Better than LM386 on batteries?

[XTL]

On Dec 24 2007, 7:44 am, Eeyore
<rabbitsfriendsandrelati...@hotmail.com> wrote:

phaeton wrote:

If you are going to build a CMOY don't don't don't use a 9v battery - the
life through a pair of 64ohm 'phones is, perhaps, 3 hours! Use some good
NiMH AAAs or consider looking at lithium polymer batteries.

The Burr-Brown chips are significantly overspec'd for my application,
at least as far as the THD figures are. Especially since I'm kinda
sorta planning a distortion circuit for the instrument input anyways.
Ideally, it would have that and some sort of mid-hump tone control,
but the more I think about it, the more features I want to add, and I
tend to overcomplicate everything I do. I've been trying to practice
the K.I.S.S. philosophy. Just go back to the original plan of 3
simple audio channels each with their own separate volume and pan
control. In any event, I've got 3 RC5532s in my parts collection.
I'll still have to use TL082s or some other opamp for the rest.

TL082s will be good for the high input impedance input stage you should have to
avoid destroying the 'tone' of your guitar. Mind you it's less of an issue with
bass guitars.

Actually, the LM386 is somewhat popular for guitar boxes because of
its ..
non-hifi tone. See http://runoffgroove.com/littlegem.html for example.
The site
also has other interesting stuff for guitar use.

My headphone box uses two Philips TDA7050 chips and individual volume
controls
for two pairs of headphones, pretty much like the datasheet says, and
fares
quite well with two AA NiMH batteries or alkalines. Haven't used it
for guitar.
I actually have a terrible sounding Korg Pandora for that

 

[phaeton]

Actually, the LM386 is somewhat popular for guitar boxes because of its crap tone.

I've actually got an assortment of LM386 type amplifiers that I use
frequently. One is similar to a Ruby that I play into a pair of 6x9
speakers. The only thing with the LM386 is that it eats batteries and
distorts VERY easily. Great sound for guitar, but I'm trying to get
something with very low distortion for mp3 playback and bass guitar
amplification.

Otherwise, yeah. I love the 386 for guitar. It makes a great
stompbox too. Check out Aron Nelson's SmashDrive.

-phaeton

 

[phaeton]

On Dec 18 2007, 2:45 am, "LAB" <n...@mail.com> wrote:

TDA7053 for a battery powered headphone amp;

NJM4580 or NJM4560 give more power if you have +/- 12 or 15V. You can
also use the two sections of a TL072 for each channel (FidoCAD,http://www.enetsystems.com/~lorenzo/fidocad_win.asp) :

Thanks again for FidoCad. Here's what I did (FidoCad source at bottom
of post):

http://webpages.charter.net/physarum/Jurd/avatar/headamptest.bmp


It works! Unfortunately, it's very faint lest I hit it with a pretty
good signal (such as with a preamp in front of this). Then it is
faint but distorted. I tried increasing the feedback resistance of
the upper opamp but that just adds more distortion. I'm running it
from an 18V supply (2 9V batteries stacked). "at rest" my headphones
measure at 19-21 Ohms in resistance. Is that kind of high? Could
that be sapping the power? I don't have another pair to try instead,
unfortunately.

Thanks for any suggestions or help

-phaeton


[FIDOCAD ]
LI 65 185 65 170
LI 65 170 90 170
MC 95 190 0 0 080
LI 90 190 95 190
LI 65 195 65 205
LI 65 205 90 205
LI 90 205 90 190
SA 90 190
LI 110 190 105 190
MC 95 150 0 0 080
LI 90 170 90 150
SA 90 150
LI 90 150 95 150
LI 110 155 110 190
LI 105 150 110 150
LI 110 155 110 150
MC 125 195 0 0 045
LI 125 190 125 195
SA 125 190
SA 125 180
LI 125 175 125 180
TY 130 180 5 3 0 0 0 * Headphone
SA 110 170
TY 70 160 5 3 0 0 0 * NE5532
MC 65 185 0 0 580
TY 75 185 5 3 0 0 0 * b
TY 90 140 4 2 0 0 0 * 100 Ohm
TY 90 180 4 2 0 0 0 * 100 Ohm
LI 90 130 90 150
LI 76 148 77 148
MC 65 105 0 0 580
SA 90 110
SA 65 105
SA 65 115
LI 90 110 90 150
LI 65 115 65 125
LI 65 125 75 125
MC 75 125 0 0 080
TY 65 130 4 3 0 0 0 * 1M Ohm
SA 90 125
LI 85 125 90 125
LI 65 115 55 115
MC 45 115 0 0 080
LI 45 115 35 115
LI 65 105 60 105
LI 60 105 60 215
MC 60 215 0 0 000
TY 70 215 4 3 0 0 0 * V / 2
TY 35 105 4 3 0 0 0 * 10K Ohm
RV 10 95 160 225
LI 125 170 125 175
MC 25 115 0 0 170
MC 110 170 0 0 170
TY 115 160 5 3 0 0 0 * 0.22uf
LI 120 170 125 170
LI 25 115 15 115
SA 15 115
TY 25 125 5 3 0 0 0 * 0.22uf

