Re: lateral mosfets vs. bjts in audio amplifier design

[[email protected]]

To do the RMS addition calculation backwards.


It was an 8903 I probably used once on one project. Clumsy to use and poorer
residuals than AP. I'd have the AP Portable One any day over that. In fact for
simple bench work I prefer the Portable One over the System One because it's
simple 'press button' stuff, no mice to arse about with and it takes up little
room. No other audio measuring set comes close in terms of performance or
convenience.

Graham

The AP system one came out later than the HP 8903B, so you would
expect it to be better. The portable was later still. The Portable One
has an HP 8903 emulation mode, so I guess you call that clumsy
mode. ;-) Anyway, the HP8903 was good for it's time, but time marches
on.

I really don't like PC based black boxes, so the Portable One would be
my choice for an upgrade. The HP 8903 is a back breaker too.

The applications group at AP is great. The system one has this digital
interface on the back. I couldn't quite get it working, called apps,
and the person asked how many people have access to the box. It turns
out the interface was done with simple gates (TTL I think) and often
people blow them up. The bigger the lab, the more likely someone blows
them up and doesn't own up to it. Well, that was the case, but
fortunately the interface chips are in sockets.

I still have my HP 334A. It is the only THD analyzer I've seen with RF
demod. I can't say the box has seen a power cord in a few years, but
you never know.

 

[Eeyore]

The AP system one came out later than the HP 8903B, so you would
expect it to be better.

Maybe. More to the point is that they were designed by the ex-Tek guys. I'd used
Tek's SG505 and AA501 previously. And they were quite OK too.

The portable was later still. The Portable One
has an HP 8903 emulation mode, so I guess you call that clumsy
mode. ;-) Anyway, the HP8903 was good for it's time, but time marches
on.

It has an IEEE interface option I believe.

I really don't like PC based black boxes, so the Portable One would be
my choice for an upgrade.

They are very nice. For a self-contained unit I think they are simply unbeatable. And
so well built too !

The HP 8903 is a back breaker too.

I agree.

The applications group at AP is great.

Always has been with regular app notes.

The system one has this digital
interface on the back. I couldn't quite get it working, called apps,
and the person asked how many people have access to the box. It turns
out the interface was done with simple gates (TTL I think) and often
people blow them up. The bigger the lab, the more likely someone blows
them up and doesn't own up to it. Well, that was the case, but
fortunately the interface chips are in sockets.

I still have my HP 334A. It is the only THD analyzer I've seen with RF
demod. I can't say the box has seen a power cord in a few years, but
you never know.

Many years ago I bought some used B&K and hardly ever used it. Technology marches on
!

Graham

 

[Eeyore]

Find fault with them then !

You're a babe in arms in this discipline.

Graham

 

[Eeyore]

---
Really?

You make those claims and then, conveniently, fail to back them up
because of the restrictions you claim are placed upon you by copyright
and contractual limitations.

I understand your frustration, but I don't think you're important.

If you were, you'd have technically relevant articles published in peer
reviewed journals or, at the very least, cogent discourse posted here,
on USENET.

As it is, all you seem to be intent on doing is promoting yourself as
some sort of audio Guru who claims to greatness but who refuses to
submit proof.

Suit yourself. Everything I claim does what it does. All you have to do is
buy one and test it yourself.

I currently have a bit of a quandary. I've been offered 2 serious jobs this
week. And I can't even begin to tell you the background, it's so
complicated.

I've made one decision though. If one outfit wants me it'll be Group
Technical Director.

Graham

 

[Eeyore]

---
Really?

You make those claims and then, conveniently, fail to back them up
because of the restrictions you claim are placed upon you by copyright
and contractual limitations.

Fine. Pay me and I'll do something similar for you and you'll own the
copyright.

<snip bollocks>

Graham

 

[[email protected]]

And it was developed under contract for a client so it's not really mine to give
away.



I don't care for Doug Self too much come to that. Long story.

Graham

You don't care for Doug Self? Then why, in the Class A thread, did
you point the OP to Doug Self's website???

