Preamp IC for hearing aid

[aetosa]

I've made a hearing aid which uses LM324 (the one of it's 4 opamps) as a microphone preamplifier.
It works at 3 volts.
Voltage amplification at the output of LM324, approximately 45. That's 33db I think.
Which microphone preamplifier IC would you suggest as a replacement of LM324?
My goal is to make a smaller circuit with fewer parts.

I think in mobile phones must be used a variety of preamplifier ICs.

 

[KlausST]

Hi,

look for low noise, low voltage rail to rail Opamps.

Klaus

 

[aetosa]

But I don't want an opamp, I'm allready using one, the LM324 (SMD) (ok, if I can't find something better I 'll go for something like **broken link removed**).
It needs 14 parts (Cout not included).
I want a preamp IC which needs less parts 'around' it to function.
I want to make the size of the circuit smaller.

Something like .

 

[KlausST]

Hi,

ahh, i see...

did you check MAX9814

or MAX9812.

Klaus

 

[aetosa]

Yes I did.
In fact they were the very first ones that seemed to suit exactly my needs.
Well...not exactly.
They have less amplification (20db) than the one I need (33db).
The importer brings only MAX9812HEXT+T which is for 4.5V to 5.5V.
And not MAX9812L which is for lower voltages (2.7V to 3.6V).
It would be perfect if it didn't have these two small...handicaps.
(of course, ok, I could take it **broken link removed**)

- - - Updated - - -

I guess I 'll go for **broken link removed**.
But I have some questions.
Pin 5 (Vcc) is the power supply and pin 2 (Vdd) is the GND?
I 'm thinking of making a simple circuit with small count of external parts like this (it's from LT):


What do you think?

They 're using a specific Panasonic microphone, maybe it has grater amplification (or something else) from a common condenser mic?

- - - Updated - - -

The preamp that I want to replace is this (works fine):

 

[KlausST]

Hi,

if you use a single power supply then 5 = VCC, 2 = GND. Mind to bias V+ input with about half VCC.

Klaus

 

[aetosa]

I guess % is pin 5!

 

[Audioguru]

The LM324 and LMV321 are much too noisy (hiss) to be a mic preamp. The Maxim ICs have even more noise but maybe they hope a deaf person cannot hear the hiss.

You posted the datasheet of the LMV321 from ST Micro who are Italian (confusing Engrish). Instead you should look at the datasheet from the American inventer who is National Semi (now bought by Texas Instruments).

Notice that the LT circuit uses TWO batteries, one is positive and the other is negative. The supply to the mic through R1 is not filtered so it might cause problems.

 

[KlausST]

Hi,

to your schematic:

Really a lot of parts for the function...

Klaus

 

[Audioguru]

The LM324 and LMV321 are much too noisy (hiss) to be a mic preamp. The Maxim ICs have even more noise but maybe they hope a deaf person cannot hear the hiss.

You posted the datasheet of the LMV321 from ST Micro who are Italian (confusing Engrish). Instead you should look at the datasheet from the American inventer who is National Semi (now bought by Texas Instruments).

Notice that the LT circuit uses TWO batteries, one is positive and the other is negative. The supply to the mic through R1 is not filtered so it might cause problems.

The mic circuit with the LM324 (it has 4 opamps but only one is used) has a 47nF capacitor at its input that cuts hiss and cuts all frequencies above only 1.8kHz (it muffles the sounds).
The Panasonic electret mic was excellent but is now obsolete. It had a very flat frequency response and the same output level of all other electret mics. A modern electret mic is a condenser mic with a 48V permanent charge in its electret material. An old condenser mic needs an external 48V power supply.

 

[aetosa]

Audioguru which opamp should I use for less noise?

The circuit of LM324, that I posted above, is from the **broken link removed** (page 7, on the right) (except for R6 and C44).
The 47nF capacitor (C6) at Mic's input is for minimizing the Larsen effect.