 

[John Popelish]

phaeton wrote:

On Dec 18 2007, 2:45 am, "LAB" <n...@mail.com> wrote:
TDA7053 for a battery powered headphone amp;

NJM4580 or NJM4560 give more power if you have +/- 12 or 15V. You can
also use the two sections of a TL072 for each channel (FidoCAD,http://www.enetsystems.com/~lorenzo/fidocad_win.asp) :



Thanks again for FidoCad. Here's what I did (FidoCad source at bottom
of post):

http://webpages.charter.net/physarum/Jurd/avatar/headamptest.bmp


It works! Unfortunately, it's very faint lest I hit it with a pretty
good signal (such as with a preamp in front of this). Then it is
faint but distorted. I tried increasing the feedback resistance of
the upper opamp but that just adds more distortion. I'm running it
from an 18V supply (2 9V batteries stacked). "at rest" my headphones
measure at 19-21 Ohms in resistance. Is that kind of high? Could
that be sapping the power? I don't have another pair to try instead,
unfortunately.

The output capacitor is way to small. For a roll off corner
frequency around 20 Hz, it should be about 100 uF. Positive
end toward the opamps. A 47 uF might be large enough (rolls
off below 40 Hz) and will prevent some inaudible low
frequency power being wasted in the opamps.

The 0.22 uF cap you have will roll off everything below
10,000 Hz, so all you get out is the highest treble. Very
thin sounding.

Note that the 0.22 uF at the input may roll off the signal
below 72 Hz. I would probable increase that to 0.47 to 1 uF.

--
Regards,

John Popelish

 

[LAB]

Hi phaeton,

change the capacitor as suggested by John Popelish

Gianluca

 

[LAB]

- change the capacitorS

 

[Eeyore]

phaeton wrote:

On Dec 18 2007, 2:45 am, "LAB" <n...@mail.com> wrote:
TDA7053 for a battery powered headphone amp;

NJM4580 or NJM4560 give more power if you have +/- 12 or 15V. You can
also use the two sections of a TL072 for each channel (FidoCAD,http://www.enetsystems.com/~lorenzo/fidocad_win.asp) :


Thanks again for FidoCad. Here's what I did (FidoCad source at bottom
of post):

http://webpages.charter.net/physarum/Jurd/avatar/headamptest.bmp

It works! Unfortunately, it's very faint lest I hit it with a pretty
good signal (such as with a preamp in front of this). Then it is
faint but distorted. I tried increasing the feedback resistance of
the upper opamp but that just adds more distortion. I'm running it
from an 18V supply (2 9V batteries stacked). "at rest" my headphones
measure at 19-21 Ohms in resistance. Is that kind of high? Could
that be sapping the power? I don't have another pair to try instead,
unfortunately.

Thanks for any suggestions or help

Well, the input impedance is only 10 k ohms so it'll heavily load a guitar pickup. And the 0.22uF coupling cap is a bit on
the small side too.

You should use a NON inverting arrangement for best noise and higher input impedance.

Also, you can't drive headphones through a 0.22uF cap ! It'll effectively be a high pass filter and is no doubt where
you're losing most of the volume. You need more like 220uF there - or DC couple the signal if it's split rail (the junction
of the 2 batteries is 'ground').

You need to learn some more basics I'm afraid.

Graham

 

[phaeton]

John Popelish wrote:

The output capacitor is way to small. For a roll off corner frequency
around 20 Hz, it should be about 100 uF. Positive end toward the
opamps. A 47 uF might be large enough (rolls off below 40 Hz) and will
prevent some inaudible low frequency power being wasted in the opamps.

The 0.22 uF cap you have will roll off everything below 10,000 Hz, so
all you get out is the highest treble. Very thin sounding.

