DIYHiFi.org

For the sake of audio
It is currently Thu Nov 23, 2017 4:31 pm

All times are UTC




Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 28 posts ]  Go to page Previous  1, 2
Author Message
 Post subject:
PostPosted: Wed Jul 19, 2006 6:52 am 
Offline
Cow
User avatar

Joined: Tue Dec 14, 2004 9:32 pm
Posts: 446
Location: Swiss Mountains
carlosfm wrote:
High output impedance? :?
Greg :?:

Yes, it's a current source...before closing the feedback loop. So, high output z.

But since the transconductance must be quite high (I estimate way more than 1000A out per 1V in, with an 8 Ohm load this is an OLG of almost 80dB), after closing the feedback loop, output z is comparably as low as with a common drain output stage.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jul 19, 2006 7:27 am 
Offline
Suspended
User avatar

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

Damping Factor on the GB150D is > 500 w.r.t 8 ohms.

Now if we are to design an amp for DIY application, we don't want to HAVE TO run it in Class A if it's meant to be Class AB and we don't want to specify it as requiring a fully regulated power supply, that would limit it's userfriendliness - but we could take half measures like regulating the early stages, or even run them from a seperate supply of a few volts higher than the outputs to drive the output stage higher to maximize the power.

All of this adds considerable complexity and often also thump causing time constants, so the amp then requires dethump relays,etc... complexity!

Best to accept the current modulation of Class AB output, and accept there WILL be harmonics developed on the impedance of a typical C smoothed power supply. DIYers can always add a bit more C according to their build budget? Double the amount will reduce impedance 6dB and so too the artefacts... can't keep doing that.

So to REALLY make a leap forward in reducing this distortion we have to look to feedback theory to help us, and see if we can identify any fundamental topological improvement possible.

Attached is the equation for a (supply borne) perturbation to a 3 stage amplifier.


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: Thu Jul 20, 2006 7:24 am 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
This is a revelation. :finga:

It tells us that if we can increase A1 significantly then we can improve the PSRR likewise.

Consider that A1A2A3,the total gain of the amplifier, is fixed and the typical amplifier has A2 (the Vas) as the high gain audio bandwidth stage.

Lets apportion some typical gain figures to a conventional amp -
A1 =26dB, A2=74dB and A3 = 0dB. then V=GVin +GVpsrr/20

If we relocate the gain mostly to the first stage -
A1=74dB, A2=0dB and A3=26dB the Vout =GVin +GVpsrr/5000

This amp would have 48dB better PSRR, simply by repositioning the gain within the loop so it has more effect at the first flip from ground ref to supply ref. This is 250 times better. :rock:

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  
 
PostPosted: Thu Jul 20, 2006 11:39 am 
Offline
Sheep
User avatar

Joined: Sun Dec 19, 2004 2:56 pm
Posts: 63
Location: The Netherlands
Hi Greg,
I understand your explanantion and from that your design choice(s) for your topology.

However I wonder what the consquences (disadvantges) are of increasing the GAIN of the input stage and the output stage? Can you also provide options how to reduce them?

I guess I am the slowest amp builderof your kits, as I am still waiting for my R-core transformers. They must arrive within 2-weeks from now, so that is why you haven't heard my impressions.

Henk


Top
 Profile  
 
PostPosted: Thu Jul 20, 2006 9:21 pm 
Offline
Sheep

Joined: Mon Dec 27, 2004 9:28 pm
Posts: 59
Location: Sydney
Calimero wrote:
I guess I am the slowest amp builderof your kits, as I am still waiting for my R-core transformers. They must arrive within 2-weeks from now, so that is why you haven't heard my impressions.


No, I am. :drinkers:

_________________
Greg Erskine


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jul 20, 2006 10:02 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
Hi Henk and Greg,

Well you have both been very slow - perhaps methodical, as you both ordered them in early January. :rolleyes: However, some have not 'reported in' at all. I can only guess they have built the amp and are overwhelmed with the results?

Of course you both have an advantage here, in that you have the full schematic for reference. :finga: Something I just cannot provide here.



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: Fri Jul 21, 2006 11:34 pm 
Offline
Sheep

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

Would it be incorrect to interpret your design in this way.

Given the equations:

Lets apportion some typical gain figures to a conventional amp -
A1 =26dB, A2=74dB and A3 = 0dB. then V=GVin +GVpsrr/20

If we relocate the gain mostly to the first stage -
A1=74dB, A2=0dB and A3=26dB the Vout =GVin +GVpsrr/5000

The SKA benefits from moving most of the gain to lower noise devices. A secondary benefit is that one complete gain stage is eliminted which reduces overall complexity in implementation. That said the biggest benefit comes from lower noise.

