I understand that we don’t want to add more. We can compensate for the known added characteristics of a amp+receiver setup, though.
If we know how a signal is transformed by a setup, we can compute the needed input to generate the wanted output.
IIRC, if I have computed the needed output signal (O(s) usually in the Laplace domain) and I known the transfer function (T(s) also in the Laplace domain), I can construct an input function, (I think it was something like I(s)=[O(s)+T(s)]/T(s),) transform it back to the time domain and offer it to the amp. That way you get clear signals from receivers if you generate the sounds digitally.
You could construct an ‘analog computer’ that transforms the signal electronically to generate the needed I(t) from a wanted O(t), but it would be for a unchanging setup. I guess that is the reason why some amps sound awful with some speakers and great with others.
As the receivers of the Oticon can deliver up to 137 at the high frequencies, it won’t be a matter of having too limited power.
Question that remains for me is whether the measurements can compensate enough for the changing setups with each different ear canal.
Lack in that department would explain why some people prefer Oticon, some Widex, some Phonak, etc., etc…
So if it is generally held that power receivers sound less for people with only severe hearing loss, it seems to me that some signal processing is not optimized: “Presuming every user has profoundly bad hearing that part remains constant.”
I really wish the info on these receivers was in the manual! I wish that next to the expert systems we could get at the backbone of the hearing aid programming. We really have a long way to go before it’s Christmas, I guess.
if I know the transfer function…, I can construct an input function, …transform it back to the time domain and offer it to the amp. That way you get clear signals from receivers if you generate the sounds digitally. You could construct an ‘analog computer’
Yes, all that can be done. To correct frequency response in speakers and rooms we used equalizers, from hand-built analog to graphic analog and today usually digital. Distortion can also be corrected (or added!). In fact there is a thing “Amp Modelers” which is a smooth guitar amp and a processor programmed to emulate the response and distortion of many “classic amps”. 5 kinds of Fenders, 6 kinds of Marshalls, 3 Ampegs, 4 Gibsons, at the touch of a button.
Adding distortion is easy because most distortion is a “falling short”. The signal rises to what should be 24 volts, the amp can only do 20 Volts, we can easily replicate this smash-top wave on another amplifier with much more than 24V max output.
Cancelling distortion is much harder. In electronics there is much use of Negative Feedback which pushes the amp when it tends to fall short. But it will not increase the MAXimum output of the amplifier. I have a 148HP engine in my car. Pushing harder on the pedal, or putting a booster on the pedal, won’t get me to 200HP.
I really wish the info on these receivers was in the manual!
Yes; speaker systems generally do not cite the specific info we really need. There are/were some measurement issues: few labs really had the facilities and calibration to get numbers which would reproduce in other labs. Mostly E-V developed their speakers by comparing with a JBL in their test lab, before that JBL tested against Magnavox in James’ garage; relative numbers, not numbers you could publish. Much everyday work is done in unrealistic couplers because the really good couplers are costly and fussy.
And for most complaints of “clarity”, we do NOT want the MAXimum output. We want Gain at low and medium sound levels. Gain is more often limited by feedback. We can’t arbitrarily turn-up-the-gain: it will squeal. And it most often squeals around 3kHz, which for us ski-slope folks is critical to speech. Yes, digital dips and phase-shifts can allow more gain before squeal. It may be these algorithms that bother some people. Note that for music-only listening we are advised to turn-off anti-squeal and accept the lesser gain.
Reading through the not-great paper on receiver , I can’t help but feel that using a receiver with greater MPO is better. The reduction in size has a negative impact on receivers and the analogy with speakers falls short, because design restrictions that have nothing to do with optimal sound design drive their size beyond what is best for sound production.
However, I did spot this curve that clearly implied that receivers have a problem with getting enough at higher frequencies:
The frequency curves in the technical specs of the PP and miniRITE also drop off at higher frequencies. However, in the technical specs I can’t find frequency response curves for the different receivers for the miniRITE. It does show that the BTE PP with a tube outperforms the miniRITE everywhere, while the slimtubes and the miniRITE are both better at different frequencies.
