#147

It’s probably not possible to have it integrated in the chip. So additional physical space is wasted. What they want to avoid in favor to space for a lager battery.
In addition to that each step up or down of power supply voltages has the downside of additional power loss.

1 Like
#148

A friend of mine designs consumer electronics, so he’s tuned into this stuff. He pointed me to this accelerometer that works down to 1.1V. The spec sheet shows current draws. They look pretty low. The listed applications include hearing aids.

https://www.digikey.com/en/products/detail/analog-devices-inc/ADXL367BCCZ-RL7/16360579

#149

Which means they’ll run off normal batteries too. So the point about making devices rechargeable in order to add this feature is moot.

2 Likes
#150

I sometimes wonder why they don’t power these higher voltage devices via a voltage multiplier.

WH

#151

I recall this starting with a post when Paradise came out, reporting that Phonak says disposable batteries don’t have enough power for the accelerometer features. I have the idea that post was from @Neville?

I couldn’t find a lower-voltage accelerometer before, so I assumed incorrectly that Phonak was referring to voltage. There is one, so my assumption was wrong. Let’s look at power instead.

The data sheet says:

► Ultralow power
► 0.89 µA at 100 Hz ODR [output data rate], 2.0 V supply
► 180 nA motion activated wake-up mode
► 40 nA standby current

Those are seriously low currents, considering that hearing aid overall current draws are measured in milliamps, and a microamp (µA) is 1/1000 of a milliamp, and a nanoamp (nA) is 1/1000 of a microamp.

Figure 43 of the data sheet shows that current draw doesn’t vary much by supply voltage. Since power is current multiplied by voltage, this accelerometer running on zinc-air voltage uses 1/3 of the power compared to running on Li-ion voltage.

So, I still don’t understand Phonak’s explanation why they can’t use an accelerometer with disposable batteries. Perhaps @AbramBaileyAuD could use his good offices to obtain more information.

#152

Might not be accelerometer that is drawing the power. Might be the extra processor power for the software that is required to power the features that use the accelerometer.

Jordan

#153

Let’s compare sound sampling rate to accelerometer sampling rate. Assuming the Paradise frequency range tops out at 8KHz, they have to sample the sound no less than 16000 times per second. The accelerometer’s output data rate is selectable between 12.5 and 400 times per second. So, even if Phonak uses the maximum output data rate, which I tend to doubt, they have to process sound at a far higher rate than they process acceleration data. For this reason, I don’t think it’s a likely explanation.

#154

I’m not talking about the extra battery required to power the accelerometer. My point is that the new speech recognition software (and the Life features) that are using the accelerometer data (i.e Autosense 5.0) are using more processor power to enable these advanced features and are probably drawing more power. Kinda like the way that some apps on your mobile phone chew more power than others.

Only a Phonak hardware and software engineer can provide the true answer but that is my guess.

Jordan

#155

This excerpt from “Fitting and Dispensing Hearing Aids, Third Edition” explains how it works. The accelerometer data (which isn’t all that computationally expensive to interpret, compared to sound) is used to activate modes that already exist in hearing aids.

IMG_8960
IMG_89601920×1440 267 KB

#156

“like most female teenagers, Messina likes to go the mall on Saturday night…and drink her favorite beverage, Orange Julius…” Holy cow, when was this written, 1964?

I still don’t think I’ve heard a plain speaking answer to my earlier question. Do rechargeable aids allow for more features/programs to be integrated into their hard and software, as compared to disposable battery aids?

I’m technically dense. from what I can gather in the discussion, there’s no consensus on this…?

#157

I agree, we are just guessing here.

#158

Going from Phonak to Oticon.

LE Bluetooth is offering me so much more (I think because I also have an iPhone?) then the Classic Bluetooth.

1 Like
#159

I agree. About 9 years ago I moved from Android to iPhone partially because of my MFi aids. At the time I had a Google Nexus phone but needed to change phones because we chose to consolidate the Damon the one carrier where my phone would not work due to a dispute with Google

#160

Informed speculation, like I’m doing, is at least a step above guessing. I can think of some other posters with the technical knowledge to weigh in. No need for circuit design expertise like I stipulated before, because we have the spec sheet for an accelerometer compatible with disposable batteries.

On the narrower question of accelerometer-based features, my plain answer is that they can be integrated into both kinds of aids. I wouldn’t know where to start answering your broad question.

His writing style is corny, true.

#161

On top of which I have to add that even in 2023 I find that in a noisy mall if I go up to a clerk at Orange Julius (um, are they still in biz?) the noise management on my aids is STILL TOO AGGRESSIVE. Likely I’d barely even hear the clerk even tho ALL noise has been dampened down in volume to an irksome LOW.

I’ve been in to see my audi 3-4 times to tweak that noise management. It’s still TOO aggressive. I’m glad I had her set up “dumb ears” (aka my Music program) cuz that one will literally keep everything as loud as it is, and I often have a better chance of hearing someone facing me if the volume of THEIR voice has not been dialed down by noise management.

1 Like
#162

That was one of the few areas where I found the ReSound Omnias decidedly better than the Lumitys trialed. Turning on aerated tap water while standing at my bathroom sink turned on very aggressive noise management that not only suppressed the fizzy running water sound but also damped down the volume and garbled the sounds of my wife’s soft voice with her standing just a few feet away at her side of the bathroom counter. I imagine in Target one could change the amount of noise suppression applied by AutoSense in the Automatic program, but I think what’s missing is better automatic processing of speech in noise in this particular type of situation. The Omnias were much less aggressive (and sudden) in damping down the noise.

BTW, I questioned my use of “damp down,” - but my brain was telling me, “It’s OK to say that…” So, I looked it up: Is It ‘Damp Down’ or ‘Tamp Down’? | Merriam-Webster. A fascinating read, including the reference to eggcorns! @SpudGunner, you’re quite erudite (although some might think that’s a special form of “rude!” :grinning: ). So, you might enjoy that history of word usage info, too.

1 Like
#163

This is the “magic bullet” that I daydream about: find that sweet spot of human speech (dB + Hz) and BOOST it for better speech comprehension in myriad environments.

I also thoroughly enjoyed the discourse on HOMOPHONES in that link - and that’s a new word for me. :smiley:

I find myself torn between “dampen down” and “tamp down” in addition to “hone in” and “home in”, “carome” and “careen” or even “who” or “whom”.

1 Like
#164

I think the real problem is that our HAs are “dumb” and not smart like the manufacturers would like us to believe, the HAs can’t actually tell what’s been spoken/human speech with all the other sounds going on around, plus that human speech constantly changes in frequencies and loudness as we speak, when we get “real AI” this problem will be eliminated, and not to be outdone, the “Smart Glasses” coming out with live captions will be a godsend fot the profound community.

2 Likes
#165

I have suspected as much. And yes, perhaps when AI really gets smart, they’ll figure it out. Human speech goes up and down, but if there is a constant drumbeat, vibration, frequency band used, etc., the AI could “figure” that it’s speech and perhaps increase the clarity or volume.

Until that feature is perfected, I can imagine YEARS of frustration with the AI also increasing volume and clarity of things that aren’t speech.

Altho when you look at that speech banana chart WHAT ELSE is in our human speech frequency? No dogs, dripping water, piano, jackhammer, spoon clinking on a glass at weddings, etc.,). Seriously, how tough can this be? If our country’s security depended upon it, there’d be a working solution right now. :slight_smile:

Hope springs eternal…

#166

Ask for NoiseBlock to be turned down. I had mine off.