DIYHiFi.org

For the sake of audio
It is currently Mon Nov 20, 2017 9:21 pm

All times are UTC




Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 28 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: Mon Jul 17, 2006 8:17 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
As many on this forum and others will know, my approach to solving the 'special' problems of Class AB are different than most. Here is the manifestation of two extremes of philosophy.

On the top is a dual mono GB150D capable of 150W into 4 or 8 ohms. Under is a 100W/channel 8 ohms, 200W/channel 4 ohm, and I expect 400W/channel 2 ohm (shall remain) un-named brand of Class AB amplifier.

Both are dual mono. :rolleyes:

Which sounds best is subjective. Which COSTS considerably more is probably obvious.

Cheers,
Greg


You do not have the required permissions to view the files attached to this post.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jul 17, 2006 8:56 am 
Offline
Dog
User avatar

Joined: Tue Dec 14, 2004 12:26 pm
Posts: 4981
Location: Lisbon, Portugal
And...???

Greg, can you be more specific? :mrgreen:

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jul 17, 2006 9:54 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Hi Carlos,

My belief is that Class AB is the answer - not only to audio nirvana, but also as an environmentally conscious solution to the needs of audiophilia.
Personally I believe, Class A is wasteful of energy which, to me as a physicist, in an overly energy hungry world without solutions, is a non-solution. Class D is super energy efficient yet, for me, loses out on 'involvement' and seems disassociating from the musical experience. As Tom Waits would say " There's something about it that just ain't right" Seems OK for bass though.

The technology's young.

So my challenge was to produce an amplifier that was 'simple' in concept and processing stages, that didn't need excessive ancilliary support systems ( like massively overbuilt power supplies) to function. It should be energy efficient, using the power supply well and idling at low consumption. Above all it had to deliver sonics that screamed YES!
I didn't want to advertise it. Let it do that.

So Carlos, you probably agree with my oration so far, bear with me... but it's late.. tomorrow..

Cheers,
Greg

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jul 17, 2006 8:59 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Class AB amplifiers have three big 'special' issues that get in the way of delivering great sonics -

1. Each output stage half only delivers ~ half the signal.
2. Power supply impedance is non-zero.
3. Power Supply rejection falls dramatically at HF in the typical Vas stage, due to conventional NFB loop stabilisation techniques.

:axe:

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jul 17, 2006 10:40 pm 
Offline
Sheep

Joined: Fri Jun 09, 2006 3:26 am
Posts: 15
Hi Greg,

I'll bite.

So how do your designs mitigate these issues?

We know that you have eliminated the VAS stage in your latest creations. For those of us that are not solid state design gurus, what does this mean?

As for the conservation issues alluded to, these can be followed by others via some very good websites that track energy issues soon to effect us all.

http://abc.net.au/4corners/special_eds/20060710/

http://www.theoildrum.com

Said succinctly, without considered conservation, we are facing a tough future. The down under dystopia movie Mad Max, may not be far from the truth.

Look forward to your thoughts.

Midrangeman


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 2:24 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Hi Midrangeman,

Quote:
So how do your designs mitigate these issues?


Appreciating that all musical waveforms can be broken down into a series of sinewaves of differing amplitude, frequency and phase, consider the simple sine wave as the most fundamental signal.

The first issue is the 'half wave' nature of the 'push-pull' output stage. If this is what's coming out of the first half into a (for the moment) constant load impedance. Then current drawn from the +V supply will be a waveform of the same shape. Likewise for the -V supply and output stage half.

So, while these half-waves combine to hopefully form a pure sinewave into the load, they demand half-wave currents from the power supplies. Pictured.

Bear with me. :roll:


You do not have the required permissions to view the files attached to this post.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 2:39 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
This current waveform demanded from each supply to output a pure sinewave from the amplifier into the load, is no longer a pure single frequency waveform but is a fourier series of even harmonics as shown.

Of these output stage generated harmonics 2HD is -8dB, 4HD -22dB, 6HD -27dB and 8HD -35dB relative to the 1KHz

This output stage current demand waveform will be drawn through the power supply impedance and develop a voltage waveform rich in the even order harmonics ON THE SUPPLIES in response to even a single frequency sine wave (example 1KHz), dependant on supply impedance.


You do not have the required permissions to view the files attached to this post.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 4:14 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
In an ideal world our power supplies would have no impedance and be pure DC voltage sources and many highly regulated power supplies can closely approach this ideal. BUT. In a power amp this comes at considerable expense and complexity.

So we mostly use capacitor smoothed power supplies. This is the first mitigating factor. Impedance of ideal capacitors falls as frequency increases - so our supply impedance will fall at 6dB/octave and will reduce the level and proportion of these generated harmonics modulating the V+ and V- supplies. -

Fund -0dB , 2HD -11dB, 4HD -30dB, 6HD -42dB, 8HD -51dB.....

But then there's supply wiring R, L, and non ideal C ESL,ESR to complicate matters. As a practical example, shown is the FFT of the supply when a GB150D runs 5Wrms at 8ohms and 1KHz sine. The supply has +/-15,000uF. Also clearly visible on the FFT is the harmonic spray from the ripple sawtooth of the upper pic charge-discharge waveform.