It does give the total harmonic distortion for an input of 70dB SPL (which is also not informative for piano…). Oddly enough it only goes up to 1600Hz. But at 1600Hz the BTE outperform the miniRITE by 1%(!!) Yes at the lower frequencies the BTE has thrice as much distortion as the miniRite. But before it hits the speech area it is already better!
So it seems that my experience with the Bolero SP’s giving a less distorted sound for the piano than the OPN1 miniRITE is not as subjective as I would have thought.
All this is dependent on the technical specs of Oticon not differentiating between the receivers 60, 85 and 105, off course. I still wonder if someone has any info on the frequency curves of the receivers. The not-so-great article implies that the only difference between low- and high-MPO receivers is the point at which the can’t cope anymore. So also in that text there is no discussion on the frequency aspect.
I suspect if you hunt around there will be specs for the other power levels besides 85, but here is a more general document that includes the 60, 85, 100, and 105 power levels.
For more lows or more highs? They are different problems. Yes, more bass suggests a larger driver; for highs small is less of a problem but some models are just too small for big anything.
Sonion’s site lists hundreds of receivers and I can’t readily find what I might be looking for. (I’m sure when they know you will be buying a bunch, they know their stuff and would be happy to steer you.) Knowles’ datasheets page seems broken this week but this paper Knowles HC series high-output receivers3 has more data, not as complete as I would like. Note that many of these tests are run with much higher drive than we can get from a single-cell aid. Both and will be lower.
Historically “high power” meant ‘for severely deaf who could not get speech’, and the focus was to get to 4kHz. “Good piano” is not something 98% of the market is clamoring for, so is more a specialty product. (The in-ear monitors used by stage musicians nominally DO aim for such highs/heights, but are not used in the all-day hearing aid products.)
I just re-looked at your . It is very different than mine. No ski-slope. Low all over, with rise at the extremes.
I assume you have investigated Conductive Loss. (I am not an expert on that.)
As an audio-guy, and you for piano, I wonder if you would do well with a box on neck-strap, 2 mikes, stereo amp (alternatively one of those micro-recorders with stereo mike on top), and good hi-fi headphones. Almost no EQ needed, just a volume control. Awkward for walking around life, but you could keep the kit at your piano or carry it to gigs.
@PRR It used to be a cookie-bite slope. Piano sounded awful, whatever I did.
However, with the years the extreme’s have been coming down and I can now use my Bolero SP’s to amplify the piano quite well. Sometimes I forget to re-tune after changing thin-tubes and I get a distortion a some notes, sometimes the piano needs to be re-tuned, but most of the time I can play the piano. Even up to such levels that people can complain about loudness!
Frequency Range
First thing that really stands out for the miniRITE is that the frequency range decreases when going from receiver 60 to 105. For the OPN1 is goes from an extreme 9200 to 7800 and for the OPN2-3 they go from 7500-6500. So it is worse to wear lower OPN models than to have heavier receivers! These values in the TUI are the 2cc values given in the technical data. The ear simulation values are 300-400 Hz higher.
Looking at the BTE with a 105 receiver you see no difference in range for all OPN models. However, the top of the range is capped at 7000Hz.
Comparing that with miniRITE, only the OPN2-3 with a 105 receiver is worse.
So clearly the BTE is behind the miniRITE for the OPN series.
Total Harmonic Distortion
2cc: For all miniRITE models with all receivers the Total Harmonic Distortion () for 2cc is given as <2% for 500, 800 and 1600Hz. Looking at the OPN BTE 105 the 2cc is 3% at 500Hz. Clearly, 2cc is a rather less informative test than ear simulation, because those values show more difference between the models.
Ear Simulation: This is where the technical data gets odd. All values are for (500, 800, 1600)Hz. The 105 receiver for the miniRITE has (<2%, <2%, <3%), while the 100 receiver has (<7%, <4%, <2%). This breaks the expectation somewhat, since the 85 and 60 both have (<2%, <3%, <2%). So while the 60 and 85 have most trouble around 800Hz, the 100 has most trouble at the lowest frequencies and the 105 has somewhat more trouble with 1600Hz. Comparing this with the BTE with a 105 receiver (7%, 5%, <2%) it is clearly the worst of all at lower frequencies but is already better than the miniRITE 105 at 1600Hz.
