variable gain on them,
          doing noise cancellation (between omni and directional mics), feedback cancellation and what not… While keeping latency low, quality high, distortion low, an AGC-like mechanism for quiet and loud places but also with a smart threshold for noise (squelch-like, for radio people), detecting and automatically blocking wind noise picked up by the mic, etc. And then I need specialized audiology-related knowledge regarding how to improve speech discrimination by adjusting the sound (it’s not about making it louder)… And of course, all of that has to be easily reconfigurable on the go. And then you start thinking about interfacing with neck loops, bluetooth and all of that.

          I couldn’t code half of that in a decade (I’m totally NOT a math/DSP person) so I’d just much rather pay $100 for an IC that does all of this for me! Doubly so when it comes with specialized software to program it. I’d happily buy a few samples. Right now I’d have to spend the price of a small new car to get my hands on some…

  2. That’s not a “hearing amplifier”; it is an amplifier. Hearing aids are much much more subtle than the article suggests.

    The ear-brain combination is exceedingly non-linear in virtually any dimension you can think of. That’s why audio compression algorithms can work. Hearing aids have to counteract and augment the non-linearities.

    They also have to do it using a tiny currents and <1V supply, which is quite a trick when using DSP technology

  3. Some of the hearing aid tech is actually really clever! My audiologist told me that they go beyond frequency compression to remapping – so sounds normally out of your hearing range are ‘moved’ to ranges within. Sadly my brain is too old to adjust to such freakish stuff, but interesting none-the-less.

      1. Yes… which makes it eminently FPGAable. I also suspect the work done around noise suppression for the likes of Amazon echo would feed well into a homebrew solution.

  4. There was something like this on TV ads years ago. I also remember hearing about lawsuits and state laws and a bunch of CYA poop.
    I had a friend many years ago who had some loss more in one ear. I had a portable stereo cassette recorder and a pair of mics angle spaced and attached to the headphone’s band. The binaural localization was good over 180 degrees front. He could wear it and walk around and hear the softest sounds of nature, and in balance.
    All of that tiny size design is for modesty. One generation was shocked at the sight of anything attached to the body. It had to be hidden. Another pokes prods and jams all kinds of stuff on and in the body. Headphones? Blurtooth gadgets. No problem.

    1. Bluetooth! Genius! It’s acceptable to wear in public, and you can get some really tiny ones. Then you can have the processing box be as big as you like (well, pocket-size but bigger than your ear canal). Would be a good start for home-made hearing aids.

      1. So, let’s have the smartphone do the audio processing of what it picks up with its microphone, and send that to the bluetooth earpiece… or yeah, the earpiece often includes a microphone.
        With a bunch of $10 cellphones, this could be a boon in the 3rd world…

  5. The hearing aid industry in for the general population is nothing more overpriced crap å la basic eye glasses costing $500 with lenses that cost a few bucks to polish out with the Rx… Picked up a show from across the border where they spent the episode cracking this little gem and eventually found one of the more popular brand’s engineers in Florida where he admitted there’s nothing more than $24 worth of parts in it – he was referring to a $1500 model. Not $24 at his scale of production, $24 “at yours” pointing to the interviewer. He refused to say what they pay but acknowledged “a lot less”. What irked me about those is that these unit “families” constitute the bulk of retail sales by stand-alone retailers.
    [Now, the really interesting stuff – high end ones – are a different story]

  6. I’ve got a DIY bone anchored hearing aid project I’m working on. Born deaf in my right ear, as in that my cochlea is completely dead on that side and absolutely no sound has ever been heard with it. In a way I’m fortunate because my good ear is indeed good, perfect hearing. So I can get away with using simple amplifiers without complex features as noted in the article.

    Being as new and green as I am with electronics I’ve cut corners here and there, been using a <$2 3W-5V stereo amplifier that I found on aliexpress along with some tiny surface exciters (transducer meant for turning surfaces into speakers). USB lipo charging module, 18650 battery, electret mic, blah blah. Works really well, better than the $5,000 Cochlear BAHA 5 that I got and hated.

    All the modules and accessible knowledge out there at this present time has made it fairly easy-mode for me to get up and running with results. Keeps me encouraged and engaged, rather than rage-quitting over my being inept.

    I've run across a handful of MEMS microphone modules, researched it a little but I wonder what benefit they may serve beyond just being smaller than electrets (like the ones shown in the video)?

    1. Also as a sidenote, there’s a project out there known as BioAid and app AUD-1 based on their algorithm that uses your iPhone/Pad/Pod Touch’s hardware and microphone to do the kind of processing that’s inside contemporary hearing aids. Frequencies, ranges, etc.

      Not perfect, not a direct replacement, but a good armchair exercise and a solution for those who have no affordable avenue to obtain a proper hearing aid. Emergencies when their hearing aid is broken/lost/dead batteries. Etc

      1. I’ve had a chance to try the BP100 Power, was a better design than the 5. The 5 has no way to turn it off other than removing the battery. Only way you can change the volume is with an iPhone and it’s unreliable and rarely works. Or you can invest in an expensive remote control to do it. My biggest complaint about the BAHA 5 is that it doesn’t work well enough to make any real difference. In most scenarios it makes it harder to hear than help.

        So good job on the BP100, no one seems to complain about it the way I can about the 5!

  7. As others have pointed out the project is of limited use without EQ for each channel as hearing loss is complex and each ear will have it’s own sensitivity curve.

    Also, the placement of the microphones is for from optimal if you also wish to retain the important spatial clues that hearing provides.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.