1W audio power amplifier driven by electret ?

[Externet]

Is there such thing among the multitude of tiny ICs new in the market ?

Am after finding, ideally :
Integrated circuit,
Analog amplifier , any class, BTL or not,
For a 3 to 4 V DC supply,
1W or near that power output,
For electret microphone input, or conventional 8Ω cone speaker as microphone,
Monophonic

Have some near candidates in sight, but do you know of something else ?
From the ringer/hands-free audio levels I have seen in cell phones and tablets, manufacturers are incorporating new ICs that could work. In other words, a 'public addressing' circuit in a single chip running on a lithium cell.

 

[Audioguru]

A 4V supply can power a single amplifier that could produce an output of 0.1W into 8 ohms just before clipping (one channel of an obsolete TDA2822M). If two channels are bridged then the output is about 0.3W.

A little arithmetic shows that for 1W into 8 ohms then the speaker needs 2.828V RMS which is 8V peak-to-peak that can be produced by an ordinary amplifier powered from about 10V to 12V.
If an amplifier could drive a speaker that is 0.5 ohms then its power with a 3V supply might be 1W.

Even though the speaker power is low, if the microphone can hear it then there will be acoustical feedback howling. You might need to record the speech then play it back when the microphone is turned off.

A speaker used as a microphone resonates and sounds like a bongo drum.

 

[Externet]

Thanks.
That is the old stuff; now it is surprising to see dozens of audio amplifiers on the market, class D and boosted. Even with DSP processor built in 4mm x 4mm. I am learning to discern among them. The feedback howling is of no concern.
Here is one of maaaany :

----> https://cdn-shop.adafruit.com/datasheets/PAM8302A.pdf

And a selection chart of only one manufacturer :

----> http://www.ti.com/audio-ic/low-powe...s-d/products.html#p89=Class D;Class D Boosted

But still trying to find one that will accept an electret.

 

[(*steve*)]

I've not looked at the modules you've linked to, but presumably there is some specification of the input level. And presumably you know the signal level you'll get out of the mic.

What's the problem with finding an op amp (or even a discrete circuit) that can run from this voltage to supply the necessary gain?

 

[Externet]

Thanks.
No problem on adding a discrete circuit to give the needed gain to an electret and feed an amplifier IC; but with so many new products on the market I wanted to make sure there is not a chip with everything included.
Am restricted in size to 19mm diametre round PC boards that I can stack a couple to a 15mm height pile.
Typical max gain of these IC amplifiers is 40dB = 100X voltage gain.

 

[Audioguru]

The amps you linked to produce higher output power when they are clipping like crazy with 10% and even at 1% distortion. The PAM8302 class-D power amplifier IC even produces non-hifi distortion when its output power is reduced to 0.1W.

The amplifier ICs with "class-H boost" have datasheets that do not say how long it takes for boost to occur. If distortion occurs before the boost is up then I do not know if the boost can be turned on all the time (wasting battery power).

If your microphone and speaker are in the same room then of course feedback howling is a concern.

I doubt that somebody makes a power amp with a built-in preamp for a microphone. You want it to be tiny but what about the size of the volume control between the preamp and the power amp?

 

[Externet]

Thanks.
I will then add a preamplifier section if any extraordinary chip does not surface. No volume control to be implemented, will have fixed high gain.

 

[Audioguru]

Fixed high gain will cause the low power amplifier and maybe the preamp to have clipping with a squarewave buzz for an output whenever the input level is higher than average.
What is the input signal? A voice (at conversation level about 10cm away from the mic)?

 

[Externet]

Hi, thanks.
The input sound is plain conversation level voice picked at a fixed 3cm from lips side by an electret,
preamplified and fed to probably a LM4871 or similar to a tiny 50mm 4Ω speaker with no acoustic feedback path towards the mic.
A manual or remotely controlled volume control is highly undesirable There is AGC modules like

----> https://learn.adafruit.com/adafruit-agc-electret-microphone-amplifier-max9814/assembly

Currently testing.
Gain to be set as loud as possible with hopefully less than 5% distortion for loud voice, without clipping, and for 'normal' conversation, the speaker should be intelligible 8 metres away.

Also playing with ----> https://www.adafruit.com/product/2130

A lot more to tweak...

 

[Audioguru]

A 2" speaker can barely produce an adult male voice frequencies, maybe a woman's voice.
A half decent 3" speaker (clock radio style) has a sensitivity of about 86dB per Watt at a distance of 1m. 80dB at 2m, 74dB at 2m, 68dB at 4m and only 62dB at 8m. The listening location better be dead quiet. Add 10dB for a fairly loud voice and use a 10W amplifier.

I do not know if the Adafruit PAM8302 amplifier uses a real Diodes Inc. IC or a cheap Chinese copy.

 

[Externet]

Thanks.
Am restricted/hancuffed to many things, including only one lithium cell, very tight dimensions...
Speaker candidates if there was any way to prove their chinese specifications :

----> https://www.ebay.com/itm/2pcs-2-inc...eakers-Loudspeaker-For-toy-Game-/263194310526

----> https://www.ebay.com/itm/2pcs-2-inc...rs-Multimedia-Card-Horn-Speaker-/262917098812

----> https://www.ebay.com/itm/2pcs-2-inc...229667?hash=item360a21eda3:g:5YUAAOSwx2dYGVFv

Already testing a couple, will see which one lies the least...

Yes, it is very quiet location.
Perhaps a typographical error ----> "80dB at 2m, 74dB at 2m, 68dB at 4m and only 62dB at 8m."

 

[Audioguru]

Perhaps a typographical error ----> "80dB at 2m, 74dB at 2m, 68dB at 4m and only 62dB at 8m."

Sound level drops 6dB per doubling of distance.
OOps, I see my error. It should be 86dB at 1m, 80dB at 2M, 74db at 4M and 68dB at 8m. 68dB is almost conversation level.

A tiny speaker probably has a peak at a few kHz that the sensitivity is based on (a beeper). If your circuit or your hearing flattens the peak then the sensitivity is much lower.

 

[BobK]

Look fot a horn type speaker to get the volume yoi need at 1 W.

Have ypu ever seen a bullhorn? Very compact, very loud.

Bob

 

[Externet]

Thanks, Bob.
Very restricted in dimensions. Speaker should not be over 40mm thick

Wondering what are the guts here, to canibalize. Perhaps uses a too much integrated circuitry for multiple functions that would be hard to hack :

----> https://www.aliexpress.com/item/Spe...lgo_pvid=49948dff-ae96-4906-b8ee-69c32b2c761d

----> https://www.aliexpress.com/item/Rol...id=407bbfb0-258c-4cca-aa1f-7d0a79bb2df4&tpp=1

----> https://www.aliexpress.com/item/Min...ml?spm=2114.10010108.1000015.7.1dccab43nRUEES

 

[Audioguru]

Why are you looking at cheap Chinese mini speakers that are 3 times larger than you want?
Do you believe their lies about the output being 5W or 10W?

 

[duke37]

Do you believe their lies about the output being 5W or 10W?

The speaker output is not likely to exceed 1mW, the efficiency of even big speakers will not be above 5%.

 

[Audioguru]

The speaker output is not likely to exceed 1mW, the efficiency of even big speakers will not be above 5%.

Nobody is measuring the speaker output, instead we are talking about the amplifier power output.

 

[WHONOES]

Nobody is measuring the speaker output, instead we are talking about the amplifier power output.

Maybe but, they will be related.