So I speculate that the 105 miniRITE re - or later designed to behave at approximately the same quality level as the other receivers with only a decrease of 200Hz of the high end frequency range for 2cc compared to the 100. Checking with a local store I see Oticon receivers of 65(!), 85 and 100, so it seems that the 105 are less commonly used.
Curves
The curves are all with a different vertical scale. So looking at them will give you the wrong impression. Obviously, the receivers with a smaller number fall off less rapidly at higher frequencies.
It also seems that the THD values are taken at the part of the curves where they are most smooth. The 2cc curves are indeed smoother at these values than the ear simulation. Would have been more interesting to see the THD values in the speech area of course. So the given THD data are best taken as an indication of the maximum performance, instead of as representative.
For the OPN1, if you look at a 7000 Hz, the miniRITE 60, 85, 100, 105 and BTE 105 reach a full on gain for 2cc of resp. 12, 28, 41, 38 and 40 (or 30 with corda miniFit). For 8000Hz it’s 9, 31, 30, 38 and 20 (extrapolating the straight line beyond the curve).
Stunningly enough looking at the OSPL90 (Outside sound pressure level of 90dB, I guess) the values for 9000Hz are resp. 98, 108, 95, 100 and 116 (or 110 with the corda miniFit). For 9500Hz the values are resp. 93, 103, 95, 93 and 108(both normal and miniFit).
This supports @PRR 's argument that driven at the same sound level larger speakers don’t necessarily have lower quality of sound. You can’t get 132 at this frequency so the range of the larger receivers is limited for those that have profound hearing loss. However, for a (moderate-)severe hearing loss the BTE 105 receivers can output the same as the miniRITE 85 receivers!
Comparing the smoothness of the curves, I do see more hard to model curves for the BTE 105: at lot of local minima and maxima, while the curves are smoother and keep getting smoother if you go from to 85 and 60. If that can be taken as an indication of expected THD, you can expect increasing distortion going from miniRITE 60, 85, 100, 105 to BTE 105. Looking at the increase in local max/min above 5000Hz, I would guess that they are all still pretty high in THD.
However, the argument of maximum power output limiting the use of power aids for higher frequencies doesn’t seem to apply to this data. The BTE 105 outperforms the miniRITE 85. The optimum in terms of in high frequencies is BTE 105 followed by miniRITE 85, 100 and 105 and lastly the 60. However, the BTE does suffer from a dip in power at 8000Hz. Only 110dB for the ear measurement, while the miniRITE 85 dips earliers at 6000Hz and goes to 113dB at 8000Hz.
Concluding based on the data I should choose the OPN1 miniRITE 85 over the BTE 105.
However, as I tested that OPN1 85 and it couldn’t handle the piano well enough, while my Phonak Bolero B90 SP’s do, I do have to wonder whether that extra oompf that the BTE brings isn’t needed for piano. Too bad I can’t test it, yet! Would it be odd to buy those ’s just to be able to test them?
Checking with a local store I see Oticon receivers of 65(!), 85 and 100, so it seems that the 105 are less commonly used.
I suspect one of the reasons for that is that the lower power receivers share the same form factor, while the SP or HP receiver is a larger unit and has to be built right into a custom mold. They probably do not come from the store, but come with the custom mold.
Excellent technical review of the aids by the way!
Concluding based on the data I should choose the OPN1 miniRITE 85 over the BTE 105. However, as I tested that OPN1 85 and it couldn’t handle the piano well enough, while my Phonak Bolero B90 SP’s do
I would agree on the choice of the 85 over the powerful . I suspect what you are hearing in the RIC model is feedback suppression. There may also be a peak limiting feature that is impacting the output. Signia/Rexton call it sound smoothing. In a music program sound smoothing and auto noise reduction are typically turned off. Feedback suppression is either slowed down or turned off.
