audio amplifier output transistors

I noticed that TR7 is PNP and TR9 is NPN here:
http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif

Then over here, TR7 is NPN and TR9 is PNP:
http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif

Are these correct, or in error?

Also - for building audio amps "on the cheap", would TIP41C/TIP42C do
for output transistors in these schematics?

Thanks,

Michael

mrdarrett@gmail.com wrote:
> I noticed that TR7 is PNP and TR9 is NPN here:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif
> 
> Then over here, TR7 is NPN and TR9 is PNP:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif
> 
> Are these correct, or in error?

It looks to me like the first one is using Sziklai_pairs for the output
stage, see:
http://en.wikipedia.org/wiki/Sziklai_pair

and the other is using Darlington pairs, see:
http://en.wikipedia.org/wiki/Darlington_transistor

Either of these should be suitable for audio output stages, though I
didn't study these in detail.


-- 
-----------------------------------------------------------------------
To reply to me directly:

Replace privacy.net with: totalise DOT co DOT uk and replace me with
gareth.harris

mrdarrett@gmail.com wrote:
> I noticed that TR7 is PNP and TR9 is NPN here:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif
> 
> Then over here, TR7 is NPN and TR9 is PNP:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif
> 
> Are these correct, or in error?

They are both right.  Note that in both cases, the arrow in 
the emitter points down in all cases (from positive toward 
negative)

The first output is made of an inverter driving an inverter, 
with 100% voltage feedback to the first emitter, so the the 
pair act like a very precise follower, but with more 
stability issues, because of the inversion inside the loop.

The second version has output stages made of just a follower 
driving a follower, with no local voltage feedback, except 
that inherent in each follower.  This produces a less DC 
accurate follower, but with a higher frequency response (if 
the transistors are the same) and the distortion is removed 
by overall amplifier feedback.

> Also - for building audio amps "on the cheap", would TIP41C/TIP42C do
> for output transistors in these schematics?

After checking the peak voltage capability, I usually check 
the gain versus current curve and pick transistors that do 
not require currents much above that which forces the gain 
down to half of its peak value.  Then I check the safe 
operating area and overall wattage rating against the 
amplifier needs and heat sinks available.

The first amplifier does not specify power supply voltage 
rails and I don't know your load resistance.  But the second 
one (with +-37 volt rails) requires transistors with at 
least something like 80 volt capability.
From:
http://www.fairchildsemi.com/ds/TI%2FTIP41C.pdf
http://www.fairchildsemi.com/ds/TI%2FTIP42C.pdf
the 100 volt rating looks like a possible.

The gain falls from a typical peak of something like 55 
around 0.3 amps, but falls to about 27 at 5 amps, so that is 
all the peak current I would ask of these transistors.  That 
would limit the load resistance to about 7 ohms or more.

Checking the second breakdown (safe operating area) curves, 
you might stay within the safe area with an 8 ohm load.  But 
I think you are risking damage with a 4 ohm load.  Lowering 
the supply voltages a bit helps a lot (lowers the peak 
current and the peak voltage across the transistor with an 
inductive speaker load that keeps the load current going a 
bit as the voltage swings back through zero, so supply 
voltage across the transistor while current is flowing, 
briefly).

So these might serve for some loads, but will not withstand 
much abuse, like paralleled speakers or speaker wiring shorts.

The first circuit does not detail the Vbias part of the 
circuit between the two output half drivers, but that bias 
has to see the transistor temperature, so it can compensate 
for the changes in base emitter drop of the hot output 
transistors.  the second schematic shows one way to make 
such a bias generator, and TR13 should be thermally coupled 
to the output heat sink.
-- 
Regards,

John Popelish

On Oct 15, 3:13 pm, John Popelish <jpopel...@rica.net> wrote:
> mrdarr...@gmail.com wrote:
> > I noticed that TR7 is PNP and TR9 is NPN here:
> >http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif
>
> > Then over here, TR7 is NPN and TR9 is PNP:
> >http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif
>
> > Are these correct, or in error?
>
> They are both right.  Note that in both cases, the arrow in
> the emitter points down in all cases (from positive toward
> negative)
>
> The first output is made of an inverter driving an inverter,
> with 100% voltage feedback to the first emitter, so the the
> pair act like a very precise follower, but with more
> stability issues, because of the inversion inside the loop.
>
> The second version has output stages made of just a follower
> driving a follower, with no local voltage feedback, except
> that inherent in each follower.  This produces a less DC
> accurate follower, but with a higher frequency response (if
> the transistors are the same) and the distortion is removed
> by overall amplifier feedback.
>
> > Also - for building audio amps "on the cheap", would TIP41C/TIP42C do
> > for output transistors in these schematics?
>
> After checking the peak voltage capability, I usually check
> the gain versus current curve and pick transistors that do
> not require currents much above that which forces the gain
> down to half of its peak value.  Then I check the safe
> operating area and overall wattage rating against the
> amplifier needs and heat sinks available.
>
> The first amplifier does not specify power supply voltage
> rails and I don't know your load resistance.  But the second
> one (with +-37 volt rails) requires transistors with at
> least something like 80 volt capability.
> From:http://www.fairchildsemi.com/ds/TI%2FTIP41C.pdfhttp://www.fairchildsemi.com/ds/TI%2FTIP42C.pdf
> the 100 volt rating looks like a possible.
>
> The gain falls from a typical peak of something like 55
> around 0.3 amps, but falls to about 27 at 5 amps, so that is
> all the peak current I would ask of these transistors.  That
> would limit the load resistance to about 7 ohms or more.
>
> Checking the second breakdown (safe operating area) curves,
> you might stay within the safe area with an 8 ohm load.  But
> I think you are risking damage with a 4 ohm load.  Lowering
> the supply voltages a bit helps a lot (lowers the peak
> current and the peak voltage across the transistor with an
> inductive speaker load that keeps the load current going a
> bit as the voltage swings back through zero, so supply
> voltage across the transistor while current is flowing,
> briefly).
>
> So these might serve for some loads, but will not withstand
> much abuse, like paralleled speakers or speaker wiring shorts.
>
> The first circuit does not detail the Vbias part of the
> circuit between the two output half drivers, but that bias
> has to see the transistor temperature, so it can compensate
> for the changes in base emitter drop of the hot output
> transistors.  the second schematic shows one way to make
> such a bias generator, and TR13 should be thermally coupled
> to the output heat sink.
> --
> Regards,
>
> John Popelish



And here I thought that with +/- 37V, with each transistor only seeing
|37| V, the 40V one would be fine...

I'll have to research in detail what "safe operating area" for
transistors means.  (I'm still new at this.)

Thanks!

Michael

mrdarrett@gmail.com wrote:

> And here I thought that with +/- 37V, with each transistor only seeing
> |37| V, the 40V one would be fine...

When one transistor is fully on, the other has the sum of 
both supplies across it.  If the load is not pure resistance 
(part inductance or capacitance), there is also some current 
through the load as it passes through zero volts, instead of 
the current being instantaneously proportional to the 
voltage across the load.  This means that it is possible 
that there will be almost the full sum of the two supplies 
across a transistor and it will still have to be conducting 
some current.

> I'll have to research in detail what "safe operating area" for
> transistors means.  (I'm still new at this.)

The two dimensions of the safe area are voltage and 
simultaneous current (instantaneous power).  The higher the 
voltage across the transistor, the smaller part of the total 
die conducts that current, so the power gets concentrated in 
small hot spots.  The die cannot heat sink those hot spots 
(and keep them below the peak allowed temperature) as well 
as it can spread the heat of full die heat production, so 
the power capability goes down and the duration of peak 
power goes down.  Very short duration heat pulses can be 
soaked up by the die, itself, but cannot get out to the 
package fast enough.  This is why the safe area chart 
contains several lines representing voltage and current 
combinations of various durations.

-- 
Regards,

John Popelish

On Wed, 15 Oct 2008 20:37:13 -0400, John Popelish <jpopelish@rica.net>
wrote:

>mrdarrett@gmail.com wrote:
>
>> And here I thought that with +/- 37V, with each transistor only seeing
>> |37| V, the 40V one would be fine...
>
>When one transistor is fully on, the other has the sum of 
>both supplies across it.  If the load is not pure resistance 
>(part inductance or capacitance), there is also some current 
>through the load as it passes through zero volts, instead of 
>the current being instantaneously proportional to the 
>voltage across the load.  This means that it is possible 
>that there will be almost the full sum of the two supplies 
>across a transistor and it will still have to be conducting 
>some current.
>
>> I'll have to research in detail what "safe operating area" for
>> transistors means.  (I'm still new at this.)
>
>The two dimensions of the safe area are voltage and 
>simultaneous current (instantaneous power).  The higher the 
>voltage across the transistor, the smaller part of the total 
>die conducts that current, so the power gets concentrated in 
>small hot spots.  The die cannot heat sink those hot spots 
>(and keep them below the peak allowed temperature) as well 
>as it can spread the heat of full die heat production, so 
>the power capability goes down and the duration of peak 
>power goes down.  Very short duration heat pulses can be 
>soaked up by the die, itself, but cannot get out to the 
>package fast enough.  This is why the safe area chart 
>contains several lines representing voltage and current 
>combinations of various durations.

---
Beautiful. 

JF

<mrdarrett@gmail.com

>I noticed that TR7 is PNP and TR9 is NPN here:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif
>
> Then over here, TR7 is NPN and TR9 is PNP:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif
>
> Are these correct, or in error?
>
> Also - for building audio amps "on the cheap", would TIP41C/TIP42C do
> for output transistors in these schematics?


** No.

 Piss off  -  you utter dickhead.



.....  Phil

On Oct 15, 5:37 pm, John Popelish <jpopel...@rica.net> wrote:
> mrdarr...@gmail.com wrote:
> > And here I thought that with +/- 37V, with each transistor only seeing
> > |37| V, the 40V one would be fine...
>
> When one transistor is fully on, the other has the sum of
> both supplies across it.  If the load is not pure resistance
> (part inductance or capacitance), there is also some current
> through the load as it passes through zero volts, instead of
> the current being instantaneously proportional to the
> voltage across the load.  This means that it is possible
> that there will be almost the full sum of the two supplies
> across a transistor and it will still have to be conducting
> some current.
>
> > I'll have to research in detail what "safe operating area" for
> > transistors means.  (I'm still new at this.)
>
> The two dimensions of the safe area are voltage and
> simultaneous current (instantaneous power).  The higher the
> voltage across the transistor, the smaller part of the total
> die conducts that current, so the power gets concentrated in
> small hot spots.  The die cannot heat sink those hot spots
> (and keep them below the peak allowed temperature) as well
> as it can spread the heat of full die heat production, so
> the power capability goes down and the duration of peak
> power goes down.  Very short duration heat pulses can be
> soaked up by the die, itself, but cannot get out to the
> package fast enough.  This is why the safe area chart
> contains several lines representing voltage and current
> combinations of various durations.
>
> --
> Regards,
>
> John Popelish


Thank you for taking the time to explain this to me.

I appreciate it.

Michael

mrdarrett@gmail.com wrote:

> I noticed that TR7 is PNP and TR9 is NPN here:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig1.gif
>
> Then over here, TR7 is NPN and TR9 is PNP:
> http://www.dself.dsl.pipex.com/ampins/dipa/dpafig33.gif
>
> Are these correct, or in error?
>
> Also - for building audio amps "on the cheap", would TIP41C/TIP42C do
> for output transistors in these schematics?

Since the original devices aren't mentioned, it's anyone's guess. Do you
know what fT means ? And why it's important for an amplifier output
stage.

TIPs are generally fairly shitty for audio IME.

Graham