Maker Pro
Maker Pro

High voltage capacitors in audio

W

Wimpie

Jan 1, 1970
0
Thanks a lot for explaining. I get what you mean. I will take a closer
look on the functions of my capacitors.

But polyester, you say. Those are the plastic block ones, right?

You are right. The cheapest ones are the metal "colored" block
capacitors (the leads will come off easy). Just look to MKT or MKS,
they come in a variety of shapes and enclosures (both radial and axial
versions).

Best regards,

Wim
PA3DJS
www.tetech.nl
 
M

MooseFET

Jan 1, 1970
0
I have a bigger problem with the choice of output
transistors.

I hadn't looked at that issue. You are right. Obviously the right
transistors in this circuit are a TIP35C and TIP36C.
Can anyone see why 300 volt transistors with
gain that falls pretty sharply above 200 mA collector
current (and at 10 volts collector to emitter drop for the
PNP) were chosen for this design?

That would be only because he din't have a TIP35 and TIP36 in his junk
box.

With the output running
80 mA class A there is not much current capability left to
drive the headphones.http://komponenten.es.aau.dk/filead.../dfs/GTE/education/industrial/tec_analogiques...

I would like to see the amp tested, side by side with one
using a TIP29C, TIP30C output pair.http://www.fairchildsemi.com/ds/TI/TIP29C.pdfhttp://www.bourns.com/pdfs/tip30.pdf

Those are way to whimpy. You need great big transistors to make it
sound good.

Of course, if you change from class A to class AB (say, 5 to
10 mA idle current), you could probably use a pair of TO-92
output transistors, like ZTX692B and ZTX792A and cut the
power supply size in

Oh, no, it has to be class-A
 
M

MooseFET

Jan 1, 1970
0
I think it and the 15 pF are an overcompensation scheme for
the NE5534.


15pF would be about right if the NE5534 is being used at a gain of 2
or so. If the AOT value is 3.3K, it at the 470pF make a pole at about
200KHz. With a 15pF compensation capacitor, the gain cross over point
is well above 1MHz. Like I said, I think it is hooked up wrong.
 
M

MooseFET

Jan 1, 1970
0
Really ?

It looks like an active filter configuration to me.

Go look again.
sound.westhost.com/project24.htm

The op-amp is wired as a non-inverting amplifier. The 470pF puts a
pole within the feedback loop.

Yes it does.

What exactly does it do?
 
E

Eeyore

Jan 1, 1970
0
MooseFET said:
What exactly does it do?

It removes a DC path between the output of the headphone amp itself and the
input buffer. Without it you could get noise from operating the volume control
even with no audio present.

Graham
 
E

Eeyore

Jan 1, 1970
0
MooseFET said:
Go look again.
sound.westhost.com/project24.htm

The op-amp is wired as a non-inverting amplifier. The 470pF puts a
pole within the feedback loop.

I see what you're saying now. I was thinking the AOT might be zero.

Intruiging. Might be interesting to simulate.

Graham
 
M

me

Jan 1, 1970
0
Subsonic energy being produced by your headphones or
speakers is a waste of amplifier power that can cause them
to distort the stuff you can hear. It is usually a good
thing to get rid of.

Sure, if it was there to begin with.
 
J

John Popelish

Jan 1, 1970
0
MooseFET said:
I hadn't looked at that issue. You are right. Obviously the right
transistors in this circuit are a TIP35C and TIP36C.


That would be only because he din't have a TIP35 and TIP36 in his junk
box.



Those are way to whimpy. You need great big transistors to make it
sound good.

I didn't see a smiley in there, anywhere, but I assume you
are being sarcastic. Lets stick to the actual situation for
a moment. What impedance headphones do you think this unit
was intended for. It may be there in the page, but a quick
look didn't find it. If I assume 32 ohm headphones, a +- 12
volt sine wave would produce about 2.25 watts, electrical, a
huge amount of power for headphones, I think. But the peak
current would still be the bias of 80 mA plus the peak load
current of 375 mA or 455 mA. That is getting pretty far
over the peak of the current gain curve for the 340 350
pair. But quite comfortable for the TIP29,31 pair.
Oh, no, it has to be class-A

For religious reasons?
 
J

John Popelish

Jan 1, 1970
0
MooseFET said:
15pF would be about right if the NE5534 is being used at a gain of 2
or so. If the AOT value is 3.3K, it at the 470pF make a pole at about
200KHz. With a 15pF compensation capacitor, the gain cross over point
is well above 1MHz. Like I said, I think it is hooked up wrong.

I wouldn't be stunned if you are right.
 
J

John Popelish

Jan 1, 1970
0
MooseFET said:
What exactly does it do?

There are two feedback loops connected to the output. One
of them is isolated from the DC applied to the top of the
pot with a downstream capacitor on the wiper, but the other
one is not isolated by that capacitor, so a DC error voltage
on the output of the first opamp would propagate to the output.
 
