Low Voltage Audio IC Amps - have I got this right?!

[HankMcSpank]

A cursory glance at the majority of low voltage auido amp ICs, shows they often apply a voltage gain of about 20.

So if working off a 6V supply (4 x AAs), you can likely only expect the IC to get within 0.8mV of rails, so that's 5.2V peak to peak - but that's for fresh batteries, therefore let's say the batteries have been in use for a short while (let's say they're down to 5V)- so max possible is about 4.2V peak to peak?

Ok, so working backwards, if most output ICs apply a voltage gain of x20, then the largest input you can deliver into on in my above example would be 4.2V divided by 20 = that's an input signal of just 21mV?!! Eh?!!

Now surely that just puts the signal way down in amongst the noise?

Aren't there any low voltage audio amp Ics that have no such chunky voltage gain - thereby allowing a larger input signal into it (that can claw its way above the noise!)

Or have I missed something here?

[macboy]

Hm. I get 210 mV, not 21 mV. That's 420 mV peak-peak, or 150 mV(RMS).
Generally, line level signals are considered to be 2 V peak-peak or 0.7 V(RMS). That's a difference of 14 dB, but that doesn't necessarily mean that your S/N will suffer by that much. Your source can be line-level, then you attenuate to the appropriate level for the amp. The noise in the source signal is attenuated too. As long as you don't add a lot of noise back in by using a bad amp design (e.g. using very high resistances) then you will be OK.

[jcx]

today there are "unity gain stable" op amps with 100+ mA output for low ps V and you are free to set the gain as you want

most are surface mount with exposed "power pad" that is hard for a diyer to solder easily though


many battery power amps will probably use NiMH rechargables which start from ~1.2 V/cell and are often rated for 0.9 V discharged

PowerGenix NiZn rechargables have a little higher cell V http://www.powergenix.com/docs/powergenix-specs-aa.pdf

[wakibaki]

What's your point?

These things are built into applications. Other devices. Gain can always be lost. It's the responsibility of the designer to ensure that the device is not driven into clipping, and that the S/N is not compromised, taking into account variations in the supply voltage. There are well-defined mathematical modelling tools for the noise performance of amplifiers (and attenuators) in cascade.

There are a lot of trick PSUs for battery apps too, and more daily.

w

[panson_hk]

You can set the gain of most audio op-amp reasonably freely.

[HankMcSpank]

Oops I can't divide (200mV is the right answer!)

But wait, my example of 4.2V was already max peak to peak voltage swing on the audio amp output...dividing by 20 yields 200mV peak to peak as an input (or about 70mV RMS!)

What's my point? Nothing other than I'm trying to grasp some fundamentals. And yes, I realise gain can be lost! It;'s just that when you have say 4V of dynamic range, to be working at 200mV peak to peak into the audio amp IC seems rather low - I'm guessing most would use a preset/trimpot just before the audio amp & set for maximum preamp dynamic range (yet avoiding power amp clipping)

BTW the point I was getting at was more the characteristics of audio IC amps more than opamps - specifically, why - in the main - do they have such chunky voltage gain? Why not say have an audio IC amp with a voltage gain of x1 or x2 with a 5V supply - you can then pump a signal level of say 4V peak to peak into it from the preamp? (serious line of questioning as I'm unsure why they design them with such voltage gain?)