In the new circuit, can you guys help me remove some of the unnecessary parts?
This is what I've done so far with LMV321: (R3 - R4 form the voltage divider for pin's 2 bias)
(Correction: in the LM324 circuit, C10 is in fact 100uF, electrolytic, bulky , that's something I'd definitely like to get rid off in the new circuit)

 

[Audioguru]

Your LM324 circuit with the 47nF capacitor parallel with the mic cuts all high frequency audio. Is that what you want? Acoustical feedback howling will still occur anyway but at lower frequencies if the mic can hear the speaker.
The LM324 is noisy, has crossover distortion and has poor high audio frequency slew rate because it is LOW POWER and IT IS OLD.
Only a few audio opamps will work from a supply as low as 4V to 5V.

Now you show a noisy LMV321 opamp as an inverting amplifier that has an input impedance that is so low (2.2k ohms) it reduces the output from the mic. Also R1 does nothing and can be replaced with a piece of wire.

 

[aetosa]

Your LM324 circuit with the 47nF capacitor parallel with the mic cuts all high frequency audio. Is that what you want? Acoustical feedback howling will still occur anyway but at lower frequencies if the mic can hear the speaker.

Let me tell you what happens in real use of the device.
Without the 47nF the Larsen effect appears very easily, let's say when the earphone is in a distance from the mic of 50cm(or less).
With the 47nf it appears when the earphone is in a distance from the mic of 5 cm.
Of course if there is any better way to achieve this, it's welcome.

After what you wrote, I think LMV321 is rejected.
But, which opamps (or mic preamps) would you suggest?

 

[KlausST]

Hi,

I have good experience with MAX4251.

Klaus

 

[aetosa]

I can easily find these: MAX4250EUK+T MAX4252EUA+.
(MAX4251 from the internet, so it'll be the last choice)
MAX4250 isn't it about the same?

BUT...I think **broken link removed** with AGC would be a far better choice.

Or why not a low priced **broken link removed** ? Is this one 'noisy'?
It has high impedance inputs: "inputs are high-impedance CMOS inputs".
Edit:
LM324: Vn Equivalent input noise voltage: f=1kHz 35nV/√Hz
MCP6001T: Input Noise Voltage Density eni: f = 1 kHz, 28nV/√Hz
It's less 'noisy' than LM324, so it's satisfactory from that point.

OR **broken link removed**:
1. The non-inverting and inverting inputs are high impedance CMOS inputs .
2. input Noise Voltage Density eni 19nV/√Hz f = 10 kHz

What do you think?

Edit:
Finally I'll go for LMV721, I think that's IDEAL, it even has a mic preamp in it's first page!

 

[Audioguru]

I agree that the LMV721 is ideal for your hearing aid. It is even available in a tiny package that can barely be seen.
I didnt look at the copy-cat Italian datasheet but National Semi also shows a mic preamp. The preamp is inverting with a fairly high input impedance of 10k. The mic is powered from 5k and the battery bypass capacitor is not shown.

 

[aetosa]

I didnt look at the copy-cat Italian datasheet but National Semi also shows a mic preamp. The preamp is inverting with a fairly high input impedance of 10k. The mic is powered from 5k and the battery bypass capacitor is not shown.

Which opamp are you refering to?

 

[Audioguru]

The LMV721 opamp was invented by National Semi (National Semi was recently purchased by Texas Instruments).
ST Micro and a bunch of other semiconductor manufacturers copy the IC and print their own datasheets for it.
Here is the mic preamp schematic by National Semi and Texas instruments:

 

[aetosa]

That's the mic preamp I was talking about. That's the one I'm going to make.
Anyway thank you all for your help!

 

[aetosa]

Finally I made this circuit. (It's the one from the datasheet of LMV721, except for C2 and C22, I added them (and mic, mine is a condenser mic). Without the C22 it gives me the impression that it stops and restarts giving a 'popping' annoying sound. Like an oscillation of 2Hz)

But sounds lousy.
Now I remembered that the same had happened when I had made this circuit with LM324 too. And that was the reason I made the 'complicated' version found in the second image of this post.

Why is this happening?