Re: Bipolar Transistors for Audio


"Alan Rutlidge" <don't_spam_me_rutlidge@xxxxxxxxxxxx> wrote in message
news:47a444ea$0$22559$5a62ac22@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

"Trevor Wilson" <trevor@xxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:60ibbkF1qaih1U1@xxxxxxxxxxxxxxxxxxxxx

"Bret Ludwig" <bretldwig@xxxxxxxxx> wrote in message
news:9f7e05cf-94d8-4eff-b883-8b596b1b852d@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
((Arny doesn't build transistor amplifiers either, so this kind of
immaterial. Oh well.-Bret))




Bipolar Transistors for Audio
By John Atwood

Toshiba and ON Semi transistors

**SNIP> TEN YEAR OLD news. Sheesh! Get with the programme. Any designer
worth his/her salt has been using the Toshiba semis for more than 10
years. In fact, there's newer, better models available. No sane designer
bothers with MOSFETs any more. The new BJTs are vastly superior and well
priced.

Trevor Wilson


Admittedly technology has moved on, but I've never had any issues with
MOSFETS.
Okay they do have low gain and in power amp output stages require a lot of
voltage drive which can be easily obtained by using high voltage bjt
devices with a good current gain. Ones with a high slew rate like those
used in TV output deflection circuits are cheap and do the job just
nicely. The only problems commonly encountered with DIY MOSFET designs is
board layout and some potential RF instability. The later can often be
controlled by slipping a ferrite bead over the gate leads of each device.
Never the less, careful consideration to the component layout is often the
best approach in the first place to prevent such problems from arising.

**Indeed. Speed is a MOSFET's greatest virtue. It lacks 'storage effects'
which causes BJTs to overheat. Additionally, MOSFETs enjoy a characteristic
known as 'negative temperature coefficient of gm'. This makes MOSFETs all
but indestructible. It is also their Achille's Heel. IOW: As more current is
demanded from the device/s, the gain throttles back (which is pretty much
the opposite of BJTs). It is this characteristic which makes MOSFET amps
sound 'wimpy' and lacking in 'impact'. But wait: There is worse to come.
MUCH worse. MOSFETs are quite linear, PROVIDED they are operated very
heavily biased. Their 'knee' occurs at currents which push MOSFET amps well
into Class A operation. BJTs, OTOH, are linear from very low bias currents.


Like lots of things, what's in vogue today is yesterday's has-been.

**Not really. MOSFETs have always been vastly over-rated as audio
amplifiers. They are, however, the best way to go for swithc mode (aka:
Class D, digital) amplifiers.

The
fickle consumer audio world rarely stops to consider some of the classic
timeless designs. IMHO the best book on amplifier design was "Amplifier
and Low Pass Filter Design" by Cherry and Hooper. An excellent text
(albeit a bit dated) and essential reading for anyone considering the
design of their own. It covers the basics and more importantly explains
the rationale behind each design type.

Newer isn't always better. Take for instance the ME 850 / 900 design.
http://www.me-au.com/html/me850hc1.html Uses bi-polar output devices that
must be 20 years old if they are a day. Still going strong and obviously
still meet the requirements of the 21st century audiophile.

**Indeed. However, in order for those devices to operate linearly, ME uses
12 devices per channel (4 would be plenty in a conventional design), thus
ensuring that the output devices rarely operate outside their linear region.
Having said that, though, ME will be releasing an upgrade to the ME850 which
incorporates the new, Japanese output devices. To compensate for the lower
SOA of the Japanese devices, there may as many as 24 output devices per
channel (final design is yet to be settled). Distortion should be halved (at
least), maximum current output should be increased to around 100 Amps and
some improvement to sound quality is expected.

Trevor Wilson

Trevor Wilson


.

Relevant Pages

  • Re: inductor sizing
    ... capacitors and maybe ainductorto help with ripple ... The mosfets will be PWM controlled ... Please post your circuit and we ... comfortable with the design and implementation of this kind of high ...
    (sci.electronics.design)
  • Re: amplifier question
    ... resolved by using MOSFETs as the output devices? ... Using MOSFETs doesn't eliminate the problem, ... make the design a whole lot easier and it does prevent the thing from ... of dollars to cut a prototype, or if you are trying to push the limits ...
    (rec.audio.pro)
  • Re: HV gain stage with fewest number of transistors
    ... >> Given that 1200V MOSFETs are readily available, ... > all_NPN design. ... It's often called a totem-pole output stage, ... it in the late 60s in DTL logic gates. ...
    (sci.electronics.design)
  • ringing with mosfets problems
    ... I've converted a H-bridge from a P-N type design to using all N-chan ... The mosfets actually start to ...
    (sci.electronics.misc)
  • Re: Stereo receievers: THD of .04 vs. .08
    ... struggling with output devices with Ft in the KHz ... output devices with load current. ... You COULD build an indifferent amplifier design with high ... a believer in commercial amps. ...
    (rec.audio.tech)