E

ectoplasm

Jan 1, 1970
0
I didn't see a smiley in there, anywhere, but I assume you
are being sarcastic. Lets stick to the actual situation for
a moment. What impedance headphones do you think this unit
was intended for. It may be there in the page, but a quick
look didn't find it. If I assume 32 ohm headphones, a +- 12
volt sine wave would produce about 2.25 watts, electrical, a
huge amount of power for headphones, I think. But the peak
current would still be the bias of 80 mA plus the peak load
current of 375 mA or 455 mA. That is getting pretty far
over the peak of the current gain curve for the 340 350
pair. But quite comfortable for the TIP29,31 pair.

My headphones are 32 ohm ones, but you might as well plug in 70 ohm
ones, ok... and if 300 ohm, it should also be ok (but they might
require higher voltage).

I think output power in one pair of headphones (32 ohm) would be
around 500 mW max? I guess output voltage of around 3 volt max is
reasonable. The circuit would have around 10x gain (adjustable by the
AOT resistor).

The output voltage is the same no matter what you plug in, isn't it?
What if you'd plug in 300 ohm headphones, wouldn't they sound too
silent?

As MooseFET says TIP35, 36 would be a good choice... I must take a
closer look on those data sheets later and make comparisons with the
original ones.
 
M

MooseFET

Jan 1, 1970
0
It removes a DC path between the output of the headphone amp itself and the
input buffer. Without it you could get noise from operating the volume control
even with no audio present.

What is the offset voltage in the two op-amps? Without the 22uF this
will be what is on the 100K pot. What fraction of that ends up as
noise?
 
M

MooseFET

Jan 1, 1970
0
There are two feedback loops connected to the output. One
of them is isolated from the DC applied to the top of the
pot with a downstream capacitor on the wiper, but the other
one is not isolated by that capacitor, so a DC error voltage
on the output of the first opamp would propagate to the output.


There is a 10uF after the pot in the path to the following op-amp. No
DC signal can get from the pot through the amplifiers.
 
M

MooseFET

Jan 1, 1970
0
I didn't see a smiley in there, anywhere, but I assume you
are being sarcastic.

Yes the suggestion of the TIP35 was intended to be funny. Even if you
assumed 8 ohm head phones, you would destroy someones hearing long
before you needed a TIP35.
Lets stick to the actual situation for
a moment. What impedance headphones do you think this unit
was intended for. It may be there in the page, but a quick
look didn't find it. If I assume 32 ohm headphones, a +- 12
volt sine wave would produce about 2.25 watts, electrical, a
huge amount of power for headphones, I think.

Yes, that is way too much power. At that volume the small bones in
the ear clattering together make more distortion than the amplifiers.

But the peak
current would still be the bias of 80 mA plus the peak load
current of 375 mA or 455 mA. That is getting pretty far
over the peak of the current gain curve for the 340 350
pair. But quite comfortable for the TIP29,31 pair.

In truth, I fully agreed with your selection.
For religious reasons?

Its marketing. A "class A shoe" is better than a "class B shoe". A
"class A dinner set" is also better so a "class A amplifier" must be
the best.

I once saw a 25W class-A audio amplifier. It also would serve as a
space heater.
 
M

MooseFET

Jan 1, 1970
0
My headphones are 32 ohm ones, but you might as well plug in 70 ohm
ones, ok... and if 300 ohm, it should also be ok (but they might
require higher voltage).

The P = V^2/R

If you double the headphone resistance, the voltage goes up by about
40%.

I think output power in one pair of headphones (32 ohm) would be
around 500 mW max? I guess output voltage of around 3 volt max is
reasonable. The circuit would have around 10x gain (adjustable by the
AOT resistor).

The OPA2604 stage has a gain of just about 10 as it is. The 1K and
10K in the second OPA2604 section sets the gain of that part.

If I was designing this, I would have made the two stages of the
OPA2604 share the work of providing gain differently. The second
section is also canceling the distortion of the output pair.

The output voltage is the same no matter what you plug in, isn't it?
What if you'd plug in 300 ohm headphones, wouldn't they sound too
silent?

As MooseFET says TIP35, 36 would be a good choice... I must take a
closer look on those data sheets later and make comparisons with the
original ones.


No, no, no! That was a joke. They are more than 10 times the
transistor you want there.

3W is more than more than enough power to blow your ears up.
 
E

ectoplasm

Jan 1, 1970
0
I see only one capacitor in that circuit that will have any
significant effect on the sound. It is the 1 nF. The rest
have essentially no changing voltage across them in the
audio frequency range, so will have almost no effect on audio.

I see. So use a film capacitor there instead of electrolytic.
Polyester would be fine then I guess, as Wimpie suggested (this one
might be too small for that perhaps).

On the other hand, MooseFET says the 10uF at the input is the only one
worth changing. But as others mentioned, electrolytic does fine for DC
coupling (zero bias coupling) & no filter purposes, which is what this
10uF cap does... MooseFET?
That said, I think some of the capacitor values are
a bit small, like the two 22 uF caps in the output stage.

What value would you suggest then?
 
E

ectoplasm

Jan 1, 1970
0
I see what you're saying now. I was thinking the AOT might be zero.

Intruiging. Might be interesting to simulate.

So the 470pF / 15pF together would provide a wrong filter when the AOT
resistor is not zero? Like when it would be 3.3k (as MooseFET
mentioned the cross over point being well above 1 MHz in that case.

So that means if the AOT is set to another value, these two capacitors
will require different values?
 
Top