Logic for three unit hearing aid system:

1. Right now we put $2500 things in both ears every day and we find that after n years of this they are toast. So then we have to buy two more $2500 things. Annual cost might be about $1500 - $2000 - not acceptable for a whole lot of people whatever our rational justifications.

2. If we can avoid putting things this expensive in harm's way we might save a lot of people a lot of money.

3. Let's instead put $200 things in our ears and put the expensive thing in a box where it's pretty much out of harm's way. How do we do that?

4. Let's take the hearing aid design, leave in the microphone and the speaker and the battery but toss everything else out. Then let's add a low power BlueTooth transceiver, connect the mic to the transmitter and the receiver to the speaker.

5. So now we have a tiny wireless earset. But these already exist! They are used for handsfree cell phone use and they sell for about $100. We'll change them a little so they are attractive to us and have a customizing option to fit us individually. They can be sold for as much as $200 each and meet our goals.

6. These are the same for everybody. Commodities. Dog gets it just go to Cosco and get another one.

7. Now that our ears are wirelessly instrumented full-duplex satellites, we need to buy the expensive hearing aid part. Its one box. It's the dog-proof, drop-proof, wireless full-duplex mother ship that links to both satellite earsets and that replaces the stuff we tossed out before - in spades. Looks like a pager. Rechargeable. Has a battery status indicator for all three components.

8. Now when we turn on the system here's what happens:
Any sound picked up by a microphone in either headset gets wirelessly transmitted to the box. The DSP in the box does processing for both channels differently but simultaneously. Because the DSP has information from both ears, it can make decisions that individual DSPs could not and enhance hearing much more readily. The processed outputs are transmitted back up to the satellite earset receivers and are heard by the user as sound. You can wear the pager-like box but you don't have to. You can put in you pocketbook or briefcase or anywhere up to 30 feet away.

9. Because the DSP in the box is a bruiser compared to typical HA DSPs everything is done better, more often and with less delay.

10. The DSP box uses conventional surface mount components and can be produced using conventional methods and assembly houses. As a result the box is repairable and has a MSRP of less than $600.

11. The result is that a better system is available to the user for about $1000 instead of $5000.

12. The result is that every 2 to 4 years when your earsets become toast you have to come up with another $400 - and your friend with the Siemens Trianos has to come up with another $5000. Your annual cost is about $220. Your friend's annual cost is $1500 +.

13. If I'm wrong and it works no better than the Trianos you still are way ahead.

14. Bonus - Now that you ears are "wirelessly instrumented full-duplex satellites" a whole lot of really first rate assistive widgets can be made available to you. Like ideal quality audio grade links to:

• Cell Phones
• Regular Phones
• PC Sound (MP3)
• TV, Hi-Fi, et.
• Baby Monitors
• Alarms, Door Bells etc.
• Hear Large Area Listening Systems w/o Add'l Devices

Reinventing the Hearing Aid

DSP based hearing aid designers have done a great job within the constraints of the product definitions they have been given to execute.

But I believe that due to these constraints, designers of DSP based hearing aids have inherited handicaps that are causing the resulting products to exhibit limited performance and unrealized potential. Perhaps it is possible today to lift some of these 20-year-old product definition constraints and reap some real price-performance benefits.

The hearing aid industry fell in love with "micro-miniaturization" and "hiding" about twenty years ago and have been busily painting hearing aid design into an ever smaller economic and technological corner. In the meantime the demand for modern, smart sound processing for hearing aids has mushroomed. So engineers, faced with putting little computers complete with power supply into a pill sized enclosures had their work cut out for them.

• They had to reduce the computer based circuit's size.
• They had to very dramatically reduce the power consumed by the DSP.
• They had to include provisions for user programming.
• They had to provide functions like adaptive feedback control and noise
• reduction without overtaxing taxing the processor.
• They had to provide up to 16 channels of processing with overtaxing the processor.
• They had to limit digital noise by using a sufficiently high resolution and sample rate without over taxing the processor.
  They had to provide pre and post conversion filtering in almost zero space.
• They had to balance DSP function versus power drain.
• They were limited by the flea power afforded by a tiny battery.
• They had to produce it for less than the highest cost the market would bear.

The fixed limitations of size and power tend to cause the parameters of a complex design like this to resemble a water-filled balloon. One parameter has to be sacrificed for another. You push the water-filled balloon in here- it comes out there. You can only do so much.

Its amazing they did as well as they did. But the trade-offs left us with a marginal result. Almost none of the functions could be realized at full potential. Some were passable some were not. One user is happy the other is not. Not all the desired programmable features could be offered.

