More than 30W per chanel Class A transistor amps

On Monday, 4 March 2019 16:06:40 UTC, Peter Wieck wrote:
> On Monday, March 4, 2019 at 9:59:57 AM UTC-5, Howard Stone wrote:
>
> > Aha, this is an argument for active pre-amps. Passive pre-amps presumably all have no gain.
>
> Correct. Passive pre-amps are no more than switches.
>
> Peter Wieck
> Melrose Park.

So I guess that, given that my project is about exploring Class A with my speakers, that IF I buy a pre-amp, it too needs to be Class A.

And I suppose I find out whether there are any Class A DACs, because I think their analogue end must involve some gain.

On Monday, March 4, 2019 at 11:06:40 AM UTC-5, Peter Wieck wrote:
> On Monday, March 4, 2019 at 9:59:57 AM UTC-5, Howard Stone wrote:
>
> > Aha, this is an argument for active pre-amps. Passive pre-amps presumably all have no gain.
>
> Correct. Passive pre-amps are no more than switches.

And noise and distortion generators.

--
+--------------------------------+
| Dick Pierce |
| Professional Audio Development |
| cartchunk.org |
+--------------------------------+

It's only money: http://www.avalondesign.com/pdfs/hr_...MPCHARTnew.pdf

Or, not so much:
https://www.aliexpress.com/item/Lusy...4eaefa96d67-25

Peter Wieck
Melrose Park, PA

On Monday, March 4, 2019 at 11:32:18 AM UTC-5, Howard Stone wrote:
> On Monday, 4 March 2019 16:06:40 UTC, Peter Wieck wrote:
> > On Monday, March 4, 2019 at 9:59:57 AM UTC-5, Howard Stone wrote:
> >
> > > Aha, this is an argument for active pre-amps. Passive pre-amps presumably all have no gain.
> >
> > Correct. Passive pre-amps are no more than switches.
> >
> > Peter Wieck
> > Melrose Park.
>
> So I guess that, given that my project is about exploring Class A
> with my speakers, that IF I buy a pre-amp, it too needs to be Class A.

The chances you're going to come across a preamp whose active
'stages are NOT biased class-A is pretty small. I can think of none
of the top of my head. (Some earlier IC-based preamp may have used
op-amps that had a push-pull output stage that may have run class
AB, but, again, I can think of none).

> And I suppose I find out whether there are any Class A DACs,
> because I think their analogue end must involve some gain.

Same goes here as well: you're not likely to find a DAC output
stage that is NOT class-A biased.

NOTE: A "DAC" really consists of two fundamental parts: the
digital-to-analog stage, which, these days, will also
include the bulk of the anti-imaging filtering, done
in the digital domain: this part feeds an analog signal
to the line driver stage, which, again, is very unlikely
to be anything but class-A.

Now, the issue of whether an audio power amp is biased class-A
vs class-AB or even class-B (and, yes, even various switch-mode
implementations) is a design issue forced on the designer byu
considerations of efficiency and thus weight (less efficiency
means more copper and iron in the power supply) heat dissipation
(larger heat sinks, bigger chassis area for heat sinks), cost
and therefore success in the chosen market segment. When you're
talking about even, say, a modest output power of 50 watts/channel
continuous, there's a very large difference in whether you
implement it as a class-A or class AB, and even withing those,
a large range of demands on power, heat and the subsequent costs.

However, in a preamp, where the total power output requirements
are orders of magnitude less, these considerations become far
less of a design driver and thus do not force decisions between
biasing classes that are either engineering or economically
driven.

Putting it more simply, it's MUCH easier to design, manufacture
and support class A amplifiers dealing with low(er) signal levels
and substantially lower powers than it is to to the same where you
need LOTS of power.

Case in point: consider a sophisticated solid state preamp with,
maybe, 4 gain stages (RIAA, gain/balance, tone and output), using
the output stage as one instance, say it's rated at 2VRMS into
10 kOhm load: that's, lessee 0.02 uW (or, to make it even more
impressive sounding: 20 billionths of a watt). And let's say each
stage has roughly the same power output requirement. The difference
in heat dissipation and power supply requirements from running all
those stages at class A vs all of them at some class-AB is probably
on the order of the amount of power required to run the front panel
lights.

