Class AB Soft Non-Switching - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 15th March 2010, 03:41 PM   #1
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Default Class AB Soft Non-Switching

hello,

attached you will find a .zip file containing three amplifiers, only differing in the vertical MOSFET output pair biasing arrangement :

- Class AB,
- Class AB Non-Switching,
- Class AB Soft Non-Switching

The three files are Tina 7 Texas Instruments schematics ready for simulation.

Are you equipped with Tina 7 Texas Instruments ?
Have you read the subject till the end ?
May I suggest that you post here, only if you satisfy the two conditions ?

1.
The Class AB arrangement is a fixed voltage Vbe multiplier, decoupled by a big capacitor in parallel, as usual.

2.
The Class AB Non-Switching arrangement samples the 0,47 ohm source resistor voltage drop of both MOSFETs, individually, and compares the voltage drop with a reference voltage. When the 0,47 ohm resistor voltage drop is less than the reference voltage, the non-switching arrangement sends a correction current into the base of the Vbe multiplier that's associated to the MOSFET. Both MOSFETs thus never go off. However cumbersome or complicated it sounds, it works on paper (on simulation).

3.
The Class AB Soft Non-Switching arangement bias servo-loop is now getting a variable loop gain. The loop gain is minimized when the instantaneous 0,47 ohm current is minimal. The trick is in the differential amplifier operating current. It is now variable so the dynamic Rbe varies accordingly (log relation), and come in series with the 180 ohm emitter resistors. The 180 ohm resistors define the maximal loop gain, when the dynamic Rbe become small in front of 180 ohm.
The waveforms look perfect with a 1KHz 500 mV signal at the input. There are no hard corners nor inflexions anymore in the MOSFETs source currents. All the edges are nicely softened. When viewed from a certain distance, this may be confused with smooth genuine Class-A operation.

***

However, inputting a 500mV 10 KHz sinus signal enables us to see that :

- The Class AB arrangement delivers the cleanest MOSFETs source currents in the changeover zone, but the current drops to zero, and there are hard corners in the changeover zone, meaning a strong high-frequency harmonic content, internally. That's not new. This is known from long, and is the specificity (and flip side) of Class AB operation.

- The Class AB Non-Switching arrangement manages to keep both MOSFETs in conduction, but is prone to HF micro-oscillations bursts (1 Mhz, 2 ÁS duration, 25 millivolts amplitude), very localized inside the changeover zone. Apart from the HF bursts, there are hard corners in the changeover zone, meaning a strong high-frequency harmonic content, internally.

- The Class AB Soft Non-Switching arrangement manages to keep both MOSFETs in conduction, and is delivering smooth rounded current edges, without harsh harmonic content. When looking closely, one can see that the Soft Non-Switching arrangement is perfectly working on the leading edges, when the MOSFETs currents go from low quiecent current to high load current. However, both MOSFETs do experience a small 5 Ás duration, 10 millivolts ampitude glitch at each trailing edge, when the MOSFETs currents need to go from high load current to low quiecent current. The wavevorm that I am publising is the extreme version of the glitch, when the output delivers a 80 Volt peak-peak 10 KHz sinus, and causing a very brief zero current.

I have selected this worst-case waveform ON PURPOSE in order to ask the diyAudio communauty "hello, can someboby help me ironing-out this glitch ?".

Any suggestion welcome,
Steph
  Reply With Quote
Old 15th March 2010, 05:26 PM   #2
diyAudio Member
 
Join Date: Nov 2007
Location: Dallas
I don't fully understand the circuit yet, but looks like tail current of one
of the driver transistors? Are any drive transistors allowed to shut off?
  Reply With Quote
Old 15th March 2010, 08:17 PM   #3
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
Quote:
Originally Posted by kenpeter View Post
Looks like tail current of one of the driver transistors ?
Both curves are the voltage drop on the 0,47 ohm resistor (source resistors). Have you conducted some simulations yet ? Are you now able to reproduce the curves ?
Quote:
Originally Posted by kenpeter View Post
Are any drive transistors allowed to shut off ?
Interesting question. If one of the drivers transistors shuts off, maybe one of the Vbe multiplier endures a zero current, with the Vbe multiplier voltage vanishing. That's a direction to be investigated, indeed ! However, do you think this is occuring with a 10 KHz sine ?
  Reply With Quote
Old 15th March 2010, 10:26 PM   #4
The one and only
 
Nelson Pass's Avatar
 
Join Date: Mar 2001
Since I don't happen to be using Tina, could you provide
all three schematics as Jpegs?

  Reply With Quote
Old 15th March 2010, 10:55 PM   #5
wahab is offline wahab  Algeria
diyAudio Member
 
Join Date: Nov 2009
Location: algeria/france
Quote:
Originally Posted by Nelson Pass View Post
Since I don't happen to be using Tina, could you provide
all three schematics as Jpegs?

Didnt you try to open his zip files?...One never know....

Last edited by wahab; 15th March 2010 at 10:56 PM. Reason: typo
  Reply With Quote
Old 16th March 2010, 08:54 AM   #6
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Quote:
Originally Posted by kenpeter View Post
I don't fully understand the circuit yet, but looks like tail current of one
of the driver transistors? Are any drive transistors allowed to shut off?
Good point. I was looking at it from another angle. The soft-non-switching circuit has to handle a very non-linear high-dynamic range across that 0.47ohms. If you look at the circuit, it's ingenious but involved. The feedback loop back to the Vbe multiplier has non-linearity and phase shift, and inho that is the cause for the small damped oscillations when the output device currents are going down to zero. That loop not only includes the regulation circuit but also the Vbe multiplier, drivers and output devices with wildly differing dynamic properties.
It's a matter of getting that loop stable and it is solved.

