For audophiles

[Cursitor Doom]

On Tue, 22 Oct 2019 00:58:30 -0700, tabbypurr wrote:

> exceptionally nice speakers.

Some top-notch manufacturers used to make electrostatic headphones, too.
I haven't seen 'em for a long time, though. The sound quality was
*amazing* but having 600V right up close to your head wasn't a great
selling point for the average buyer, I'd imagine.




--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[bitrex]

On 10/22/19 2:31 AM, Jan Panteltje wrote:

On a sunny day (Mon, 21 Oct 2019 21:42:44 -0700 (PDT)) it happened
tabbypurr@gmail.com wrote in
2147376f-2e89-441c-b6a4-143353e7ae67@googlegroups.com>:

On Monday, 21 October 2019 16:43:16 UTC+1, bitrex wrote:
On 10/21/19 5:17 AM, tabbypurr wrote:
On Sunday, 20 October 2019 19:28:54 UTC+1, Jan Panteltje wrote:

Needed a bit stable headphone amp for my Sennheiser HD201.
Lots of long audio cables from various sources with varying impedances around in the house..

OK, maybe LM380 (yes old days) or LM386, make a peeseebee, potmeter, connectors, housing
$+$+$+$ plus local shipment > 5 Euro (> 5 USD)...

mm ebay?
Sure!
https://www.ebay.com/itm/293190924115
9 $ and free shipping, cannot buy the parts for that (only 1 LM3.. in junkbox).

lucky you avoided the LM386. Sound of those not good.


NT


They're nice for a lot of tasks like solenoid and servos, driving small
brushed DC motors and other applications other than audio. The DIP-8
version LM386N is very rugged and can dissipate a fair amount of power
(a small copper heat sink thermal epoxied to the top helps.) It can
drive a 30mm brushed DC motor like in an RC car.

I think it's the world's most popular audio power amp IC. (Maybe the one with the most distortion too.)


NT

Did some more tests with the Chinese thing last night.
I think the damping on my headphones is not so good, that results in some
suspicious 's' sounds in some speech.
It is a large headphone that behaves a lot like a dynamic speaker...

So, maybe I will add a LM380 or LM386 to compare,
and see if I can get any of the chips bitrex mentioned.
Space enough in that plastic housing.
I have to keep resisting buying a couple of Quad electrostats on ebay for a thousand or so..
I really like those speakers (were also used as monitor in the studio sound control room),

They had, in their amps an interesting feed forward (IIRC) distortion compensation.
Build a circuit like that once for about 100 kHz for an inductive loop system.

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

 

[bitrex]

On 10/22/19 3:58 AM, tabbypurr@gmail.com wrote:

On Tuesday, 22 October 2019 07:32:25 UTC+1, Jan Panteltje wrote:
21 Oct 2019 tabbypurr wrote in
2147376f-2e89-441c-b6a4-143353e7ae67@googlegroups.com>:
On Monday, 21 October 2019 16:43:16 UTC+1, bitrex wrote:
On 10/21/19 5:17 AM, tabbypurr wrote:

lucky you avoided the LM386. Sound of those not good.

They're nice for a lot of tasks like solenoid and servos, driving small
brushed DC motors and other applications other than audio. The DIP-8
version LM386N is very rugged and can dissipate a fair amount of power
(a small copper heat sink thermal epoxied to the top helps.) It can
drive a 30mm brushed DC motor like in an RC car.

I think it's the world's most popular audio power amp IC. (Maybe the one with the most distortion too.)

Did some more tests with the Chinese thing last night.
I think the damping on my headphones is not so good, that results in some
suspicious 's' sounds in some speech.
It is a large headphone that behaves a lot like a dynamic speaker...

So, maybe I will add a LM380 or LM386 to compare,
and see if I can get any of the chips bitrex mentioned.
Space enough in that plastic housing.

The 386 has a very characteristic sound. I don't know what type of distortion causes it - I wonder if it might be slew rate limitation, but don't know.