Note that the 0.22 uF at the input may roll off the signal below 72 Hz.
I would probable increase that to 0.47 to 1 uF.

Thanks John. I did just that (change the caps) and it made a huge
difference. It's not quite the "ear splitting" volume level that I was
expecting, but it's definitely usable now. Is there a specific name for
the equation that you're using to calculate the rolloff freqs? I've
always used the freq=(1/(2*PI*R1*C1)) for R/C pairs but when there's no
resistor involved I'm not sure where to go. I initially chose .22uF
because most guitar stompbox circuits have .1uF input and output caps,
and since this is for a bass guitar (one more octave down) I figured
that .22uF should be appropriate.

I also noticed that the difference in sound quality and volume is very
minor between running it from 9V or 18V. In that case, I may just go 9V.

Eeyore wrote:


You need to learn some more basics I'm afraid.

Graham

Agreed. Learning by doing and reading, when I have time. Asking lots
of questions, but trying hard to not be a pest. If I ever am, feel free
to kindly send me away with "go read up on X" or "go read this link".
For some things though, I just don't know where to start, and I just
need someone to point me in the general direction of the pinata.

Fwiw, I'm going to also go with your previous suggestion of using JFET
type opamps as the 'preamp section' for the high impedance. I'll use
that area for tone shaping and signal splitting. These NE5532s will
drive the headphones.

Right now I am also using the pair of 9V batteries to make +9V and -9V
with their junction point as ground (like you mentioned).

Thanks everyone!

-phaeton

 

[John Popelish]

phaeton wrote:

Thanks John. I did just that (change the caps) and it made a huge
difference. It's not quite the "ear splitting" volume level that I was
expecting, but it's definitely usable now. Is there a specific name for
the equation that you're using to calculate the rolloff freqs? I've
always used the freq=(1/(2*PI*R1*C1)) for R/C pairs but when there's no
resistor involved I'm not sure where to go. I initially chose .22uF
because most guitar stompbox circuits have .1uF input and output caps,
and since this is for a bass guitar (one more octave down) I figured
that .22uF should be appropriate.
(snip)

That is the right formula, but the resistance for the output
is the parallel combination of the two opamp output
resistors in series (plus) the headphone resistance or
50+20. Set frequency to 20 Hz and solve for C1.


--
Regards,

John Popelish

 

[Eeyore]

phaeton wrote:

John Popelish wrote:

The output capacitor is way to small. For a roll off corner frequency
around 20 Hz, it should be about 100 uF. Positive end toward the
opamps. A 47 uF might be large enough (rolls off below 40 Hz) and will
prevent some inaudible low frequency power being wasted in the opamps.

The 0.22 uF cap you have will roll off everything below 10,000 Hz, so
all you get out is the highest treble. Very thin sounding.

Note that the 0.22 uF at the input may roll off the signal below 72 Hz.
I would probable increase that to 0.47 to 1 uF.

Thanks John. I did just that (change the caps) and it made a huge
difference. It's not quite the "ear splitting" volume level that I was
expecting, but it's definitely usable now. Is there a specific name for
the equation that you're using to calculate the rolloff freqs? I've
always used the freq=(1/(2*PI*R1*C1)) for R/C pairs but when there's no
resistor involved I'm not sure where to go.

The resistor was 10k.

I initially chose .22uF
because most guitar stompbox circuits have .1uF input and output caps,
and since this is for a bass guitar (one more octave down) I figured
that .22uF should be appropriate.

I also noticed that the difference in sound quality and volume is very
minor between running it from 9V or 18V. In that case, I may just go 9V.

On 9V you will lose maximum signal level handling (it'll clip earlier) but
that may not be an issue for you.

Eeyore wrote:


You need to learn some more basics I'm afraid.

Graham


Agreed. Learning by doing and reading, when I have time. Asking lots
of questions, but trying hard to not be a pest. If I ever am, feel free
to kindly send me away with "go read up on X" or "go read this link".
For some things though, I just don't know where to start, and I just
need someone to point me in the general direction of the pinata.

Fwiw, I'm going to also go with your previous suggestion of using JFET
type opamps as the 'preamp section' for the high impedance. I'll use
that area for tone shaping and signal splitting. These NE5532s will
drive the headphones.

Right now I am also using the pair of 9V batteries to make +9V and -9V
with their junction point as ground (like you mentioned).

Here's a good source of info on op-amps. Theory, practice the works.
http://focus.ti.com/lit/an/slod006b/slod006b.pdf

Graham