In a more philosophical sense, in my experience, trying to control the bias current with a closed loop almost always brings up two issues: loop stability, and/or bias modulation by the signal due to the high dynamic range of the Re voltage.

jd
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W. S. Maugham
Check out Linear Audio!

Last edited by jan.didden; 16th March 2010 at 08:58 AM. Reason: spelling
  Reply With Quote
Old 19th March 2010, 06:53 PM   #7
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
hello dear friends,

thanks for your interest in this novel amplification sub-class. By the way, is this really a "novel" amplification sub-class ?

It may be ...
I know there is a lot of litterature about servo-controlling the minimum bias current in the output devices of power amplifiers, and I know some practical circuits materializing this.
However, I could not find litterature about doing this in a well controlled, smooth and soft approach, minimizing high-frequency generation within the circuit. And I never came across a practical circuit materializing this smooth and soft approach.

Anyway, being the inventor or not of the Class AB Soft Non-Switching amplification sub-class, I'm now offering you an improved schematic. This new release manages to get both MOSFETs delivering smooth current waveforms. This time there are no glitches, no hard corners, and no high-frequency contents anymore. I managed to iron-out everything by adding a few simple components from my previous version delivered in this thread. The positive current now blends admirably smooth with the negative current during the crossover.

Truly amazing ... on simulation.
Class AB Soft Non-Switching now exists ... on simulation.

Have fun playing with the attached .zip containing the Tina 7 T.I. circuit ready for simulation.

Any candidate for evaluating the (novel ?) Class AB Soft Non-Switching schematic in the lab ?
  Reply With Quote
Old 19th March 2010, 09:21 PM   #8
diyAudio Member
 
Join Date: Nov 2007
Location: Dallas
I too been scratchin to figure, "what is this amplifier class?"
Are we perhaps working different paths to the same end result?
My curves look just like yours, but derived by entirely different
method and circuit. (See: Visch or not to Visch thread)...

Compare to 3rd Screenshot of Post#7 above.
Attached Images
File Type: gif Currents.gif (11.3 KB, 163 views)
  Reply With Quote
Old 19th March 2010, 09:51 PM   #9
diyAudio Member
 
steph_tsf's Avatar
 
Join Date: Mar 2008
What about the + and -15V auxiliary power supplies that are needed, centered on the output of the amplifier ? Weird stuff ... horribly expensive and cumbersome ? Maybe not ...

I remember some time ago Jan Didden's PAX amplifier in Elektor. PAX needs two 15V power supplies centered on the amplifier's output for powering a AD844 IC. PAX elegant circuit relies on two bootstrap capacitors feeding two 15 V zener voltage stabilizers.

I have used the same layout. However, conducting some extreme simulations (like 10 Hz audio signal near clipping), I found that some non-sinus ripple was developing using a single zener layout.

A proper tradeoff needs to be found : fast recharging speed (low value charging resistances) against low ripple amplitude (high value reservoir caps). Speed is indeed needed for enabling a proper 15V generation even on a single low duration burst near clipping. Percussive music is a sequence of short bursts.

I think it is tolerable to get a small slow-moving sinus modulation on the 15V supplies in some extreme situations. I think it is less tolerable to get small (maybe smaller) non-sinus (charging waves) on the 15V supplies.

In the proposed bootstrap scheme, I have thus raised the voltage up to 18V using a 1st zener, filtered to 15V by a 2nd stage using a 15V 2nd zener. Depending on personal preference, one can can opt for the 2nd stage, or one can drop it.

Anyway, provided one can reasonably tailor the speed/smoothness compromize in extreme situations, the bootstrap scheme provides an elegant solution to the requirement of + and -15V supplies centered on the output.

One may argue that it draws current, non sinuso´dal current maybe, from the output of the amplifier. Well, with bootstrap resistors equal to 1K, the bootstrap current is confined to small values. I won't say this is neglectible, but I will say it is reasonable.

Are audiophile people rejecting this elegant bootstrap scheme ?
Are there other solutions, really ?

In the attached files, there is a Tina 7 T.I. schematic ready for simulation.

Cheers,
Steph
Attached Images
File Type: jpg Bootstrapped 15V Aux Supply.jpg (216.5 KB, 263 views)
File Type: jpg Bootstrapped 15V Aux Supply (working).jpg (147.3 KB, 220 views)
Attached Files
File Type: zip Bootstrapped 15V supply.zip (2.3 KB, 39 views)
  Reply With Quote
Old 19th March 2010, 10:22 PM   #10
diyAudio Member
 
Join Date: Nov 2007
Location: Dallas
I still don't understand the theory of operation of your circuit.
By what analog computation does it generate these curves?
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Soft start for class AB VMat Solid State 14 13th June 2011 10:04 PM
Class AB, Class AB Non-Switching, Class AB Soft Non-Switching steph_tsf Solid State 4 15th March 2010 04:39 PM
New class A biasing (with non-switching class AB overflow) Steven Solid State 34 24th March 2007 07:40 PM
soft switching mflorin Power Supplies 8 30th June 2005 03:01 PM
A good Brazilian Amplifier, class A switching to AB in hi power destroyer X Solid State 65 24th March 2005 09:12 PM


New To Site? Need Help?

All times are GMT. The time now is 09:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ę1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2