Plus the fact that, in the context of audio amplifiers, the choices
over whether to run things class-A vs class-AB is ONLY applicable
to push-pull topologies. To flip this around, if you wanted to have
the choice of running things class-A vs class-AB in a preamp, you
would have had, first, to make the design choice that you're designing
a push-pull preamp from input to output.

Don't spend a lot of time worrying about whither the preamp you
might buy is class A or class AB: because you'd be spending a HUGE
amount of time searching for the class-AB preamp that you don't want
to buy.

--
+--------------------------------+
| Dick Pierce |
| Professional Audio Development |
| cartchunk.org |
+--------------------------------+

On 4/03/2019 11:45 pm, Peter Wieck wrote:
> On Sunday, March 3, 2019 at 7:21:11 PM UTC-5, Howard Stone wrote:
>
>> Everyone says that I should try it with am Audio Research valve pre -- but how can this possibly be an improvement over no pre at all???
>
> It is called "Gain" and "Transient". For the sake of simplicity and discussion, let us agree on three things:
>
> 1. Whatever amplifier is in use, it is rated at its full RMS output against a 2V input at 1,000 hz.

**Just a nit: There is no such thing as RMS power. It is either
continuous Watts, or just plain ole Watts. The moniker RMS can be
applied to Volts or Amps, but not to the product of the two.

>
> 2. It is a well-designed device using quality parts and assembled with great care.
>
> 3. The speakers in use are also good, well-made and rated at 90 dB @ 1 meter @ 1 watt. And rated at an 'average' of 50 watts.
>
> Some basics: The transient associated with a snare-drum rimshot is approximately 120 dB. The transient associated with the onset of the organ Bombard pipes in the Saint-Saens organ symphony on a well-made recording very nearly, or actually peaks at 30 dB over average.
>
> Most amplifiers are capable of momentary peaks and transients far above their rated RMS.

**Most amplifiers that employ crappy power supplies, yes. The Krell
KSA50 is not such an amplifier. It's CONTINUOUS power rating is very,
VERY close to it's transient power rating. The continuous power of a
KSA50 is around 75 Watts. The transient figure is something like 76
Watts. Had Krell decided to skimp on the power transformer, then the
difference between the transient figure and the continuous figure would
be far greater. The weird thing is the manufacturers make a big song and
dance about the size of the transient number vs. the continuous one,
when the bigger the difference, the crappier the amplifier.


Most speakers can handle transients and peaks far above their
'average' rating.

**Oh, absolutely. One need only examine the voice of a typical bass
driver and relate that to the power handling. Almost all domestic
speakers are rated to be connected to an amplifier of XX Watts and being
driven with music. Since music has, at worst a peak to average figure of
around 10dB, then one may assume that the average power applied to a
speaker is always somewhere less than 10% of the continuous power rating
of the amplifier.


>
> If a pre-amp has no gain over the original signal source, which is a nominal 2V, the peaks and transients will not be expressed - unless the actual listening level is a minimum of 30 dB below the average.

**Well that depends on the sensitivity of the amplifier.

>
> Pre-amps are awful things, just like democracy. Except for the alternatives.

**A decent preamp need not be expensive.


--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On 5/03/2019 3:32 am, Howard Stone wrote:
> On Monday, 4 March 2019 16:06:40 UTC, Peter Wieck wrote:
>> On Monday, March 4, 2019 at 9:59:57 AM UTC-5, Howard Stone wrote:
>>
>>> Aha, this is an argument for active pre-amps. Passive pre-amps presumably all have no gain.
>>
>> Correct. Passive pre-amps are no more than switches.
>>
>> Peter Wieck
>> Melrose Park.
>
> So I guess that, given that my project is about exploring Class A with my speakers, that IF I buy a pre-amp, it too needs to be Class A.

**Almost all preamps operate in Class A.

>
> And I suppose I find out whether there are any Class A DACs, because I think their analogue end must involve some gain.

