Over on Slashdot, there’s a post that caught my eye because it is so simplistic and so wrong. It’s something I have deep personal experience with, and while not the normal fare for WUWT I thought I’d share my detailed response for the benefit of others. First, here’s the Slashdot story:
An anonymous reader writes “The price of a pair of hearing aids in the U.S. ranges from $3,000 to $8,000. To the average American household, this is equivalent to 2-3 months of income! While the price itself seems exorbitant, what is even more grotesque is its continuous pace of growth: in the last decade the price of an average Behind the Ear hearing aid has more than doubled. To the present day, price points are not receding — even though most of its digital components have become increasingly commoditized. Is this a hearing aid price bubble?”
My response: [As noted in my WUWT About page] I wear two ITC/CIC hearing aids with DSP processors built in. Let me tell you a little bit about why they are so expensive. The largest supplier of hearing aids in the USA is Starkey in Minneapolis. I’ve been to the factory, and have experienced the process from start to finish courtesy of the president of the company.
1. Because hearing aids, especially BTE (behind the ear) and ITC/CIC (completely in the canal) types use a single cell 1.5 volt battery, which can drop as low as 1.3 volts through its useful operational life, the amplifier circuits must be of extremely low power consumption and low voltage. The only chip material that works well for this is germanium, which has a diode junction forward voltage of ~ 0.3V as opposed to the ubiquitous silicon used in consumer electronics which has an ~ 0.7V forward voltage. While germanium was once very common for transistors and some early integrated circuits, it has fallen out of favor in the microelectronics hearing aid world. There are only a handful of sources and companies now that work with germanium, thus the base price is higher due to this scarcity. You can’t just take an off the shelf silicon chip/transistor and put it in these aids. Each one is custom designed in germanium. [Added: power consumption is a big issue also, aids are expected to last a few days on a single battery, if most of the power is being used to overcome the forward diode voltage, it gets lost as heat instead of being applied to amplification use.]
2. The process of properly fitting a hearing aid is labor intensive. Custom ear molds must be created from latex impressions, and these need to be fitted for comfort. A small variance or burr can mean the difference between a good fitting mold and one that is painful to wear. Additionally, if the mold doesn’t maintain a seal to the inner ear properly the hearing aid will go into oscillatory feedback. Sometimes it takes 2 or 3 attempts to get the fitting right.
3. On the more expensive aids, labor is involved in doing a spectral hearing loss analysis of the user’s hearing problem, so that the aid doesn’t over-amplify in the wrong frequencies. Just throwing in a simple linear amplifier is destructive to the remaining hearing due to the sound pressure levels involved.
4. Construction of aids is done by hand by technicians, especially with the popular ITC (in the canal) aids. At the Starkey company, a technician is assigned to create the aid from the ear mold, fit the chips and microphone/receiver and battery compartment, and connect it all with 32 gauge wire and make sure it all fits in the ear mold. This can be a real challenge, because human ear canals aren’t often straight, but bend and change diameter. Imagine a room with a hundred technicians sitting at microscopes assembling these. Each is a custom job. There’s no mass production possible and thus none of the savings from it.
5. After the aid is created, then there’s the fitting. This process is also hands on. Getting the volume and the audio spectrum match right is a challenge, and audiologists have to have chip programming systems onsite to make such adjustments withing the limits of the aid. Sometimes aids are rejected because the user isn’t comfortable with the fitting, and then the aids go back to the factory for either a new ear mold, new electronics, or both.
6. There’s a lot of loss in the hearing aid business. Patients don’t often adapt well, especially older people. There may be two or three attempts at fitting before a success or rejection. Patients only pay when the fitting is successful. If it is not, the company eats the effort and the cost of labor and materials. Imagine making PC’s by hand, sending them out to users, and then having them come back to have different cases or motherboards or drives fitted two or three times, and software adjusted until the customer is happy with it. Imagine 4 out of 10 PC’s coming back permanently after trial and error with a customer.