It's just distortion I think. It's a very simple 3 stage amp inside the
diff pair doesn't even have a current source in the tail and its open
loop gain is relatively limited. THD is already at 10% with a 9V supply
pushing 500mW.

I have to keep resisting buying a couple of Quad electrostats on ebay for a thousand or so..
I really like those speakers (were also used as monitor in the studio sound control room),

exceptionally nice speakers.


NT

They had, in their amps an interesting feed forward (IIRC) distortion compensation.
Build a circuit like that once for about 100 kHz for an inductive loop system.

 

On Tuesday, 22 October 2019 16:47:27 UTC+1, bitrex wrote:

On 10/22/19 3:58 AM, tabbypurr wrote:

The 386 has a very characteristic sound. I don't know what type of distortion causes it - I wonder if it might be slew rate limitation, but don't know.

It's just distortion I think. It's a very simple 3 stage amp inside the
diff pair doesn't even have a current source in the tail and its open
loop gain is relatively limited. THD is already at 10% with a 9V supply
pushing 500mW.

Of course it's distortion, the question is what sort.


NT

 

[Jan Panteltje]

On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
<user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low distortion.
I have used those before in a (not HiFi) monitoring system in a radio transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a lot... not sure if I can ..
Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover compensation?
Output current is a bit low though.

 

[piglet]

On 22/10/2019 20:00, bitrex wrote:

On 10/22/19 2:02 PM, Jan Panteltje wrote:
On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low
distortion.
I have used those before in a (not HiFi) monitoring system in a radio
transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a
lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover
compensation?
Output current is a bit low though.


Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

Did that morph into Peter Walker of Quad's current dumping topology?

piglet

 

[bitrex]

On 10/22/19 3:00 PM, bitrex wrote:

On 10/22/19 2:02 PM, Jan Panteltje wrote:
On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low
distortion.
I have used those before in a (not HiFi) monitoring system in a radio
transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a
lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover
compensation?
Output current is a bit low though.


Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

With grounded load: <https://www.tubecad.com/2004/12-2-2004_Sandman11.gif>

 

[bitrex]

On 10/22/19 2:02 PM, Jan Panteltje wrote:

On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low distortion.
I have used those before in a (not HiFi) monitoring system in a radio transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover compensation?
Output current is a bit low though.

Another interesting amp topology is this:

<https://www.tubecad.com/2004/12-2-2004_Sandman2.gif>

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

 

On Tuesday, 22 October 2019 10:02:01 UTC+1, Cursitor Doom wrote:

Some top-notch manufacturers used to make electrostatic headphones, too.
I haven't seen 'em for a long time, though. The sound quality was
*amazing* but having 600V right up close to your head wasn't a great
selling point for the average buyer, I'd imagine.

I once built a pair of electrostatic headphones to a design published
in Wireless World. It involved rubbing Aqua Dag (graphite colloid)
onto cling film to make the high resistivity diaphragm. They were
driven with a hybrid transistor-valve amplifier. The sound was
superb, but they were a bit impractical.

John

 

[Rick C]

On Monday, October 21, 2019 at 5:23:15 AM UTC-4, Jan Panteltje wrote:

On a sunny day (Mon, 21 Oct 2019 01:22:08 -0700 (PDT)) it happened Kill
Allison <pallison49@gmail.com> wrote in
7b9108f7-5a0e-43ee-a99d-d69f418aa0b1@googlegroups.com>:




But all the time I was wondering about the schematic:
http://panteltje.com/pub/chinese_headphone_amp_schematic_IXIMG_0191.JPG

** Not Found Error

You need a computah for that.


Later ...
I remembered I had an downed under I{ that could not access my servers at that time those ran at home
As sysadm I contacted them on behalf of their customer (who complained to me)
and had them check their routing

But as you are such a genius you could do the following in a real Linux computer:

Type
host panteltje.com
that should show:
panteltje.com has address 37.148.205.129

Then type
traceroute 37.148.205.129
see where it gets stuck

Then look up the IP where it gets stuck.
(ip_to_country)
Very likely down under, problem you get when everything is upside down and water flows all over y'r hat.