Michael

 

[Kevin Aylward]

Not too bad,

Yes. "not too bad" in the strict sense of the phrase. That is, it is a bad
design, but not so bad as to be totally unusable. However, this does depend
on the definition of "good" somewhat. Assuming we give meaning to the term
"average", then this design can not in any reasonable way be classed as a
"good" or even an average design". It is a quick knock up, that serious
audio designers don't even give a second look to, well except when
commenting on how bad it is.

except for the biasing. Iq could have been set by a
pot+resistor from IC1 pin 4 to 7, and the DC offset trim should be
into the opamp, not fighting it.

A "good" audio amp, imo, will have at most, only one trim pot, and this is
to set the output bias.

The biasing/crossover of this type of circuit is potentially perfect,
in that both stages run at some idle current and signal makes one pick
up,

For very high speed, audiophile performance, this output device
configuration is very poor. The gates of the transistors are connected via
two much speed lag. This usually results in very, large shot through
currents when hit by 100n pulses.

but leaves the opposite side idling. As opposed to a lot of
circuits where conduction on one side actively shuts off the other.

DC bias on the output fets depends on their thresholds, not so good;
closing local loops on the fets (with more opamps) would be better.

Oh...a sure-fire recipe for disaster, if done correctly...It will generate
additional de-stabilising poles. That particular technique can be useful for
reducing LF. distortion, but it does it at the expense of BW, i.e. higher HF
distortion. I would be surprised if this "design" could stand having its HF
IMD compromised any further.

Driving the load from the drains means the amp has a very high
open-loop output impedance,

Well, not so high in this particular case. Assuming a nominal early voltage
of 30V, at 1A would be 30 ohms. Its highly variable though.

which makes it harder to stabilize.

Can be. This is actually a bit subtle. Depending on just what the actuall
compansation is. A correctly designed amp will have a stabilastion load
network, in which case, the stability may well be dominated by a higher
frequency unity gain point In which case, it turns out that the high
frequency UGF of a source follower and a drain ouput is the same.

Flipping things over, driving from sources, makes the crossover
biasing more interesting and costs swing.

At least it's not the same old 40-year vintage class AB thing.

Maybe it should be. It would then, presumably have the output inductor and
|| resistor with an output RC zobel network, that is pretty much mandatory
for stability under all load conditions. I would also suggest looking at a
direct zener clamp on the mosfets.

Unfortunately, I don't have the time to go over the other considerable
shortcomings of this er.. "design"

Kevin Aylward
www.kevinaylward.co.uk
www.blonddee.co.uk
www.anasoft.co.uk -SuperSpice

 

[Eeyore]

You don't care for Doug Self? Then why, in the Class A thread, did
you point the OP to Doug Self's website???

Because whatever I may feel about him, the info is good.

Graham

 

[Eeyore]

A "good" audio amp, imo, will have at most, only one trim pot, and this is
to set the output bias.

I was delighted with the consistency of my D Series of bipolar amps. No trim
pots at all. Designed out in the design process. In fact the entire amp had not
one trimpot. That means faster production and final test plus a cost reduction.

Graham

 

[Eeyore]

It would cost you LOTS more to build one yourself and you wouldn't replicate the
pcb pattern which can be critical for top performance.

Graham

 

[Eeyore]

Well, let's face it, you don't have the necessary skills.

Graham

 

[Eeyore]

---
You wish.

I cut my teeth on audio and was designing and building bridge amplifiers
in the early '60's, even before RCA came out with them, as I recall.

The world has moved on John.

Graham

 

[Eeyore]

I'd argue
that adding an opamp per power fet makes things faster, stability
better, and compensation simpler, since the gate/Miller capacitance
disappears... each fet now looks like a very fast, pF input
capacitance, DC-perfect device, and essentially disappears from the
overall loop dynamics. Now just pile on as many opamp+fet pairs as you
need. DC balance and current sharing become as good as the opamp
offset voltages, microvolts if you like, so fet gate threshold
variations and transfer curves don't matter any more. So use very
small source resistors and cut losses.

It's certainly a very interesting topology.

Graham

 

[Eeyore]

Make sure you have a shovel, you'll need it for that hole
your about to dig for your self.

I'm quite sure that JF is far too out of touch with the technology currently
used in the best audio amplification.

Also, it has to be economic to manufacture. That's easily half of it.

Graham

 

[Jamie]

John Fields wrote:




It would cost you LOTS more to build one yourself and you wouldn't replicate the
pcb pattern which can be critical for top performance.