Almost missed this. A later edit, I presume? Thanks, it gives some background.
p1: “Additionally, advanced technology originally developed for the FC receiver series is incorporated to the HC receiver design.” So they developed their high-output receivers after the lower powered FC series. The same seemed to hold for the receivers used in Oticon OPN’s.
p.4&p5: “Figure 5 (previous page) shows the maximum output level that is achieved with
a low-impedance source (voltage drive) before 10% distortion is reached.”
Apparently the power output at which the receivers have 10% (Total Harmonic Distortion) increases after 1000Hz. Conversely, for the same power output the percentage THD will drop after 1000Hz. If the same holds in general, than the reason that Oticon only gives values for 500, 800 and 1600Hz is that after 1600Hz it stops being an issue! That assumption generates a rather dramatic conclusion, that destroys the assumption that non-smoothed curves have higher THD.
p6: “The green curve shows how a shorter tube (5mm x 1mm) raises the resonance frequencies, especially on the second resonance peak.”
The peaks are associated with resonance frequencies. Apparently, their shift in frequency is interesting to engineers, not their abundance. That said, the curves in the article are all much cleaner than those of the Oticon OPN’s.
Without further info, the comparison of Oticon’s data shows that the high frequencies have enough output also in the high frequencies and this article implies that THD is not an issue in those frequencies, too.
Yes. While OSPL90 is taken “under specified conditions”, I believe it reflects the basic efficiency of the several drivers. I replotted to same scale, then re-pasted all on one graph.
The “60” mainly features slightly lower power consumption. This may be the basic curve of a BA receiver with “no tricks”. The others have several <1kHz bass outputs. But at 3kHz you could cover them all with a dot. The 100/105 add a 5kHz 10+ resonance. At 7kHz they are all alike. This is in accord with general loudspeaker design: treble is tough and there are tricks.
You want good output at 4kHz. It is possible the 100/105 drivers suit you better; and that your 90 driver is more-like these types than the 85 in the graph.
Single numbers may guide the maximum useful level, but don’t say what happens on less-than-max levels. For systems without ‘scratch’, THD normally declines to very-small at very-small output. There’s not a ton of data on speaker THD/drive (normally we feed constant tone and plot THD against frequency). But nonlinearity analysis is similar from mechanical systems to simple electronic systems. Here’s typical power amplifier THD vs level.
There’s reasons two tubes don’t mesh well at low level. The BA driver is a single moving system and should be cleaner at low level; I’ve added my guess on that.
The rise from 2% to 10% THD happens in just a few dB. 2% to 0.5% is probably just a few dB drop. So a driver that hits 10% at 100dB may be <1% at 95dB, and “negligible” over the main range 40dB to 90dB. (Yes, a sudden rise of THD on 96dB piano peaks may annoy you. In hi-fi we say “Turn it down!”. However when sitting AT a good piano with full hearing of rumble and tizz, turning down the midrange boost is not satisfying.)
More to ponder: Horn Speakers is really about the drivers used on theater horns. Much of it needs interpretation. These drivers can eat 1000X the power we can get from a . They will make 150dB in a pipe, 100+SPL spread around a theater, 194++SPL if funneled to an ear-hole. So they could do 130SPL with HA power. But they are about the weight of your head. The last graph shows a fall-off above 1kHz similar to HA drivers; even with a quite different electrical drive system we run into mass-of-metal limits at high freqs.
Bottom of Tweeter Test1 shows THD vs frequency of a speaker. THD tends to stay low in the “useful range”. In open air, THD will rise in bass; things are different in the small closed world of a HA.
I could be wrong about this, but I believe in most brands the S, M, and P receivers all use the same hearing aid, so share the same amplifier, with typically a 312 battery. And, I think most of the highest SP or UP receivers use a different body aid, and I suspect larger amplifier, and a larger battery for the increased current draw. Not sure if you can use the lower power receivers with the larger body aid.
Rotterdam, Netherlands
Downeast Maine
The reduction in size has a negative impact on receivers and the analogy with speakers falls short, because design restrictions that have nothing to do with optimal sound design drive their size beyond what is best for sound production.