7. Early hearing aids weren’t anything but simple amplifiers. Even until the mid 90’s there was very little spectral customization. Now many aids are getting features like frequency equalizers and DSP noise reductions that we take for granted in even the cheapest silicon based consumer electronics. Hence, price has increased with complexity, but there’s still the high cost of custom special chips, and lots of labor.
So for those who think mass production techniques used on iPods would work just fine for making a delicately balanced instrument that must fit in your ear, please think again. As a hearing aid user since 1969, do I think the price tag of the special hearing aids today are worth the price compared to the simple linear amplifiers I used to have to deal with? Absolutely.
For more on hearing loss, see the Starkey Hearing Foundation, which I support.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Applegate (September 10, 2011 at 7:50 am) wrote:
[…] I recently obtained BTE aids for $7000 the pair. These should require no such custom manufacture since the aid itself is in a large standard size/shape package that connects to the ear through a simple tube. I now question whether those types should be so expensive. Even if I returned them, the aid itself could be resold.
Welcome to the wonderful world of medical devices! The US Food and Drug Administration (FDA) has an endless collection of rules about medical devices. You certainly couldn’t sell a returned hearing aid as “new”, if the FDA permitted it to be resold at all. Whatever the outcome, the manufacturer takes a bath on each returned device.
Janice (September 10, 2011 at 7:49 am) wrote:
About a dozen years ago, I had a nice expensive pair of hearing aids that worked beautifully. There was just one small problem: I was allergic to the hypo-allergenic plastic they were made from. … The hearing aids are in a box in a drawer, and I simply ask people to repeat things. …
If you want to try again, look into behind-the-ear hearing aids. Then all that will be in your ear is the earmold. Before you commit to buying anything, check around to see if there are alternative materials available for earmolds. You can get a hearing aid from one manufacturer and an earmold made by another.
If all else fails, you can try my mother’s trick for things that irritated her skin. She’d paint a couple of layers of clear nail polish on the parts that touched her skin. Worked every time for her. When the nail polish began to wear off, she just painted on another coat.
The only real problem I have with hearing aids is the absolute loss of common sense in people around us.
I probably wrecked my ears with 72 years of flying airplanes, driving, running chainsaws and lawnmowers and such, all without ear protection.
The problem is that people think that (for all problems including differing languages) by making things louder the problem will be solved.
For people like me, if you talk so loud that my ears bleed, I will still be unable to understand you,
The damage to my hearing (which I do not understand at all) is such that I can not understand what I hear.
In an anechoic chamber (a really really quiet room) listening to an audiologist read a list of words at a fairly low volume (my best chance) through an earpnone there are some words I can not make out if I can’t see her lips.
So, to stop family members and other “friends” from hammering on me about getting over my vanity and getting a hearing aid, I got one (a BTE, not-custom-molded) at a “reasonable” (see discussion above to calibrate “reasonable”) price and with it, I can hear and understand people sitting or standing (or lying) to my right in a fairly quiet room. Even it the HA is in my pocket.
Somebody asked about on-going costs. The batteries cost a bit over a dollar apiece and last about a week. I replaced the ear-wax filter for the first time in nearly a year that I have had the device–there at most of about a dozen of the original package left–I have no idea what they cost if I live long enough to use them all up. The soft-rubber cone thing the receiver is in looks like it will fail eventually (it turns wrong-side-out sometimes when I take it out of my ear), I have another in the original set, and I have no idea what they cost.
And someone is surely to wonder what IGFETs are: today they call them MOSFETs! Showing my age again!
Dave says:
September 10, 2011 at 9:25 am
Anthony or readers. For me this is a timely discussion, over the last 2 years my hearing has degraded ,but worse I have a constant ringing in the ear. There’s lots of info in the net but does anybody know if there is a real cure or research that will help this problem or am I dammed?
