Need your opinion, how small gain can I use with LM3886 (how close to "flat amp")?

[danielwritesbac]

Need your opinion, how small gain can I use with LM3886 (how close to "flat amp")?

I need a good opinion on how little (small) gain I can use with lm3886 and still be able to use its full power?

I'd like to do an experiment that involves much more input, much less gain. If I use the minimum gain of 10, would I be overdriving the input before the LM3886 is producing its 68 watts? If so, how much gain do I need in order to stay out of trouble?

Thanks!

 

[theAnonymous1]

It says right in the datasheet the IC needs at least a gain of 10 to be stable. I don't think you will be overdriving the input at full power; it can take a 60V differential input signal when used with an 80V supply.

 

[Nickolay Shvydky]

With appropriate feedback circuit, you can make it work even with gain = 1 (in inverted mode). I can give you exact schematic if you'll tell me what gain you need.

Nick

 

[Will]

With appropriate feedback circuit, you can make it work even with gain = 1 (in inverted mode). I can give you exact schematic if you tell me what gain you need.

Hi there, do publish the feedback circuit for a gain=1 amplifier. That's interesting. Many thanks.

 

[danielwritesbac]

I'd like to see it at 2x, or 4x, or 8x--even multiples if you don't mind.

 

[Nickolay Shvydky]

This one for example has a gain = 3.
--------------------------------------------------
Sorry, I was simulating and did not notice your new posts. I'll post tommorow what you need.

 

[danielwritesbac]

Its a good illustration.

My question was really about a non-inverting LM3886.
How little gain can I use there and still have all of its power available, but without overdriving the LM3886 input?

In a related question, can 100R or 1k be used for input load and and NFB? (pre=headphone amp)

 

[Nickolay Shvydky]

Common Mode Input Voltage ( V+ or V- )and | V+ | + | V- | <= 80V. So I think it could work for a noninverting LM3886 too, but distortion will be much higher than in inverting.

 

[Nickolay Shvydky]

In a related question, can 100R or 1k be used for input load and and NFB? (pre=headphone amp)

You mean R3 and R14 ?

 

[danielwritesbac]

Common Mode Input Voltage ( V+ or V- )and | V+ | + | V- | <= 80V. So I think it could work for a noninverting LM3886 too, but distortion will be much higher than in inverting.

Inverting is fine then. I take it that we can invert both the amp and the pre so that it will come out straight?

 

[danielwritesbac]

You mean R3 and R14 ?

I mean R14 (feedback) Is it doable at 100R?
And can one also set the amplifier's input impedance resistance at 100R?

 

[Nickolay Shvydky]

I mean R14 (feedback) Is it doable at 100R?

It is doable, but has no sence, because in this case R14 will dissipate a lot of power (Uout^2/200) and it could cause high distortions too (thermal).

 

[Nickolay Shvydky]

I mean R14 (feedback) Is it doable at 100R?
And can one also set the amplifier's input impedance resistance at 100R?

It is possible, but I did not get you mean impedance of source or R3 ?
Could you please give me some more detailed information about your project ? Is it pre-headphone amplifier ? If yes, why LM3886 ?
Thank you.

 

[danielwritesbac]

The project is in using a higher gain headphone amplifier as the predrive for LM3886. The source is built to drive 100R.

 

[Nickolay Shvydky]

The project is in using a higher gain headphone amplifier as the predrive for LM3886. The source is built to drive 100R.

Ok, and what is a gain and maximum output voltage of your headphone amplifier ?

 

[danielwritesbac]

Ok, and what is a gain and maximum output voltage of your headphone amplifier ?

Gain is 48 and output voltage is 5vac peak, into 100R.

 

[Nickolay Shvydky]

I do not understand why the gain is so high. I think it will work better if you make a buffer-inverter with a gain about 7-10 and maximum output voltage about 10, and will not use your headphone amp for LM3886.

 

[danielwritesbac]

I do not understand why the gain is so high. I think it will work better if you make a buffer-inverter with a gain about 7-10 and maximum output voltage about 10, and will not use your headphone amp for LM3886.

The gain is available from 10 to 48. The gain is high because the source is a digital VCA, which is approximately half of line level.

EDIT: The gain can be reduced at the predrive; however, the "total" amplifier does need from 45 to 48.

 

[Nickolay Shvydky]

The gain is available from 10 to 48. The gain is high because the source is a digital VCA, which is approximately half of line level.

I would make a buffer-inverter with gain about 7-10, then inverted LM3886 with gain about 3. In this case you can connect it to any regular source or to your headphone amp. Also it would have a very low distortions. I guess AD8067could be a good choice for the buffer-inverter.
If you would like to go with headphone amp with no buffer-inverter, I think you need LM3886 gain about 6-10.

 

[danielwritesbac]

I would make a buffer-inverter with gain about 7-10, then inverted LM3886 with gain about 3. In this case you can connect it to any regular source or to your headphone amp. Also it would have a very low distortions. I guess AD8067could be a good choice for the buffer-inverter.

Well, howabout let's make one?