I am just trying to intuit the benefit looking at the math.


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jul 22, 2006 5:42 am 
Offline
Suspended
User avatar

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

Sorry I glossed over your question -

Quote:
I wonder what the consquences (disadvantges) are of increasing the GAIN of the input stage and the output stage?


I suppose high voltage gain in the first stage has implications for input capacitance with miller multiplication, but it seems to be a non-issue. Output stage gain means drain output so immediately there is a concern of high output Z, but with measured DF of > 500 over the audio OL BW, it's another non-issue.

Hi Midrangeman,

Yes I guess it seems intuitive that it's a noise thing, especially when the signal is being amplified so much in the first stage before it's referenced to the supplies at ~ 2V.
However, no stages are eliminated as the 0dB voltage gain stage is a very necessary buffer between two gain stages and which provides the bootstrap function. All stages are simple stages though. :finga:

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: Sun Jul 23, 2006 10:10 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
To prove this concept, I developed a new topology to achieve very high gain with audio bandwidth in the first stage. Symmetrical is nice so I used a complementary differential front end with current source tails. The buffer stage could then be a complementary EF with a shared emitter resistor, effectively a CCS. The low impedance EF output could the be used to bootstrap the load R's of the first stage providing very very high first stage gain. The Class A emitter followers could then drive a complementary common source MOSFET output stage for low output losses (swinging to within 2V of supply rails) and further gain. The high Vgs of vertical MOSFETs became a non-issue ( even a +) so these rugged devices became a clear performance/$ favourite. :finga:

Would this topology realise the PSRR desired? The first stage tail current sources were idealised to determine the potential of the new topology, and the sim -


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: Thu Jul 27, 2006 10:34 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
So, comparing the two simple symmetrical topologies - low gain dual diff'l input stage, followed by high gain Vas followed by unity gain EF or SF output, versus very high gain dual diff'l input stage followed by complementary EF buffer followed by CS output stage, we have comparative topological PSRR of -

freq 10Hz 100Hz 1KHz 10KHz 20KHz
standard topo 86 86 86 80 74
SKA topology 100 110 110 110 110

- assuming 30dB closed loop gain. Since harmonics congregate at the higher frequencies, the considerable improvement at 10KHz+ is a big plus, allowing cleaner HF without blurring by a miasma of PS artefacts.
30 dB is 30 times better!

However this is comparing two ideal simple topologies. Both of these require real current sources to power the first stage and these WILL waste some of the topological PSRR advantage of the SKA, due to their inherent non-idealness.

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: Sat Aug 19, 2006 10:23 pm 
Offline
Suspended
User avatar

Joined: Mon Jan 16, 2006 9:06 pm
Posts: 214
Location: australia
So the hunt was on for a biassing arrangement using input stage constant current sources that -

1. Gave excellent stability of operating point.
2. Retained as much of the high PSRR of the topology.
3. Allowed for wide DIY supply voltage operating range for versatility.
4. Did not introduce input stage common mode distortion.
5. Provided temp tracking current reduction to Tcomp the MOSFETs.
6. Was simple and cost effective.

A variety of possibilities was considered -

1. Resistive tails (poor CCS but supposed to sound nice).
2. BJT current sources
3. FET current sources
4. Current diodes

....

_________________
"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  
 
PostPosted: Thu Aug 24, 2006 6:48 pm 
Offline
Sheep

Joined: Fri Feb 25, 2005 2:26 am
Posts: 53
Location: il
henk:

hee, hee ...
i can guarantee that i'm slower than you in building my ska. i haven't even started yet. my kit sits unopened collecting dust. :-(
my wife and my day job have other plans for my time these days.

someday this year, i hope ...

BTW, where are you finding an r-core transformer suitable for (this) power amp?

mlloyd1

Calimero wrote:
Hi Greg,
...
I guess I am the slowest amp builderof your kits, as I am still waiting for my R-core transformers. They must arrive within 2-weeks from now, so that is why you haven't heard my impressions.

Henk


Top
 Profile  
 
 Post subject:
PostPosted: Fri Aug 25, 2006 8:25 am 
Offline
Suspended
User avatar

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

That's really slack. [-X .

I'm sure your wife would jump at the opportunity to help form all the wires to populate the boards, by your side. :squabble1: Together you could make sweet music :music:

Oh gawd...

_________________
"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 Previous  1, 2

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