searching better alternative for LM386

[ian field]

Tinnitus can come from drug side effects. It's thought that mine is a
result of continual aspirin therapy for arthritis. But many other
regular-use drug therapies have been implicated as well.

It can also come from too much exposure to leftist weenies ;-)

.................and also lunatic fringe right wing extremists!

 

[ian field]

At least quads of KT66s.

Nah! - you mean those transmitter tubes (TT something or other) with 2
prongs on top and run off several kV HT.

 

[Homer J Simpson]

I'm willing to bet you that many current commercial offerings don't have
the level of interlocks you suggest, given that determining "power
levels... known to cause hearing damage" is a complex function of wave
shape, duration, the listener, etc.

Yup. See the interlock system on the cargo doors of Boeing 747s. Cost 11
lives in one incident alone. Good luck getting a better system on a piece of
electronics.






--

 

[Phil Allison]

"Joel Kolstad"

"Phil Allison"


I'm willing to bet you

** Its a dead cert YOU are fucking half wit.

Not that I know,

** Its a dead cert YOU are fucking half wit.

Of course, some amount of interlocking should exist.

** As a separate, fail safe circuit, monitoring the signal level at the
earpiece.

Be criminally reckless not to have it.




........ Phil

 

[John Popelish]

(snip)

I NEED to generate something in the order of 150 dB SPL
briefly during these tonebursts (at the maximum power setting,
obviously in most cases (healthy people etc.) the required power will
be much lower).

(snip)

This need for a very high sound pressure level, combined
with a very quiet noise level between pulses implies a need
for an inherently low noise amplifier, and the LM386 just
isn't going to do that. What sound pressure level do you
require for the quiet times between pulses? You may have to
switch the power off to the amplifier to get 120+ db (1
million to 1) ratio between the level of the pulses and the
silence between them.

 

[Joel Kolstad]

Thanks Phil, this was exactly the kind of feedback I was hoping for!

 

[Joel Kolstad]

Yup. See the interlock system on the cargo doors of Boeing 747s. Cost 11
lives in one incident alone. Good luck getting a better system on a piece of
electronics.

This one: http://en.wikipedia.org/wiki/United_Airlines_Flight_811 ? Or
something else?

 

[ian field]

(snip)

This need for a very high sound pressure level, combined with a very quiet
noise level between pulses implies a need for an inherently low noise
amplifier, and the LM386 just isn't going to do that. What sound pressure
level do you require for the quiet times between pulses? You may have to
switch the power off to the amplifier to get 120+ db (1 million to 1)
ratio between the level of the pulses and the silence between them.

Many of the "boomer" series amplifiers from NS, already mentioned - such as
the LM46xx, 47xx & 48xx series have silent "pop-less" shutdown controls,
this might work for the silent operation between signals that the OP needs.

 

[Homer J Simpson]

This one: http://en.wikipedia.org/wiki/United_Airlines_Flight_811 ? Or
something else?

Yep. They tried to pretend it was something else - until the 1991 incident
involving malfunction of a UAL Boeing 747 cargo door occurred at New York's
JFK airport.

If that's what we fly in - how safe is a small gadget going to be?





--

 

[Adrian Jansen]

John,


Yes, I have provided the 12V from a battery (with short, twisted leads)
instead of from the DC/DC converter and the noise remains the same.

(if the 12V from the DC/DC is insufficiently filtered there is
additional noise in addition to this ever-present noise, but is caused
by the switching in the DC/DC converter, and is not white at all)

So I'm pretty sure it is the LM386 generating the noise. I can also
measure with the scope that the supplies are clean and the output has
significant noise, even if loaded with only 100 ohms instead of 10
ohms.

greetings,
Tom

What level is the output noise with a 10 ohm load ? Some figures would
be helpful.

Have you got the correct snubber network on the output - 47 nF and 10
ohm in series to ground ? Also correct power supply bypass ? These
things are critical to prevent the chip going into oscillation.

