"Don Pearce" wrote in message
...
On Wed, 4 Nov 2009 07:52:21 +1100, "Trevor Wilson"
wrote:


"Don Pearce" wrote in message
...
On Wed, 4 Nov 2009 07:06:02 +1100, "Trevor Wilson"
wrote:


"Pete" wrote in message
.. .
I am looking to build an amp for home theatre/hifi use.
currently have an amp using Hitachi's default design for mos fets from
the
'80's ( 2sk49 2sj50 I think ). It is used with a tagmclaren av32r and
ipl
acoustics tl4 speakers at the front and Ariston ( small black tubular
things that sound quite good ) at the back.

I plan using LM4702 but have three options for the output stage.

bipolar
SAP15P and SAP15N
2SA1943 and 2SC5200
mosfet
2SK1530 and 2SJ201

I know I'm inviting WW3 but would there be any differance in sound
between
them ?

**Depends on implementation. The MOSFETs, of course, will always sound
like
MOSFETs - dull, boring, undynamic and compressed. The 2SA1943/2SC5200
devices are amongst the most advanced, lowest distortion output devices
developed. The SAP15s are easy and convenient to implement, but you run
the
risk of using a device unique to a single manufacturer. Once that
manufacturer deletes them from the product line-up, you'll be screwed
for
spare parts later on. Personally, I like the 2SA1943/2SC5200 devices,
Collectors coupled to the load.

Compressed?

**Yep.

You can of course produce a transfer characteristic to
back this up?

**Negative Tempco of gm.

As far as I am aware it is perfectly possible to produce
a MOSFET design without a trace of compression - which would of course
show up as an unimaginably huge distortion figure.

**What can I say? It what they sound like, when operating in low bias
Class
A/B. When operating in high bias Class A/B or Class A, they show no signs
of
compression. At low currents, they suck. Ask anyone who has taken the time
to do a double blind listen. I have. Many times. Shockers.

Negative tempco makes no difference. The gain of a power amplifier is
set by a pair of feedback resistors. Provided there is sufficient open
loop gain available (and we must assume there is) there can be no
compression. The system will perform at constant gain up to the
clipping point just like any amplifier.

I presume you are treating this as an interview for a job reviewing in
Stereophile?

**I doubt that. I have been critical of Stereophile many times. I have also
performed several DBTs involving MOSFET amps and BJT amps. The first was way
back when Perreaux first burst onto the scene with their MOSFET amps. I
wondered at what I was hearing, so I set up a Phase Linear 400 amp for
comparison. The Phase Linear was determined (not just by me) to be much more
dynamic sounding. The only MOSFET amps I've heard which do not suffer this
problem are those manufactured by Pass Labs.

Tell me about your blind listening experiences with MOSFET amps and BJT
amps.


--
Trevor Wilson
www.rageaudio.com.au