**That will depend on how any OP amps used happen to be configured. I
wouldn't lose sleep over it.


--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On Monday, March 4, 2019 at 2:33:50 PM UTC-5, Trevor Wilson wrote:

> **Most amplifiers that employ crappy power supplies, yes. The Krell
> KSA50 is not such an amplifier. It's CONTINUOUS power rating is very,
> VERY close to it's transient power rating. The continuous power of a
> KSA50 is around 75 Watts. The transient figure is something like 76
> Watts. Had Krell decided to skimp on the power transformer, then the
> difference between the transient figure and the continuous figure would
> be far greater. The weird thing is the manufacturers make a big song and
> dance about the size of the transient number vs. the continuous one,
> when the bigger the difference, the crappier the amplifier.

Back to those nits: I keep a power-amp that will make a continuous 175 watts @ 8 ohms all day and all night. However, due to its power-supply reserve, output devices and so forth, it will make a transient (as a function of value and time) up to 1,000 watts for a very brief moment, less for longer moments.

> Most speakers can handle transients and peaks far above their
> 'average' rating.
>
> **Oh, absolutely. One need only examine the voice of a typical bass
> driver and relate that to the power handling. Almost all domestic
> speakers are rated to be connected to an amplifier of XX Watts and being
> driven with music. Since music has, at worst a peak to average figure of
> around 10dB, then one may assume that the average power applied to a
> speaker is always somewhere less than 10% of the continuous power rating
> of the amplifier.

Um. Not so much. Back in the day, AR made a huge point on the power-handling capacity of their speakers, and used "magic" such as ferro-fluid and so forth for heat dissipation.

a) Their speakers were rated rather low as compared to their contemporaries..
b) But their transient capacity was rated massively higher.
c) A great deal of my signal has a P/A of 20 dB. Some few examples have a P/A of 30 dB. Gregorian chant, single instrument, single voice may have the P/A of 10 dB that you reference, and those people may never turn the volume past 9:00 - but not all of us.

> > If a pre-amp has no gain over the original signal source, which is a nominal 2V, the peaks and transients will not be expressed - unless the actual listening level is a minimum of 30 dB below the average.
>
> **Well that depends on the sensitivity of the amplifier.

No. That is a fact. If the pre will put out no more than 2V, then it will drive the amp to peak for some percentage of the time. But not past the peak, ever. So, if the P/A is 10 dB, that will be well within the capacity of the amp (and the speaker). A close-run thing at 20 dB, impossible at 30 dB. But, using the example of a transient, and 90 dB efficient speakers - a perfect example of "impossible" and why I picked that particular item.


> > Pre-amps are awful things, just like democracy. Except for the alternatives.
>
> **A decent preamp need not be expensive.

That is absolutely true. A decent vintage pre may be had for well under US$200, tube or SS, and may, for approximately the same again or less, be tweaked to as-good-as anything off the shelf today.
>
>
> --
> Trevor Wilson
> www.rageaudio.com.au
>
> ---
> This email has been checked for viruses by Avast antivirus software.
> https://www.avast.com/antivirus

On 5/03/2019 7:11 am, Peter Wieck wrote:
> On Monday, March 4, 2019 at 2:33:50 PM UTC-5, Trevor Wilson wrote:
>
>> **Most amplifiers that employ crappy power supplies, yes. The Krell
>> KSA50 is not such an amplifier. It's CONTINUOUS power rating is very,
>> VERY close to it's transient power rating. The continuous power of a
>> KSA50 is around 75 Watts. The transient figure is something like 76
>> Watts. Had Krell decided to skimp on the power transformer, then the
>> difference between the transient figure and the continuous figure would
>> be far greater. The weird thing is the manufacturers make a big song and
>> dance about the size of the transient number vs. the continuous one,
>> when the bigger the difference, the crappier the amplifier.
>
> Back to those nits: I keep a power-amp that will make a continuous 175 watts @ 8 ohms all day and all night. However, due to its power-supply reserve, output devices and so forth, it will make a transient (as a function of value and time) up to 1,000 watts for a very brief moment, less for longer moments.

**And mine, with it's 5.5kVA (continuously rated) power transformer will
maintain it's power for as long as the mains breakers hold out. It's
transient power is VERY close to it's continuous power. Load impedance
is irrelevant (above 0.5 Ohms anyway). If the load impedance dips to a
low enough figure, then it can maintain a a couple of kW for a long time
(till the breaker opens).


