Noise in Amplifier.

[L.A.T.]

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:648nv7F2a2gqnU1@mid.individual.net...

"L.A.T."

** Can you ( or anyone here ) post the schematic of this unit ?
On ABSE, or elsewhere please.


Done.


** Still not visible to me, but FZ has posted a link to the schem now.

OK, all you gotta do is modify the headphone output circuit.

1. Replace the 22 ohm with a 2k2 ohms.

2. Fit the 22 ohm in parallel with one of the 1N5817 diodes.

Then crank up the VOL pot to compensate for the 40 dB loss of signal.

Should improve the signal to noise by lots.




....... Phil

Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain. Drawers
full of resistors but nothing closer. So off to Dick Smith (which is a town
or two north of us) to spend twenty or thirty cents on resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there is
something basically wrong with the way I made the amp in the first place.
But it is hard to imagine what I could do wrong with such a simple little
circuit.

 

[Phil Allison]

"L.A.T."

"Phil Allison"
Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain. Drawers
full of resistors but nothing closer. So off to Dick Smith (which is a
town or two north of us) to spend twenty or thirty cents on resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there is
something basically wrong with the way I made the amp in the first place.
But it is hard to imagine what I could do wrong with such a simple little
circuit.

** OK - if that has not helped, then the " hiss " is being generated
inside the electret microphone.

Such mics have a fet pre-amp, so it generates hiss.

BTW:

is the same level of hiss evident when monitoring on headphones??



........ Phil

 

[Franc Zabkar]

On Wed, 19 Mar 2008 03:39:43 GMT, "L.A.T." <tonyt@sci.net.au> put
finger to keyboard and composed:

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:648nv7F2a2gqnU1@mid.individual.net...

"L.A.T."

** Can you ( or anyone here ) post the schematic of this unit ?
On ABSE, or elsewhere please.


Done.


** Still not visible to me, but FZ has posted a link to the schem now.

OK, all you gotta do is modify the headphone output circuit.

1. Replace the 22 ohm with a 2k2 ohms.

2. Fit the 22 ohm in parallel with one of the 1N5817 diodes.

Then crank up the VOL pot to compensate for the 40 dB loss of signal.

Should improve the signal to noise by lots.




....... Phil


Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain. Drawers
full of resistors but nothing closer. So off to Dick Smith (which is a town
or two north of us) to spend twenty or thirty cents on resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there is
something basically wrong with the way I made the amp in the first place.
But it is hard to imagine what I could do wrong with such a simple little
circuit.

You stated in your OP that the hiss was present even with the
microphone disconnected. Try shorting the microphone input and then
check for hiss. Next short pins 2 and 3 of the LM386 amp and check for
hiss. That should help you locate the source of the noise.

Looking at the PCB layout, I'm wondering whether there is any
potential for a low input signal to be affected by any noise current
flowing along the outside trace between pins 3&4 of the amp and the
negative terminal of the battery. This noise current could give rise
to a small noise voltage which would be added to the mike voltage, and
this combination would then be amplified by the transistor stage which
has a gain of about 20. You could try adding an additional ground wire
between the MIC- and Headphone- points on the PCB.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[gcd]

Hi,
I'm a little confused, you've got a microphone (the electret) and without
the amp it's too low in volume? (even with the laptop mixer gain maximised?)

Then with the amp it's too high in volume and high in noise? and without the
electret the noise is still there.

Well the overall amp has again of some 300 odd and it's going into a mic
input. How about totally bypassing the 386, take the output from the pot
wiper.

A sound card max input for microphone is around 100mv. the electret should
output from 10uV up to 2-5mv. depending on its sensitivity and the sound
pressure.

I'd take the ouput from the wiper and perhaps even increase the 220ohm in
the emitter circuit up to 470.

If there is still noise then i'd start looking elsewhere including psu
coupled noise, or earth loops in cable grounds, installing twisted pairs,
reducing noise bandwidth etc.

Greg



"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
news:nu51u3d6fc78bgbbt2nqgjv9071kohjp2q@4ax.com...

On Wed, 19 Mar 2008 03:39:43 GMT, "L.A.T." <tonyt@sci.net.au> put
finger to keyboard and composed:


"Phil Allison" <philallison@tpg.com.au> wrote in message
news:648nv7F2a2gqnU1@mid.individual.net...

"L.A.T."

** Can you ( or anyone here ) post the schematic of this unit ?
On ABSE, or elsewhere please.


Done.


** Still not visible to me, but FZ has posted a link to the schem now.

OK, all you gotta do is modify the headphone output circuit.

1. Replace the 22 ohm with a 2k2 ohms.

2. Fit the 22 ohm in parallel with one of the 1N5817 diodes.

Then crank up the VOL pot to compensate for the 40 dB loss of signal.

Should improve the signal to noise by lots.




....... Phil


Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain. Drawers
full of resistors but nothing closer. So off to Dick Smith (which is a
town
or two north of us) to spend twenty or thirty cents on resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there is
something basically wrong with the way I made the amp in the first place.
But it is hard to imagine what I could do wrong with such a simple little
circuit.

You stated in your OP that the hiss was present even with the
microphone disconnected. Try shorting the microphone input and then
check for hiss. Next short pins 2 and 3 of the LM386 amp and check for
hiss. That should help you locate the source of the noise.

Looking at the PCB layout, I'm wondering whether there is any
potential for a low input signal to be affected by any noise current
flowing along the outside trace between pins 3&4 of the amp and the
negative terminal of the battery. This noise current could give rise
to a small noise voltage which would be added to the mike voltage, and
this combination would then be amplified by the transistor stage which
has a gain of about 20. You could try adding an additional ground wire
between the MIC- and Headphone- points on the PCB.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[L.A.T.]