###
Your are probably only dammed if you are a river. Seriously, I have been dealing with sever tinnitus for decades. For me, according to the doctors, it is caused by nerve damage, and therefor not treatable.
I have tried to learn Anthony’s trick, but have been unsuccessful. I can think away the pain and even somewhat the effects of burns, a skill I learned as a cook by applying what I was taught in Karate. But maybe because of my ASD, I just can not ignore sounds, though I am sure most normal people can learn to do it.
The cost of materials is surely a factor, but the main driver of pricing is going to be labor and insurance. Start with taxes which more then doubles the cost of everything, but especially labor. Then add the insurance which more then triples the cost of medical devices. Then on top of this, add Government regulations which impede innovation and distort the market. E.g., does anyone have any idea how much it would cost and how many years it would take to qualify Li-ion batteries for hearing aid use? Is it no wonder that the cost of hearing aid has sky-rocketed.
Re:Marcos says:
September 10, 2011 at 11:30 am
“i’ve been wearing CiC’s since i was about 20 (i’m 39 now) and what i’ve always been curious about is how the price of an aid can drop 30-50% just by going down a quality tier or two when the labor is still the same. you might get 8 instead of 16 channels and maybe (slightly) less sound processing but you can save thousands of dollars….”
Labor hasn’t that much to do with the phenomenon you describe. A truism of the electronics business is that you’d better recover all your R&D costs and rake in probably 90% of all the profit you ever expect to make on a new widget in the first year after introduction to the marketplace. After that, competition (some of it copy-cat) will drive down the price to commodity proportions. I try to wait until they’re practically giving away long-in-the-tooth electronic widgets as door stops (can’t resist cutting edge artificial limb technology though).
I am a retired electrical engineer (BSEE, MSEE and MBA, btw) who has done both electronic linear amplifier design and digital logic design in bipolar (both silicon and germanium) and CMOS (an insulating gate field efffect transistor, IGFET, which has been referred to in this thread) in my career. I am not aware that it is possible to build DSP processors with germanium… the chemistry of germanium does not have a well behaved oxide (like silicon dioxide) that allows what is called Planar Processing which is the manufacturing-process basis of all silicon-based highly integrated circuits, or what is commonly called “chips.” Germanium is only capable of discrete transistors, to my knowlege. I doubt that more than a relative few discrete transistors are being used in a preamplifier circuit and/or an output driver circuit. By comparison even a quite small CMOS silicon low power processor (whether DSP or microcontroller) can have many tens of thousands of transistors, and cost less than a buck in a general purpose version. The larger ones in your PC these days have upwards of millions of transistors (and the power dissipation adds up even though they operate at less than one volt!).
There are CMOS technologies that would allow a custom DSP processor plus any required linear audio amplification all on one chip and operate at 1.3volts. Getting the absolute lowest power dissipation is a matter of custom design of such a chip (it would not do to use a general purpose DSP chip), and this kind of custom mixed analog/digital design was my career forte, although we never applied it to hearing aids as a product. We never saw enough market volume there to justify the Non-Recurring Engineering costs of a design effort. It may be different today, but it could be quite expensive per chip compared to general purpose chips that have huge markets.
While it is true that a bipolar germanium transistor can dissipate less in the diode junction than a bipolar silicon transistor operating at the same current, due to the higher silicon forward voltage, as two commenters have already said this can be totally obviated by analog CMOS transistors where the insulating gate has virtually zero current (power = voltage x current).
On the other hand, I have no doubt that engineering a good hearing aid has challenges and problems as well as solutions that I am entirely ignorant of. It would be interesting to know more about the claims behind the use of germanium in a hearing aid.