>
>> Most speakers can handle transients and peaks far above their
>> 'average' rating.
>>
>> **Oh, absolutely. One need only examine the voice of a typical bass
>> driver and relate that to the power handling. Almost all domestic
>> speakers are rated to be connected to an amplifier of XX Watts and being
>> driven with music. Since music has, at worst a peak to average figure of
>> around 10dB, then one may assume that the average power applied to a
>> speaker is always somewhere less than 10% of the continuous power rating
>> of the amplifier.
>
> Um. Not so much. Back in the day, AR made a huge point on the power-handling capacity of their speakers, and used "magic" such as ferro-fluid and so forth for heat dissipation.

**Ferro-fluid is used only in HF drivers and some mids and, except for
NEAR drivers, not in bass drivers. And it is the bass driver that limits
the maximum thermal limits of a speaker system. Except for some pro
drivers, few domestic bass drivers can cope with sine wave power levels
of more than 50 Watts or so.

>
> a) Their speakers were rated rather low as compared to their contemporaries.

**[ANECDOTE] Back when I was a young(er) man, I built a pair of Bailey
T-lines, using KEF drivers. The bass driver was the venerable KEF B-139.
Rated at a massive 30 Watts. This was back in the day when manufacturers
like KEF rated their speaker as if they were required to handle sine
waves. As a young-ish man (early 20s) I tended to throw the odd party
for my friends. Much alcohol was consumed and much rock music was played
at high levels. Over time, I gradually improved my amplification,
culminating in my prized, Marantz Model 500 power amp. Rated at 250
Watts/channel, I measured the Model 500 at 310 Watts @ 8 Ohms. It's
superb fan cooling ensured that it could maintain high power levels at 8
or 4 Ohm loads for long periods (yes I know that maximum PDiss occurs at
around 40%). After one particularly grueling party, one of the B-139
drivers failed. It would have been at the end of around 5 hours of near
clipping levels for the Model 500. A couple of days later, I stripped
the driver down and repaired the voice coil lead in wire. No big deal.
There was absolutely no sign of heat distress on the voice coil. I
firmly believe that it was (slightly) sub-standard manufacture. However,
the same cannot be said about the electrolytic capacitors in the
crossovers. These had been severely damaged and were replaced with
plastic film types. IME, AR have always used adequately sized voice
coils, capable of handling far more power than my KEF b-139 drivers.

> b) But their transient capacity was rated massively higher.

**No, it wasn't.


> c) A great deal of my signal has a P/A of 20 dB. Some few examples have a P/A of 30 dB. Gregorian chant, single instrument, single voice may have the P/A of 10 dB that you reference, and those people may never turn the volume past 9:00 - but not all of us.

**I did say: "Worst case".


>
>>> If a pre-amp has no gain over the original signal source, which is a nominal 2V, the peaks and transients will not be expressed - unless the actual listening level is a minimum of 30 dB below the average.
>>
>> **Well that depends on the sensitivity of the amplifier.
>
> No. That is a fact. If the pre will put out no more than 2V, then it will drive the amp to peak for some percentage of the time. But not past the peak, ever. So, if the P/A is 10 dB, that will be well within the capacity of the amp (and the speaker). A close-run thing at 20 dB, impossible at 30 dB. But, using the example of a transient, and 90 dB efficient speakers - a perfect example of "impossible" and why I picked that particular item.

**Again: It depends on the sensitivity of the amplifier.

>
>
>>> Pre-amps are awful things, just like democracy. Except for the alternatives.
>>
>> **A decent preamp need not be expensive.
>
> That is absolutely true. A decent vintage pre may be had for well under US$200, tube or SS, and may, for approximately the same again or less, be tweaked to as-good-as anything off the shelf today.

**Perhaps.


--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On 5/03/2019 5:11 am, wrote:
> On Monday, March 4, 2019 at 11:32:18 AM UTC-5, Howard Stone wrote:
>> On Monday, 4 March 2019 16:06:40 UTC, Peter Wieck wrote:
>>> On Monday, March 4, 2019 at 9:59:57 AM UTC-5, Howard Stone wrote:
>>>
>>>> Aha, this is an argument for active pre-amps. Passive pre-amps presumably all have no gain.
>>>
>>> Correct. Passive pre-amps are no more than switches.
>>>
>>> Peter Wieck
>>> Melrose Park.
>>
>> So I guess that, given that my project is about exploring Class A
>> with my speakers, that IF I buy a pre-amp, it too needs to be Class A.
>
> The chances you're going to come across a preamp whose active
> 'stages are NOT biased class-A is pretty small. I can think of none
> of the top of my head. (Some earlier IC-based preamp may have used
> op-amps that had a push-pull output stage that may have run class
> AB, but, again, I can think of none).