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:64bh01F2aujp2U1@mid.individual.net...

"L.A.T."
"Phil Allison"
Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain.
Drawers full of resistors but nothing closer. So off to Dick Smith (which
is a town or two north of us) to spend twenty or thirty cents on
resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there
is something basically wrong with the way I made the amp in the first
place. But it is hard to imagine what I could do wrong with such a
simple little circuit.


** OK - if that has not helped, then the " hiss " is being generated
inside the electret microphone.

Such mics have a fet pre-amp, so it generates hiss.

BTW:

is the same level of hiss evident when monitoring on headphones??
I haven't tried. I will be off the air for a few days, and when I come back

I will try that first thing.

 

[L.A.T.]

"L.A.T." <tonyt@sci.net.au> wrote in message
news:RW4Ej.277$n8.77@news-server.bigpond.net.au...

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:64bh01F2aujp2U1@mid.individual.net...

"L.A.T."
"Phil Allison"
Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain.
Drawers full of resistors but nothing closer. So off to Dick Smith
(which is a town or two north of us) to spend twenty or thirty cents on
resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there
is something basically wrong with the way I made the amp in the first
place. But it is hard to imagine what I could do wrong with such a
simple little circuit.


** OK - if that has not helped, then the " hiss " is being generated
inside the electret microphone.

Such mics have a fet pre-amp, so it generates hiss.

BTW:

is the same level of hiss evident when monitoring on headphones??
I haven't tried. I will be off the air for a few days, and when I come
back I will try that first thing.

The plot thickens. I have just found a couple of files of recordings that we

made when we were first testing the thing and the noise was nowhere near
what it is now. It was, in fact, acceptable.

 

[L.A.T.]

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:64bh01F2aujp2U1@mid.individual.net...

"L.A.T."
"Phil Allison"
Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain.
Drawers full of resistors but nothing closer. So off to Dick Smith (which
is a town or two north of us) to spend twenty or thirty cents on
resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there
is something basically wrong with the way I made the amp in the first
place. But it is hard to imagine what I could do wrong with such a
simple little circuit.


** OK - if that has not helped, then the " hiss " is being generated
inside the electret microphone.

Such mics have a fet pre-amp, so it generates hiss.

There is no difference in the noise level whether the microphone is
connected or not. I think this is possibly good news.

BTW:

is the same level of hiss evident when monitoring on headphones??

I think so, but it is hard to be sure because the headphone
"environment" is different from listening to the laptop's
speakers.
I guess that is another way of saying that the difference is not
immediately perceptible.


....... Phil
By the way, more than one person has suggested by-passing the amp
altogether.
I originally mounted the LM386 in a socket, so could I achieve this
by removing the amp and jumpering pins 2 and 5?
Thanks

Tony

 

[L.A.T.]

"L.A.T." <tonyt@sci.net.au> wrote in message
news:Cl5Ej.289$n8.111@news-server.bigpond.net.au...

"L.A.T." <tonyt@sci.net.au> wrote in message
news:RW4Ej.277$n8.77@news-server.bigpond.net.au...

"Phil Allison" <philallison@tpg.com.au> wrote in message
news:64bh01F2aujp2U1@mid.individual.net...

"L.A.T."
"Phil Allison"
Curious. I didn't have a 2k2 so I put together a couple of resistors to
achieve about 2k6. Not enough gain. So I tried 1k5. Too much gain.
Drawers full of resistors but nothing closer. So off to Dick Smith
(which is a town or two north of us) to spend twenty or thirty cents on
resistors.
Trouble is, I don't think the hiss has decreased much, so I guess there
is something basically wrong with the way I made the amp in the first
place. But it is hard to imagine what I could do wrong with such a
simple little circuit.


** OK - if that has not helped, then the " hiss " is being generated
inside the electret microphone.

Such mics have a fet pre-amp, so it generates hiss.

BTW:

is the same level of hiss evident when monitoring on headphones??
I haven't tried. I will be off the air for a few days, and when I come
back I will try that first thing.

The plot thickens. I have just found a couple of files of recordings that
we made when we were first testing the thing and the noise was nowhere
near what it is now. It was, in fact, acceptable.

For the record:
I have found some more of the original recordings. Recorded with the
original circuit with the volume control advanced very slightly, through the
microphone input on the laptop, and they are good. Signal to noise is
acceptable, any noise from the amplifier masked by the noise of the breeze.
I think it is evident that something has changed in the original amplifier
circuit.
>

 

[L.A.T.]

snip

http://www.techbuy.com.au/p/48770/SOUNDCARD/Creative/ACSB0271.asp

Thank you for your recommendation.
Mr. Lyrebirdman http://lyrebirdman.googlepages.com/
has bought this Creative soundcard and it works well. I have reset the
little Oatley amp to its previous specs and the output at full volume is now
enough to drive headphones and/or to drive the Creative soundcard but only
just. The quality of the recording is much improved as you promised. With
Audacity we can amplify the track after we get it home.
I am thinking of mounting the dish and its microphone in a little
eggcrate-lined box, firstly to diminish unwanted sound but more crucially to
protect it from knocks and mishaps in the pre-dawn, pre-birdcall darkness.
Is there a practical limit to the length of the cable from the mic to the
amp? At the moment it is about 4cm.
Would I suffer any adverse effects from making it, say, 10cm?
Thank you all, once again.