The resulting physical assembly was far from industry standard practice (far from mass-producible) so no assemblies/components/methods could be borrowed from mainstream consumer products. Result: costs skyrocketed.

It was a nice try but if the product concept was made a little more realistic their job would be a little easier and the results could be greatly improved.

How can we alter the constraints placed on the engineers so that we get a better result? Lets first list some goals then decide which constraints can be relieved. Here are some broad goals:

• Lower cost. Target cost to consumers is $1000 for equipping two ears (not including any professional fees).
• DSP circuitry is to be industry standard IC packaging and employ industry standard DSPs.
• DSP power supply is to be increased to ten times the currently available power (energy).
• Processing functions are to be optimized given the much-increased DSP power.
• State-of-the-art rapid adaptive feedback cancellation is to be implemented.
• Ear mounted devices are to be standardized - uniform electrically and acoustically for all users.
• Ear mounted devices are to offer option of custom or non-custom ear fitting.
• Remote (meaning not at the ear device) control of volume and modes is to be added.
• Remote (meaning not at the ear device) real time trimming of response by the user is to be added.

The goals above envision a hearing system that includes earsets that are low cost commodities (not programmable). Because they are the part of the system most susceptible to wear and tear, they can be easily and economically replaced. The goals suggest a third device that is programmable and that contains the controls and brains of the system.

It is not likely that any of the present industry standard hearing aid configurations would ever meet the goals listed above. So how can we change the product definition so enough constraints are lifted to meet the above goals?

There seem to be at least a few possibilities as always some desirable, some not.

Simplest Configuration:
Earsets (each with a speaker and microphone) are connected to a pager-size unit by wire (body aid style). The body unit contains all the controls and brains of the system as described above. It does not contain a microphone.

Advantages:

• Lowest cost (one DSP serves two ears)
• Very Low Maintenance (replace wired earsets)
• Full DSP performance.
• MP3 Type Styling Options

Disadvantages:

• Body aid stigma
• Wire Nuisance 
• Body Unit Mounting Issues

Wireless Configuration:
Earsets (each with a speaker and microphone) are connected to a pager-size unit by wireless link. The pager-size unit contains all the controls and brains of the system as described above. It does not contain a microphone. This unit can be located anywhere within 30 feet of the user (for example: in a pocket book, in a drawer, in a briefcase)

Advantages:

• Low cost (one DSP serves two ears)
• Conventional BlueTooth wireless interface (2 full-duplex channels)
• Low Maintenance (Replaceable standard wireless earsets)
• Full DSP performance

Disadvantages:

• The pager-size unit must be recharged weekly.
• The Earsets must have batteries replaced as hearing aids now do.

Wireless BlueTooth/Cell-Phone Configuration:
Earsets (each with a speaker and microphone) are connected to a special cell phone (or alternatively a PDA) by a standard BlueTooth link. The cell phone contains all the controls and brains of the system as described above. (It does not contain the hearing aid microphone.) The cell phone can be located anywhere within 30 feet of the user (for example: in a pocket book, in a drawer, in a briefcase). You can use the system as a hearing aid or use it to make cell phone calls with a perfect HOH interface.

Many cell phone designs already contain lots of DSP power and a full-duplex BlueTooth wireless interface used for hands free operation with general-purpose wireless earsets like the Jabra FreeSpeak or the Nextlink earsets. Changes to such phones for hearing aid operation may be mostly firmware changes.

Advantages:

• Low cost (one DSP/cell phone serves two ears)
• Conventional BlueTooth interface
• Low Maintenance (Replaceable standard wireless earsets)
• Full DSP performance

Ideal HOH interface to Cell Phone

Disadvantages:

• The Earsets must have batteries replaced as hearing aids now do.

Summary:
It may be feasible to build hearing systems using electronics-industry standard components and techniques and in doing so hold the equipment cost of provide state-of-the-art hearing aid to consumers to a target of less than $1000. In the process it would appear that substantial performance improvements could be realized due to an order of magnitude more DSP power.

Because the earsets can be standardized (same from user to user) and because they can be sold in other markets, economy of scale should drive down the cost. Because the earsets (not the processing unit) are likely to be the items needing replacement every three years or so, the system can be maintained economically for a long time.

A BlueTooth compatible cell phone could swallow the hearing aid processor function and providing further economy while extending the hearing aid function to cell phone calls in an ideal way.

The BlueTooth link could also be extended to other audio connections including standard telephone service.