Then contact your ISP's sys administrator, complain, show them the log you just made
and hope they know enough to fix it.

Bit isolated down there I guess without access to EU internet

If you were a bit polite I would do it 4 you, but I leave the chance to you to actually learn something more
than insulting.

killfile

I don't think really know Phil. That *WAS* Phil being polite!

--

Rick C.

- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209

 

If you even knew what an audiophile is you would not use those chips, or if you're really good NO CHIPs at all.

That fucking thing is like what I would put on a test and make the kids figure out the gain. All, voltage, current and power.

Easier said than done but then I am a real prick.

 

On Tue, 22 Oct 2019 11:47:21 -0400, bitrex <user@example.net> wrote:

On 10/22/19 3:58 AM, tabbypurr@gmail.com wrote:
On Tuesday, 22 October 2019 07:32:25 UTC+1, Jan Panteltje wrote:
21 Oct 2019 tabbypurr wrote in
2147376f-2e89-441c-b6a4-143353e7ae67@googlegroups.com>:
On Monday, 21 October 2019 16:43:16 UTC+1, bitrex wrote:
On 10/21/19 5:17 AM, tabbypurr wrote:

lucky you avoided the LM386. Sound of those not good.

They're nice for a lot of tasks like solenoid and servos, driving small
brushed DC motors and other applications other than audio. The DIP-8
version LM386N is very rugged and can dissipate a fair amount of power
(a small copper heat sink thermal epoxied to the top helps.) It can
drive a 30mm brushed DC motor like in an RC car.

I think it's the world's most popular audio power amp IC. (Maybe the one with the most distortion too.)

Did some more tests with the Chinese thing last night.
I think the damping on my headphones is not so good, that results in some
suspicious 's' sounds in some speech.
It is a large headphone that behaves a lot like a dynamic speaker...

So, maybe I will add a LM380 or LM386 to compare,
and see if I can get any of the chips bitrex mentioned.
Space enough in that plastic housing.

The 386 has a very characteristic sound. I don't know what type of distortion causes it - I wonder if it might be slew rate limitation, but don't know.

It's just distortion I think. It's a very simple 3 stage amp inside the
diff pair doesn't even have a current source in the tail and its open
loop gain is relatively limited. THD is already at 10% with a 9V supply
pushing 500mW.

What is the noise characteristics of the LM386, it doesn't appear in
those data sheets I have seen.

With headphone sensitivities in the order of 100 dB/mW so to reach 120
dBspl at least 100 mW of power is needed. Thus 0.9 Vrms is needed for
8 ohm, 3.1 Vrms for 100 ohm or 7.7 V for 600 ohm load impedance.
Multiply by 3 (actually 2sqrt(2)) and add two or four Vbe drops and
you get the minimum supply voltage needed. Of course, this is the
clipping limit, so even a higher supply voltage is needed to cleanly
produce 100 mW. In practice, the LM386 could produce 100 mW into 8 ohm
without too much THD at 1 kHz using a 6 Vdc supply, but not much
higher impedances.

The question about noise is interesting. In power amplifiers signal to
noise ratio is specified at full nominal rated power. Even in good
amplifiers this is often only 100 dB SNR. If the usable low
distortion output power is well below rated power, this directly
reduces the available SNR by many dB.

At least with closed headphones more than 100.to 110 dB SNR should be
available. Thus the output noise density for 8 ohm load should be
less than 63 nV/sqrt(Hz). Divide this by voltage gain to get input
related noise density.

Is the LM386 up to this ?