About 25 years ago I worked with a venture capitalist who wanted to create a national chain of hearing aid stores, in essence doing for hearing aids what Pearle Vision and Lenscrafters did when they revolutionized the eyewear retailing industry. His hypothesis was that the “mom and pop” nature of hearing aid retailing was grossly inefficient and with changes in how the product was retailed, the costs could be lowered considerably. After a month or so we dropped the idea because we found that the density of people with hearing problems in most metro areas was not sufficient to support even a modest sized retail establishment. Given reasonable assumptions of the market share the new business concept might earn, the number of potential customers that lived within a reasonable driving distance of most strip malls was insufficient to generate a reasonable return on investment. We found that there were only to market areas in the country where the customer density was adequate — Sun City, AZ and Tampa, FL. People that believe the “market works” will not be surprised to learn that high volume strip mall hearing aid retail outlets already in operation in both of those markets. In spite of technological advances that we knew were coming, demographics and the frequency of hearing loss within the general population dictated that hearing aid retailing was and would remain for some time a cottage industry.
Anthony:
Re the 100 MPG Carburetor. It’s call the Garrett Carburetor. It’s an electrolysis, hydrogen supplement device. I refuse to get caught in a discussion of its “validity” for a variety of reasons.
But the patent exists.
Whoops, posted automatically when I went to get the patent number: 2,006,676
Please note the “assignee”. Sunoco Oil company. NO they did not “buy Garrett out”. Consider the engineering level of the time, the Sunoco ENGINEERS looked at it, determined it had NO MERIT, and it was never pursued.
Origin of the URBAN LEGEND. Pure and simple.
@JDN 8:36 am
http://www.wikinvest.com/stock/Sonic_Innovations_%28SNCI%29/Data/Income_Statement#Income_Statement
Sonic Innovations is losing money. Losing big money. Just thought those folks laughing off the price as pure profit ought to know, even if it likely won’t matter to them. “OK,” I hear them say, “but that’s just one money grubbin’ evil corporation.” OK, fine, lest I be accused of cherry picking:
http://www.wikinvest.com/stock/Sonic_Innovations_%28SNCI%29/Data/Derived_Net_Income
Comparison of Sonic’s competitors. Only three of twelve are turning a profit.
Tried looking up Starkey, they’re private so no SEC fillings.
Retired Engineer (September 10, 2011 at 9:10 am) wrote:
The silicon / germanium argument is a red herring. Perhaps true for bipolar transistors but not for FET’s. Several companies sell CMOS amplifiers, pico power, and operation below 1 volt.
The answer isn’t a collection of discrete parts, no matter how capable or low-power, but an integrated solution. It’s a very expensive undertaking, though, and takes several years to go from “let’s do it” to “it works”.
The company I work for has integrated the major functionality of its pacemakers and implantable defibrillators. The pacemaker we released a couple of years ago, built on the integrated chipset, uses a 2.7 volt lithium primary cell for power, which is switched down to run the (silicon) CPU and DSP at 1 volt. With reasonable settings, the battery can be expected to last more than 10 years.
A hearing aid requires more output power than a pacemaker, but has the advantage of using replaceable batteries (pacemaker batteries are most emphatically not replaceable). Output power can be significantly reduced over older designs by using a Class D amplifier (switched output transistors, followed by a filter).
The ultimate goal should be to integrate the microphone amplifier, the A/D converter, the DSP, the controls, the output stage, and even the Bluetooth interface onto a single chip. From then on, you differentiate your products by changing the software. Occasionally, you update the chip set as newer technologies become available, allowing you to do more with less power, or to do new things you couldn’t do before.
Retired Engineer (September 10, 2011 at 9:10 am) continued:
… My complaint is with non-rechargeable batteries. Why not Li-ion or Li-PO? I talked to an engineer at Starkey and he said “we can’t do that” and nothing more. Illegal? or an ‘arrangement’ with battery makers?
My last job was at a manufacturer of test and measurement instruments. We had a product (which I hadn’t designed) that contained a rechargeable battery for memory backup. We had two suppliers for the battery. One manufacturer’s batteries would grow metal whiskers through the separator, shorting out the cell, which allowed the memory contents to evaporate. The other manufacturer’s batteries would lose their electrolyte through the seals, eventually becoming an open circuit, which allowed the memory contents to evaporate. We had a second source for the part, but lousy results no matter which one we used. (My designs all used primary cells — non-rechargeable batteries.)
In general, batteries are the pits. During their discharge life, the battery voltage and impedance change in ways that can cause the circuitry the battery powers to misbehave. Primary cells have fewer odd behaviors than rechargeable batteries, they have higher energy densities, and they’re more reliable.
If your hearing aid battery (a primary cell) dies in the middle of the day, you can replace the battery and be back in business in minutes. If you had to recharge your hearing aid before you could use it again, you may well be out of operation for the rest of the day. For hearing aids, replaceable batteries are the way to go.
Please be mindful of SiGe IC technology, which has evolved to the point of being used in commercial product today … albeit bipolar (NPN and PNP) transistor technology I think vs MOS (think: “Field Effect” transistors) …
.
Full disclosure (on this thread’s subject): My sister is a Cochlear Implant (direct connection of 5 – 10 electrodes to the sound ‘nerves’ that go to the brain with a DSP used to create the ‘stimulus’ signal) user/subject, in one ear for +18 years now and just last year had the 2nd ear ‘implanted’; if ever there was a miracle technology, CI would be it!
.
Excellent commentary.
The author of the Slashdot post would also be upset, presumably, that an iPhone costs so much more than wireless household phones did in the 90s.
Not even sure where you were going with this Roger, but, substantially anything you can do with a Si device I can do with a Ge device, notwithstanding some limitations for power dissipation on account of the tendency for Germanium devices to ‘self forward bias’ at elevated temperatures as minority-majority (hole-electron) carrier pairs self-generate regardless of control input (Base or Emitter) currents … thermal runaway, secondary breakdown etc are some of the issues moreso with Ge than Si bipolar technology (veteran of moderate to high-power power supply and RF transmitter design and test here) …
.
An OEM of hearing aid subsystems/processors (ONSEMI EZAIRO™ 5900 digital sound processor – click the video for a quick review):
http://www.onsemi.com/PowerSolutions/content.do?id=16488
Ultra low-power digital-based product from them:
http://www.onsemi.com/PowerSolutions/content.do?id=16851
Ultra low power consumption devices “Sound Design Technology” acquired by ON Semiconductor:
http://www.sounddesigntechnologies.com/
Some possible insight on the fabrication technology they use:
Sound Design Technologies Foundry Solutions
Low Voltage Low Power Bipolar JFET Process (Bi-FET)
Note in the above a standard analog processing ‘cell’ that can used to develop various functions (e.g. low pass, high pass filters etc.)
Further research finds that:
Primarily, “Sound Design Technologies” provided Si wafer processing of analog, not digital circuity …
.
An article on WUWT that I know about! 🙂 In my real life, I’m the CTO of a hearing aid company (and I spend way too much time in Xiamen, China – the heart of hearing aid assembly worldwide). The costs borne by the customer aren’t just for the hardware – it’s the fitting, diagnosis, and tuning that is the expense.
The average person goes back to the audiologists between 12 and 18 times in their first year – and that is almost always covered in the purchase price of the aids. At an average cost of $100 per office visit, it’s quick to see that upwards of $2500 of a set of hearing aids can be just for the audiologist’s office time. The profit on the hardware sales side is considerably less because of the large margin given to the seller to cover their costs on the tuning and service of the client for the first year.
This makes me feel very humble. Thank you everyone. It puts in perspective the £500 it costs here for me to get a pair of hearing aids with Specsavers, or free for one non-adjustable hearing aid on the NHS.
I’ve become painfully aware of my hearing problems in the last month, after flying to Norway, losing a substantial part of my hearing after the flight, and having the embarrassment of asking my hostess cousin to repeat almost everything she said, whether she spoke Norwegian or English. It really was not nice after a week.
Cochlear implants are one way to go. Another is Cochlear’s Baha hearing aid system that will work when a conventional hearing aid is not the optimal solution due to ear canal problems or permanently perforated eardrums. The Baha system works through direct bone conduction. A titanium implant is emplaced in the mastoid bone behind the ear and the custom tailored hearing unit is attached to the implant. Details can be found at: http://www.cochlear.com
Polistra:
“This particular bubble has been around for a long time. It’s not the germanium in the DSP, it’s just an ordinary cartel arrangement between the few HA manufacturers.”
—
You might have been told that it was cartel, but in the real world there are very few cartels that operate successfully. It can only “work” with things of limited supply. Hearing aids, or anything that can be manufactured in infinite quantities are a poor choice for a cartel.
Like commenter Shanghai Dan, I too spent way too much time in China. Shenzhen, during the boom years of late 1980’s to late 1990s in my case & high tech. Any time a company tries to gouge the consumer, they are very quickly undercut by competition. An attempted cartel is undermined by competition that will appear very quickly.
Therefore – if there really is a cartel, I would propose to Shanghai Dan that we go into business together, launch a cheaper copy and break their cartel 😉
To the commenter who questioned labor competing, thus lowering labor costs. Yes and no. Labor is in finite supply, especially skilled labor, which is why CEO’s and sports stars make so much. If a labor market has over supply wages go down but if it is tight, as it will be in any good economy, wages go up. Just like any market. The role of government should be limited to just making sure markets are free, fair, open and functioning. Beyond that, freedom leads to prosperity, with a few bumps along the way of course.
I have been wearing hearing aids for about 15 years, I am 39 now. I can add a few notes. Expect average replacement at 2-3 years. Also, they have increased in cost, about double, during the last 6-10 years for a comparable product. Just my 2 cents.
Perhaps a better ‘picture’ of hearing aid low-power digital developments:
http://www.sounddesigntechnologies.com/aboutUs_History.php
2000’s – Innovations in ultra low power Digital Signal Processing as Audio and Wireless Division of Gennum Corporation
2007: INSPIRIA™ Digital preconfigured DSP with Adaptive Directionality, Feedback Cancellation with up to 30 dB of Added stable Gain, 128 bands of Adaptive Noise Reduction
– Shipped over a Million FOUNDATION® DSP hybrid packages
2006: Introduction of 128-band psycho-acoustics based Adaptive Noise Reduction with VENTURE™ family of products
– Shipped over a Million PARAGON® Digital hybrid packages
2005: VOYAGEUR™ DSP
– First 130 nm fully programmable DSP product
– Unsurpassed MIPS vs. current and processing capability
2003: FOUNDATION® Digital
– Enabling affordable digital hearing aids
2001: PARAGON® Digital
– First pervasive off-the-shelf digital product
– First DSP with extreme wide bandwidth (16kHz)
More info – a white paper titled:
Signal Processing in High-End (Digital) Hearing Aids:
State of the Art, Challenges, and Future Trends (starting pg 12)
Note: It appears that the first 7 pages are blank, there is content on page 8 onward …
Example digital product from Gennum/Sound Design Technologies back in 2001
Power PARAGON™ DIGITAL Four Channel DSP System with FRONTWAVE
Note the minimum supply voltage shows to be 1.0 volts …
.
I’m not losing my hearing; you guys are just mumbling, and you’ve been getting worse for a long time! But if I ever do get hard-of-hearing, I’ll buy the “green” version–no dangerous EM radiation or toxic batteries to foul the environment. Al Gore approved!
http://antiquescientifica.com/hearing_aid_German_silver_Tiemann_1889_figure.jpg
this is very simple to solve: buy it on uk, from prices ranging from 6 to 160 pounds. See here:
http://www.amazon.co.uk/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=hearing+aid&x=0&y=0