Graham

LOL!!
the crap is getting deep!


http://webpages.charter.net/jamie_5"

 

[Jamie]

John Fields wrote:




Well, let's face it, you don't have the necessary skills.

Graham

Make sure you have a shovel, you'll need it for that hole
your about to dig for your self.


http://webpages.charter.net/jamie_5"

 

[Kevin Aylward]

My NMR and MRI gradient amps use a similar topology... the driver
opamp has its supply rails cascoded to make signal currents, and the
upper and lower power stages are active current mirrors. I'd argue
that adding an opamp per power fet makes things faster, stability
better, and compensation simpler, since the gate/Miller capacitance
disappears...

And your argument would be incorrect I am afraid to say. I have already
outlined these issues recently in another post.

There is no such thing as a free lunch. You cannot achieve better stability
by using extra feedback loops in this manner. The transfer function of the
op-amp and fet combined will cause considerable grief. Only if you roll off
the system much earlier will it be stable.

each fet now looks like a very fast, pF input
capacitance, DC-perfect device, and essentially disappears from the
overall loop dynamics.

Unfortunatly not. Consider a two stage amp circuit, one first stage, being
the conventional gain stage, followed by a second stage connected with a
local loop to make a CLG of unity. Additionally, the second amp connecting
back to the 1st (via a beta network if used) to form the composite. Now
calculate the system loop gain by breaking both feedback connections *at
once* an compare it with the 2 stage system both running open loop. The
transfer functions are identical. So, having the two stages means at least 2
roll offs. That is including an *additional* op-amp (to get better lf
performance) around the mosfet, will generate an additional pole that would
not have been there. Therefore the system is inherently more unstable.

I suggest you actually perform the *detailed* simulations of these type of
circuits.

I point out two circuits

http://www.kevinaylward.co.uk/ee/circuits/VeryLowDistortionAmp1.jpg
http://www.kevinaylward.co.uk/ee/circuits/VeryLowDistortionAmp2.jpg

They are similar, but one has a local feedback loop around the output
devices, one doesn't. One has better LF accuracy, one can be stabilised to a
higher UGF. Which is which?

The fundamental trade off here is basic physics, there is an inherent
constraint of H(Power, Accuracy, Speed) = 0, e.g.
http://www.kevinaylward.co.uk/ee/cmospafl/cmospafl.html


You just can't just get better LF performance, and expect to get it all. It
don't work that way. Like,....


When I was young, I prayed and prayed to the Lord to get me a bike. After
many years, and no success, I realised that HE, doesn't work that way...So I
stole the bike, and preyed for his forgiveness...

Now just pile on as many opamp+fet pairs as you
need. DC balance and current sharing become as good as the opamp
offset voltages, microvolts if you like, so fet gate threshold
variations and transfer curves don't matter any more. So use very
small source resistors and cut losses.

For LF, the op-amp approach is very nice, for HF, its not so nice.

My amps often work in pulse mode, when doing chemical NMR. They settle
to PPMs of the target value (which is current, since we're driving
gradient coils) in 10's of microseconds.

This is very slow, by about 2 orders of magnitude at least. At this slow
speed, op-amps are probably a good choice. You just clobber with a big cap.
By fast, I meant at the < 100ns level.

Kevin Aylward
www.kevinaylward.co.uk

 

[Kevin Aylward]

I was delighted with the consistency of my D Series of bipolar amps.
No trim pots at all. Designed out in the design process. In fact the
entire amp had not one trimpot. That means faster production and
final test plus a cost reduction.

It would be a good idea if those lateral mosfet makers had current sense
transistors so that a current mirrored, push pull, source follower can be
done.

Kevin Aylward
[email protected]
www.kevinaylward.co.uk

 

[Jamie]

Jamie wrote:




I'm quite sure that JF is far too out of touch with the technology currently
used in the best audio amplification.

Also, it has to be economic to manufacture. That's easily half of it.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
that being the case, then it must be shit you're making!
I think any beginner can accomplish that.

Graham

http://webpages.charter.net/jamie_5"

 

[Eeyore]

LOL!!
the crap is getting deep!

You clearly haven't the tiniest clue about the economics of high-volume manufacturing.
Or the importance of pcb layout. Not to mention the cost savings of offshore
manufacturing too.

One product of mine (in its various channel sizes) sold over 100,000 units.

Graham