Have you also tried using the higher gain configuration - pins 1 and 8
open ?

Have you bypassed the unused input ?


--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.

 

[Tom]

Adrian,

What level is the output noise with a 10 ohm load ? Some figures would
be helpful.

I don't have the exact figures handy here (posting from home), but IIRC
it's in the order of 5-10 mV peak-peak over the audio bandwidth at the
output side.

Have you got the correct snubber network on the output - 47 nF and 10
ohm in series to ground ? Also correct power supply bypass ?
Yes.

These things are critical to prevent the chip going into oscillation.

It looks like noise (random), not oscillation. There's no obvious
ringing on step signals either, so it looks quite stable.

Have you also tried using the higher gain configuration - pins 1 and 8
open ?

Yes, that's what I started with. I then reduced the gain from 20 to 9
to reduce the output noise accordingly.

Have you bypassed the unused input ?

Since the source impedance is less than 10 kOhm I have shorted the
unused input to ground, per the datasheet's recommendation. Ground is a
solid plane, by the way.

greetings,
Tom

 

[Phil Allison]

"Dangerous Tom"

I don't have the exact figures handy here (posting from home), but IIRC
it's in the order of 5-10 mV peak-peak over the audio bandwidth at the
output side.

** No way a ( correctly functioning) LM 386 generates THAT much residual
noise.

Tom is totally clueless with analogue audio.




....... Phil

 

[Joerg]

Tinnitus can come from drug side effects. It's thought that mine is a
result of continual aspirin therapy for arthritis. But many other
regular-use drug therapies have been implicated as well.

Mine happened right after a session on the artillery range. Usually the
sound was weeeeeeoooou ... BOOM. One of them wasn't like that. The pitch
didn't change on that one and an otherwise never cussing sergeant next
to me screamed "Oh S..t!". Then we both hit the dirt.

 

[Jim Thompson]

Mine happened right after a session on the artillery range. Usually the
sound was weeeeeeoooou ... BOOM. One of them wasn't like that. The pitch
didn't change on that one and an otherwise never cussing sergeant next
to me screamed "Oh S..t!". Then we both hit the dirt.

My tinnitus sometimes stops for short periods of time... VERY
startling when it's suddenly quiet ;-)

...Jim Thompson

 

[Joerg]

My tinnitus sometimes stops for short periods of time... VERY
startling when it's suddenly quiet ;-)

Yep, same here. I think most of them do that.

 

[Ken Smith]

Martin,


Thank you, that looks like an interesting part. Its ~8 ohm output
resistance is on the high side for a 10 ohm load, but if it stays
sufficiently constant I can compensate for that.

Beware of the LT1010. It really hates to drive inductive loads. It seems
to like capacitive loads OK. I would suggest the LT1206 as a better part.

 

[Ken Smith]

Jan,


Cute, thanks, but doesn't quite meet the compactness & low power
requirements...

What you really want is a pair of power-tech PT-9503.

But seriously, I just had a thought of a last ditch protection system.
Make the DC supply the thing runs on unable to support the load for very
long. This will limit the damage if all the other failsafes fail.

 

[Phil Allison]

"Ken Smith"

But seriously, I just had a thought of a last ditch protection system.
Make the DC supply the thing runs on unable to support the load for very
long. This will limit the damage if all the other failsafes fail.

** Yes, maybe a current source in the 12 volt DC rail suppling an electro
cap sized to just hold up voltage under the required ( low) duty cycle.



........ Phil

 

[Michael A. Terrell]

The PhilTard is an engineer wanna be troll.

Its sad, really. Australia used to produce the highest quality trolls
in the world. There is no way that Phil should have slipped through
quality control.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Michael A. Terrell]

My tinnitus sometimes stops for short periods of time... VERY
startling when it's suddenly quiet ;-)

Thats when you pick up the newspaper to see if you'r still alive.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida