MOSFET output stage

[Mr.T]

Good
power amps sound the same and they sound like a piece of wire with gain.

There are tons of power amps that can't be distinguished from a piece of
wire with gain, while driving well-designed speakers.

Ah, but since audiophools argue endlessly about cable and wire
characteristics, you need to specify exactly what TYPE of wire with gain.


MrT.

 

[Mr.T]

Please explain how you obtain even distortion products from a
symmetrical (suppose paired mosfets) push-pull.

Any symmetrical function produces odd (and no even) harmonics.

But isn't that part of the problem, producing Mosfets with identically
symmetrical characteristics in N-ch and P-ch ?
Of course good circuit design can compensate for many such factors, so the
final result is what is important, not some evangelical belief, IMO.

MrT.

 

[Kevin McMurtrie]

Agreed. Actually, what Kevin said is the exact reverse of generally accepted
practical knowledge. Bipolar is generally more efficient when the impedances
get really low. MOSFETs were trendy for a while, but most new power amp
designs seem to have bipolar outputs.

Time to check the specs for this decade. Visit IRF. You can get a
surface mount MOSFET in package similar to TO-220 that has 0.0008 Ohms
on resistance, 300W dissipation, 24V rating, and a 1600A surge rating.

Now I remember why I left this newsgroup.

 

[Phil Allison]

"Mr.T"

"Fred Bartoli"
Eeyore

But isn't that part of the problem, producing Mosfets with identically
symmetrical characteristics in N-ch and P-ch ?

** For switching MOSFETs ( ie Hexfets etc ) there are no true
complimentary P and N pairs - not even close. The P ch types all have
much higher "on" resistances and lower max drain current ratings compared to
their Nch counterparts.

OTOH - lateral MOSFETs ( ie 2SK176 and 2SJ56 etc ) are near perfect
Nch and Pch compliments - varying slightly in gate capacitance.

More symmetrical in an audio output stage than nearly any BJT compliments
around.




...... Phil

 

[Phil Allison]

"Kevin McMurtrie"

Time to check the specs for this decade. Visit IRF. You can get a
surface mount MOSFET in package similar to TO-220 that has 0.0008 Ohms
on resistance, 300W dissipation, 24V rating, and a 1600A surge rating.

** So you can use it to replace the engine starter relay in a car.

Yawnnnnnnn ....




....... Phil

 

[Chronic Philharmonic]

The first two are highly relevevant in ANY circuit using NFB. Basic
stability criteria.

Only if the bandwidth and the phase margins are small relative to the target
audio bandwidth, which is unlikely.

 

[Eeyore]

But isn't that part of the problem, producing Mosfets with identically
symmetrical characteristics in N-ch and P-ch ?
Of course good circuit design can compensate for many such factors, so the
final result is what is important, not some evangelical belief, IMO.

Fred is thinking of push-pull transformer coupled arrangements. He needs to
advance about 4-5 decades.

That's one reason I never call modern amps push-pull but complementary
symmetry instead.

Graham

 

[Eeyore]

Time to check the specs for this decade. Visit IRF. You can get a
surface mount MOSFET in package similar to TO-220 that has 0.0008 Ohms
on resistance, 300W dissipation, 24V rating, and a 1600A surge rating.

These are NOT used in audio amps.

Learn something about LATERAL mosfets that were designed for audio. I've already
given part number and links to data sheets.

Graham

 

[Eeyore]

IR is famous for creating incredible specs, with tiny footnotes
retracting the wilder numbers. No TO-220 is going to last long at 300
watts; not many milliseconds.

You simply can't get that many watts through that tiny tab of copper fast enough. I
think the highest continuous ratings are in the 50-70W area.

Graham

 

[Eeyore]

Only if the bandwidth and the phase margins are small relative to the target
audio bandwidth, which is unlikely.

Modern audio amplifiers are expected to have a bandwidth far greater than 20kHz.
If not, you may experience slew limiting problems and associated THD and IMD
IIRC.

Yet another one totally out of date with modern practice.

Graham

 

[Spehro Pefhany]

IR is famous for creating incredible specs, with tiny footnotes
retracting the wilder numbers. No TO-220 is going to last long at 300
watts; not many milliseconds.

John

Anything approaching 1600A would blow the leadwires off a TO-220. The
fusing current of a long wire of that cross-section area is only in
the 30-50A range. 100A-rated wire is around 6mm (1/4") in diameter.


Best regards,
Spehro Pefhany

 

[Phil Allison]

"Spehro Pefhany"

Anything approaching 1600A would blow the leadwires off a TO-220.

** Not if it is only for a few microseconds.

That is how the things are usually rated.



...... Phil

 

[MooseFET]

Not in any relevant way for audio power amps.

This is not true.

The phase shift in the output section of a power amplifier is a
significant issue. If you want to have much of a gain sacrifice
factor in the audio band, the gain cross over point will be well above
the audio band.

NFB can't make a silk purse out of a sows ear. It can only lower the
distortion by about the gain sacrifice factor. This makes if better
to start with a low distortion topology.

Rumors aside MOSFETs are not really easier to operate in parallel.
You still need emitter / source resistors.

 

[MooseFET]

The first two are highly relevevant in ANY circuit using NFB. Basic
stability criteria.

For Christ's sake has everyone except Phil's and my brain turned to jelly
overnight ?

Checking brain ..... Mashed potatoes detected

Mine hasn't.

 

[MooseFET]

The device properties of BJTs are superior to those of MOSFETs in all
respects, except for offset - there MOSFETs have the advantage. Whether you
will actually hear this depends on many more factors.

Radiation hardness
Safe operating area
Bandwidth

 

[Phil Allison]

"MooseFET"

Rumors aside MOSFETs are not really easier to operate in parallel.
You still need emitter / source resistors.


** Which MOSFETS ????

Amazing how so many MOOSE like idiots have no idea there are TWO kinds
!!!!!!!!!!!!!!

The kind known as " lateral " share current just fine when in parallel in
LINEAR applications with no source ballast resistors.

Septic Tank Imbecile.





...... Phil

 

[MooseFET]

Fine, fine, but that doesn't have any direct influence on what you'll hear,
because it's a boundary condition.

When I hear BLAM-FIZZZT-POP-KAPOW from an SOA problem, I gain and
understanding of how the SOA of a BJT is not better than that of a
MOSFET.

Like I said, it's a boundary condition. It can influence the performance of
the circuit, but only indirectly, through other design decisions.

You can design around a problem with a device but that doesn't make
the device not have that problem.

Now you're being the moron (not that I admit to being a moron before). It
seems that either you don't know what you're talking about, or personal
attacks are a hobby of yours. Because, frankly, everything I said was
true...

Fact - MOSFETs have lower offset than BJTs.

How are you defining offset?

Fact - The fidelity of the sound depends on much more than just device
properties.

Yes a Class-C audio output sounds quite bad regardless of the type of
device used. It sure gets rid of that annoying hiss.

 

[Eeyore]

Anything approaching 1600A would blow the leadwires off a TO-220. The
fusing current of a long wire of that cross-section area is only in
the 30-50A range. 100A-rated wire is around 6mm (1/4") in diameter.

Hence those ceramic power puks and the like.

Graham

 

[MooseFET]

[... BJTs are better ....]

As for offset, here's one explanation (in the context of opamps):http://en.wikipedia.org/wiki/Operational_amplifier#DC_imperfections

Note that I never claimed that this is relevant in audio applications - but
it's there.

You may not have claimed it mattered but you sure implied it. That
aside since we are talking about the output section of an amplifier,
this sort of offset has no effect at all on the quality of the op-amp.

If you look for an op-amp with an extremely low offset voltage, you
will find that it uses MOSFETs to obtain that extreme low offset.

 

[Eeyore]

"Eeyore" wrote


Agreed. Speakers that require heroic amplifiers (heroic amps definitely
exist and are easy enough to obtain) are IMO not well designed.


Agreed. However we have only part of the equation - the name of the
speaker. What was the amp?

With our own D Series amps. It was the only speaker we had trouble with. The VI
limiting was quite generous for a 4 ohm rated amp but not enough for an SX500. I
slackened off the VI protection and in the later AX series took a totally
different approach to protection.

The SX 500i EDS suggests that it might be a handful for weak-kneed amps. Its
impedance curve does go down to 4 ohms at several points below 1 KHz.

It's the phase angle that's insane more than the modulus of impedance. A D
series amp would drive 3 typical 8 ohm speakers in parallel quite happily.

http://www.electrovoice.com/documents/ev/sx500-eds.pdf

Using pairs paralleled can be a short road to amp problems.

In contrast, the sequel ZX-5 stays above 8 ohms over the same range:

http://www.electrovoice.com/documents/ev/ZX560PI_EDS.pdf

Needless to say, I have 4 ZX-5s. ;-)

A better choice I imagine. I didn't even like the sound of the SX500s, very
harsh and boxy.

Graham