 

On Tue, 22 Oct 2019 15:00:19 -0400, bitrex <user@example.net> wrote:

On 10/22/19 2:02 PM, Jan Panteltje wrote:
On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low distortion.
I have used those before in a (not HiFi) monitoring system in a radio transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover compensation?
Output current is a bit low though.


Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

Sounds like the Quad 405 feed-forward system.

Of cause these days, the dumb high signal patch could be implemented
with a class-D amplifier, while the low power path could be a class-A
linear stage.

 

[Jan Panteltje]

On a sunny day (Tue, 22 Oct 2019 15:25:04 -0700 (PDT)) it happened
jrwalliker@gmail.com wrote in
<45e35081-d922-40ed-aa19-6a120fc78973@googlegroups.com>:

On Tuesday, 22 October 2019 10:02:01 UTC+1, Cursitor Doom wrote:

Some top-notch manufacturers used to make electrostatic headphones, too.
I haven't seen 'em for a long time, though. The sound quality was
*amazing* but having 600V right up close to your head wasn't a great
selling point for the average buyer, I'd imagine.

I once built a pair of electrostatic headphones to a design published
in Wireless World. It involved rubbing Aqua Dag (graphite colloid)
onto cling film to make the high resistivity diaphragm. They were
driven with a hybrid transistor-valve amplifier. The sound was
superb, but they were a bit impractical.

John

Somebody once showed me one, I listened to it for a short while, it sounded nice,
transparent.. no idea who made it, but i as not excessively big,
had some small box that drove it.

Like 'electrets' could you not use a self biasing membrane?

 

[Jan Panteltje]

On a sunny day (Tue, 22 Oct 2019 15:00:19 -0400) it happened bitrex
<user@example.net> wrote in <8lIrF.173159$Jx2.30119@fx36.iad>:

On 10/22/19 2:02 PM, Jan Panteltje wrote:
On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low distortion.
I have used those before in a (not HiFi) monitoring system in a radio transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover compensation?
Output current is a bit low though.


Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

I do not see that argument, the lower one will have to be able to supply the same power to the load I'd think?

 

[Piotr Wyderski]

Jan Panteltje wrote:

But all the time I was wondering about the schematic:
http://panteltje.com/pub/chinese_headphone_amp_schematic_IXIMG_0191.JPG

Please remove it ASAP and pray your reputation hasn't been damaged.
Man, this one is READABLE.

Best regards, Piotr

;-)))

 

On Wed, 23 Oct 2019 06:45:08 GMT, Jan Panteltje
<pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Tue, 22 Oct 2019 15:00:19 -0400) it happened bitrex
user@example.net> wrote in <8lIrF.173159$Jx2.30119@fx36.iad>:

On 10/22/19 2:02 PM, Jan Panteltje wrote:
On a sunny day (Tue, 22 Oct 2019 11:56:11 -0400) it happened bitrex
user@example.net> wrote in <wEFrF.70497$hB2.52985@fx42.iad>:

It doesn't have to be an 8 dollar chip but for a headphone amp with two
op amps in parallel if you're going to use op amps as drivers the
parallel combo I think should be able to at least push 100mA peak,
symmetrically, across the full audio band, into your headphone impedance
of choice. NE5532 not really up to the task for 64 or 32 ohm headphones.

Yes OK.
I found 5 LM380 for 3$50 free shipping on ebay, ordered.
That chip has some real power, maybe not HiFi, but reasonable low distortion.
I have used those before in a (not HiFi) monitoring system in a radio transmitter 19 inch rack many years ago.
That LT1364 and LT1365 seems interesting, 70 MHz gain-bandwidth is a lot... not sure if I can ..

Sounds like a recipe for oscillation to me in that circuit as-is, with a
GBW that high.

Output is a complementary emitter follower x 2,
feed-forward via Rc Cc (LT datasheet page 10), interesting, crossover compensation?
Output current is a bit low though.


Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

I do not see that argument, the lower one will have to be able to supply the same power to the load I'd think?

While of course the lower amplifier needs to handle the full load
current, the needed voltage swing is much smaller. Thus, a lower power
amplifier.