**I'm not certain about that. Obviously, depending on the load
impedance, many OP based preamps may, in fact, be operating largely as
Class A/B amplifiers. A system popularised back when the 5532/4 was
released, was to force it's operation into Class A mode, by inserting a
current source into the feedback loop. I've never measured it, but I
assume that suggests it is a largely Class B output stage, but by using
the kludge, becomes Class A (load dependent, of course). Maybe I'll
measure it one day....

>
>> And I suppose I find out whether there are any Class A DACs,
>> because I think their analogue end must involve some gain.
>
> Same goes here as well: you're not likely to find a DAC output
> stage that is NOT class-A biased.

**Not that it matters either, I agree.


>
> NOTE: A "DAC" really consists of two fundamental parts: the
> digital-to-analog stage, which, these days, will also
> include the bulk of the anti-imaging filtering, done
> in the digital domain: this part feeds an analog signal
> to the line driver stage, which, again, is very unlikely
> to be anything but class-A.
>

>
> However, in a preamp, where the total power output requirements
> are orders of magnitude less, these considerations become far
> less of a design driver and thus do not force decisions between
> biasing classes that are either engineering or economically
> driven.
>
> Putting it more simply, it's MUCH easier to design, manufacture
> and support class A amplifiers dealing with low(er) signal levels
> and substantially lower powers than it is to to the same where you
> need LOTS of power.

**Quite so.

>
> Don't spend a lot of time worrying about whither the preamp you
> might buy is class A or class AB: because you'd be spending a HUGE
> amount of time searching for the class-AB preamp that you don't want
> to buy.

**Load depending, yes.


--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

On Monday, March 4, 2019 at 4:47:15 PM UTC-5, Trevor Wilson wrote:
> On 5/03/2019 7:11 am, Peter Wieck wrote:
>>
>> Um. Not so much. Back in the day, AR made a huge point on
>> the power-handling capacity of their speakers, and used
>> "magic" such as ferro-fluid and so forth for heat dissipation.

Back in "which" day? Ferrofluid was not commonly used until the
1980's, long after the original AR days, and more like into
the Teledyne/Acoustic Research era and later.

> **Ferro-fluid is used only in HF drivers and some mids and,
> except for NEAR drivers, not in bass drivers

Quite correct. Some tweeters used it, very few midranges used and
not a single bass or wide-band driver of any commercial viability
used it. It's use or non-use is well correlated with the maximum
excursion capability of the driver: it's completely unsuitable
for drivers with large excursions (typically greater than 0.5mm)
because it literally is pumped out of the gap.

> And it is the bass driver that limits
> the maximum thermal limits of a speaker system.

For the most part, the actual data from the field contradicts this.
<Looking at data from many of the clients I have dealt with over
the years, power-related (either thermal or mechanical) speaker
failures by far are dominated by thermal failure of tweeters,
while the remainder are mechanical failures of woofers. For certain,
there are overlaps (e.g., mechanical tweeters failures and thermal
breakdown in woofers), but, based on field data across a very large
number of systems, these latter are very much in the minority.

> Except for some pro drivers, few domestic bass drivers can cope
> with sine wave power levels of more than 50 Watts or so.

It's not so simple a relation as that. For sure, one can say that
if the voice coil is subject to sufficient input such that you're
dissipating the vast majority of the input power as heat in the voice
coil, that's largely true.

But consider the following: all other things being equal, for what
might be the output voltage of the amplifier that would result in
50 watts being produced at, say, lower midrange frequencies would
produce MUCH less power (and thus thermal dissipation) at below
100Hz or say at 2000 Hz, where the impedance is substantially higher.

Especially on the low bass, where the impedance is dominated by the
resonant peak from the woofer, and combined with the fact that the
excursion of the woofer gores as the inverse square of frequency,
the thermal problems a secondary compared to the mechanical limits.

--
+--------------------------------+
| Dick Pierce |
| Professional Audio Development |
| cartchunk.org |
+--------------------------------+