You do not have the required permissions to view the files attached to this post.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 6:42 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Now that's what we can expect on our power supplies, let's look at the typical simple conventional symmetrical circuit with a Vas stage and look at it's PSRR to see how well it will ignore those artefacts on the power supply.

Attached is a sim of a complementary differential, complementary Vas and complementary MOSFET source follower output, very common in Hi End audio. Very Barry Manilow.

The transistors are BC556C and BC546C for the Vas (low capacitance types ~ 3pF and yes operating beyond their voltage and power limits here) so the compensation miller C is very minimal.

Look at the high frequency PSRR. ( add 30dB for amp gain).

This 75dB @20KHz and 80dB @ 10KHz will let through a emphasized harmonic structure that would be causing audible smearing of fine spatial and timbral detail, causing a defocussing of the image and blurring at HF.


Heavier compensating C than the 10pF or a lower speed Vas transistor will degrade the PSRR considerably more.


You do not have the required permissions to view the files attached to this post.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 6:48 am 
Offline
Cow
User avatar

Joined: Tue Dec 14, 2004 9:32 pm
Posts: 446
Location: Swiss Mountains
One possibility to mitigate those artifacts would be to build a complete balanced amp.
Of course, parts count doubles, but there are several advantages...

Could you bridge two of your amps and do the same measurements? Of course the positive and negative supply in that case has to feed simultaneously both half's, preferably with equal wire lengths etc...so, it's not just to takte two amps and bridge them, ending up with two power supplies for a bridged version. Most advantages would be eliminated.

Thanks, Tino


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 7:55 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Hi Tino,

Yes to drive 2 amplifiers in antiphase on the same supplies and take a bridged output is a way around the PSRR issue - at more than double complexity. :roll:

So now what we have is -

1. A Class AB output stage that generates harmonics in the power supply current.
2. Power supply impedance that converts these to supply voltage modulation.
3. An amplifier topology that lets them in, at a location that they are amplified greatly to the output, the input to the Vas.

It doesn't sound much good at all. What to do.

There are a number of options.....

For now.

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 9:59 am 
Offline
Dog
User avatar

Joined: Tue Dec 14, 2004 12:26 pm
Posts: 4981
Location: Lisbon, Portugal
Greg, I'm trying to 'visualize' your amp and it doesn't look much different than an op-amp with gain and a discrete complimentary stage inside the feedback loop.
You just made it full discrete.

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 11:12 am 
Offline
Cow
User avatar

Joined: Tue Dec 14, 2004 9:32 pm
Posts: 446
Location: Swiss Mountains
What we have to try to visualize in the post above is not Gregs amp, AFAIK. His "killer" amp has the output stage in common source configuration, i.e. with gain (as the Zen V5 :mrgreen: ). And no conventional VAS.
Voltage gain is acheved through the input differentials and the output stage. In between I believe is just a follower.

But then, my memory may fool me.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 1:31 pm 
Offline
Dog
User avatar

Joined: Tue Dec 14, 2004 12:26 pm
Posts: 4981
Location: Lisbon, Portugal
zinsula wrote:
Voltage gain is acheved through the input differentials and the output stage. In between I believe is just a follower.


High output impedance? :?
Greg :?:

_________________
Carlos Filipe

"Any intelligent fool can make things bigger, more complex, and more violent. It takes a touch of genius - and a lot of courage - to move in the opposite direction." Albert Einstein


Top
 Profile  
 
 Post subject:
PostPosted: Tue Jul 18, 2006 9:11 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
That's correct Tino.

Confronted with this problem of Class AB output generated distortion sitting on the supply impedance and getting into the amplifier and subsequently appearing at the output along with the pure tone that generated the harmonic spray, designers have a few options to stop this from degrading their sound quality.

The Beast pictured above appears to take the approach -

1. Increase the idle bias to a couple of amps so the amplifier runs in Class A to 30W.

Benefit: No AB artefacts until 60W peak is reached.
Downside: 200W dissipation at idle (/ch). 1/2 kW idle power total.

2. Regulated supplies even separate ones for output and other stages.

Benefit: Great reduction in PS impedance so reduced AB artefact appearance on supplies.
Downside: Massive increase in complexity (and cost) and need to overbuild power supply as reg +/-Vout must be below the worst case sag of mains voltage, heaviest anticipated load, and reg dropout. So transformers have to go up in size and PS C's go up in V rating and value (Vripple is proportional to V). The trafos here are 2 x 1200W+EI.

3. Change to amplifier topology - widebanding with fast output BJTs and minimal Ccomp.

Result achieved. A nice sounding amp for the well heeled.:rock:
Especially those with a forklift and power station shares.

Cheers,
Greg

_________________
"Clearer and nearer still," cried the Rat joyously,
"Now you must surely hear it! Ah - at last - I see you do" ............The wind in the Willows


Top
 Profile  
 
Display posts from previous:  Sort by  
Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 28 posts ]  Go to page 1, 2  Next

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group