 

[Jan Panteltje]

On a sunny day (Wed, 23 Oct 2019 11:21:09 +0300) it happened
upsidedown@downunder.com wrote in
<e830red4vha8glqcftf2m82o4qkc9g980k@4ax.com>:

On Wed, 23 Oct 2019 06:45:08 GMT, Jan Panteltje
pNaOnStPeAlMtje@yahoo.com> wrote:

On a sunny day (Tue, 22 Oct 2019 15:00:19 -0400) it happened bitrex
user@example.net> wrote in <8lIrF.173159$Jx2.30119@fx36.iad>:

Another interesting amp topology is this:

https://www.tubecad.com/2004/12-2-2004_Sandman2.gif

The upper amp is a slow, beefy class B or lean AB amp with meh THD
specs. The lower is a fast lower-power amp with much better THD. The
second is used to clean up the first and keep it in line.

I do not see that argument, the lower one will have to be able to supply the same power to the load I'd think?

While of course the lower amplifier needs to handle the full load
current, the needed voltage swing is much smaller. Thus, a lower power
amplifier.

Yes of course same current, you are right.

 

[Jeroen Belleman]

Piotr Wyderski wrote:

Jan Panteltje wrote:

But all the time I was wondering about the schematic:

http://panteltje.com/pub/chinese_headphone_amp_schematic_IXIMG_0191.JPG

Please remove it ASAP and pray your reputation hasn't been damaged.
Man, this one is READABLE.

Best regards, Piotr

;-)))

Readable, yes, but it also has the + and - inputs reversed on
the output amps.

Jeroen Belleman

 

[Jan Panteltje]

On a sunny day (Wed, 23 Oct 2019 09:04:26 +0300) it happened
upsidedown@downunder.com wrote in
<inqvqeps4op523hb6b7mip7g09jq3eh8ku@4ax.com>:

What is the noise characteristics of the LM386, it doesn't appear in
those data sheets I have seen.

With headphone sensitivities in the order of 100 dB/mW so to reach 120
dBspl at least 100 mW of power is needed. Thus 0.9 Vrms is needed for
8 ohm, 3.1 Vrms for 100 ohm or 7.7 V for 600 ohm load impedance.
Multiply by 3 (actually 2sqrt(2)) and add two or four Vbe drops and
you get the minimum supply voltage needed. Of course, this is the
clipping limit, so even a higher supply voltage is needed to cleanly
produce 100 mW. In practice, the LM386 could produce 100 mW into 8 ohm
without too much THD at 1 kHz using a 6 Vdc supply, but not much
higher impedances.

The question about noise is interesting. In power amplifiers signal to
noise ratio is specified at full nominal rated power. Even in good
amplifiers this is often only 100 dB SNR. If the usable low
distortion output power is well below rated power, this directly
reduces the available SNR by many dB.

At least with closed headphones more than 100.to 110 dB SNR should be
available. Thus the output noise density for 8 ohm load should be
less than 63 nV/sqrt(Hz). Divide this by voltage gain to get input
related noise density.

Is the LM386 up to this ?

All good and well, indeed my Sennheiser HD201 does attenuate much of the outside noise.
But... 2 PCc with fans, airco fan in room, OK turn it all off for the ultimate experience
clock ticking in kitchen OK pull battery out, far away farmer working at night with tracker in field,
OK wait till he is gone (>2 at night), now early traffic on main road, mm I have double glazing it helps,
dogs, maybe for me a more realistic achievable value is 50 dB or do, conversations,..

I know a real audiophile has his/her own dead room, but I have no progressed that far yet,
The LM380 is a bit better but very much like the LM386, but can be run from a 3 cell lipo if must be.
And then there is the sort of music you listen to, I m sure there is classical music that need that high dynamic range
but lot of stuff I have comes from old vinyl .... live audience noise, etc etc..
It is gonna be allright!!!