SE distortion - Page 5 - diyAudio
Go Back   Home > Forums > Member Areas > The Lounge

The Lounge A place to talk about almost anything but politics and religion.

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 8th May 2012, 03:56 PM   #41
diyAudio Member
 
pauldune's Avatar
 
Join Date: Oct 2009
Location: Netherlands
Hi Johann,
No problem, I wasnt overly offended, it is ok. I read some of your posts, and I can see what you mean. It looks like they don't like you or something, and some of the replies were not really nice from them.
But, I have to be honest, some of your replies were also not so nice.
I see you have a strong opinion about things, and some people don't like that.
If I were you, I would try not to respond everytime someone doesnt agree with you.
OK enough with that!

You think the 1 ohm series resistor with the primary is to small? I thought to minimise the effect of the measuring resistor on the rest of the setup.
But I think you are right; if I use a larger resistor I get a larger drop, which is easier to measure.
I would suspect that with a small voltage-swing the hysteresis effect would be clearer. So if I use a swing of say 10v PP, at a 5Kohm primary transformer 1 ohm is definitely to small, it would result in ~2mV pp swing in the 1 ohm resistor. That is maximum sensitivity of my scope.
Hmm, I think 100 Ohm would be better. And 8 Ohm at the secondary of course.
Dont know when I can try it, but I will put results here.

Thanks, Paul
  Reply With Quote
Old 8th May 2012, 04:33 PM   #42
dady is offline dady  Balearic Islands
diyAudio Member
 
dady's Avatar
 
Join Date: Oct 2002
Location: Punta Prima, Menorca, Spain
Send a message via MSN to dady Send a message via Skype™ to dady
Hi again, the distortion in SE is the same of the push pull and the same of all the topologies, in PP if you have a balanced output with well charge driving you lost the pairs or even harmonics. But always you need a matched pair and a well made transformer.
One important thing of SE that always work in A class. With PP you can make more easy without passing all the current in all the cycle of the tube. The efficiency is better but you still have 3th and odd group of harmonics.
The power supply in SE is transcendent for hum but no for intermodulation distortion because the consume is at 100%, how you don't need a big bank or capacitor like in AB or B classes that in the fortissimos need more power you have on charge there in caps. Also in SE The hum is too ostensible because the transformer doesn't cancel. But in PP you have this class of benefices.
I am not an engineer i am an hobbyist one. I work in health on ICU at hospitals. But i will try you suggestion Mr big Johan Stephan Popilin Frind. If is possible to show a diagram for see this hysteresis envelope and maybe you suggest the time base of the oscilloscope for help us the mortal people that hope to learn from the big monsters of the audio investigation
__________________
Esteban Bikic
  Reply With Quote
Old 8th May 2012, 04:47 PM   #43
popilin is offline popilin  Argentina
diyAudio Member
 
popilin's Avatar
 
Join Date: Oct 2010
Location: Traslasierra
Hi Paul
Thank you for your understanding, but if you were systematically persecuted, attacked, and ashamed by someone, you culd understand me better, is true, sometimes react badly, but the experience makes me that, there is a forum member that make the same with me for almost two years!!!
Fortunately there are people like you and Dady and many more, who are friendly and civilized.

With measurements going in the right direction, be careful not to overload the generator.
What else I can tell? the same thing I told Dady, without wishing to offend you, you look like an electronics engineer, and I do not mean quacks, "opinionologyst" and critics by profession, but a good one !!!
I think you are a great hobbyist!!!

BTW, I envy your transformers, 0.05mm, wow!!! mine is NOSS C core, but 0.25mm. But is what we have and can pay. Core only, I must to winding someday

Best regards
Johann
  Reply With Quote
Old 8th May 2012, 05:00 PM   #44
popilin is offline popilin  Argentina
diyAudio Member
 
popilin's Avatar
 
Join Date: Oct 2010
Location: Traslasierra
Hi Dady
I am so slow to write, when I finish and post there are five more.

You never cease to surprise me!!!

You do not need any particular scheme, our friend Paul interpreted my suggestion very well, does what he, and just play with the scope, time base depends on frequency you use, just play, the scope will survive.

Best regards
Popilín
  Reply With Quote
Old 8th May 2012, 07:58 PM   #45
popilin is offline popilin  Argentina
diyAudio Member
 
popilin's Avatar
 
Join Date: Oct 2010
Location: Traslasierra
Quote:
Originally Posted by Michael Koster View Post
Magnetic hysteresis is based on the reversal of direction, not crossing some absolute zero magnetization level.

This is exactly borne out in Maxwells equations, specifically Faraday's law.

You should find magnetic hysteresis based distortion at very low signal levels, in both SE and PP output transformers.

I don't find evidence or theory for an additional source of distortion in PP output transformers due to "magnetic zero crossing". What I do see is a general change in permeability due to the SE magnetic gap and DC bias which changes the shape of the B-H curve overall.
Hi Michael Koster
Do not want you to feel offended by my comments, but in post # 12 I said no magnetization, because the analysis is much more complicated.
Magnetization is given by

B = H + 4(Pi) M

And zero crossing of hysteresis loop means that B=0 at two points in the hysteresis loop curve, B is a vector, and a change in its sign means a reversal in its sense of direction, this cause an impact in magnetic domains of the core IMHO, I called zero crossing because my knowledge is limited, and my English is worse.
But what else I can do? I'm just a TV repairman.

Best regards
Johann
  Reply With Quote
Old 8th May 2012, 09:25 PM   #46
diyAudio Member
 
Michael Koster's Avatar
 
Join Date: Jan 2008
Location: Eureka, CA
Hi Johann,

I meant to say rate/reversal of voltage change rather than magnetization. And english is my first language

So I find that the magnetic hysteresis occurs when the direction of the voltage change reverses, i.e. increasing to decreasing, rather than the polarity change (plus to minus) of the applied voltage.

This means to me that there isn't any additional source of hysteresis or distortion in the case of PP vs. SE. In fact, the typical B-H curve shown in your first plate has a polarity reversal in the middle and there is no bump

It seems to have more of an impact on small signals because the magnetization energy is small compared with signal level in a good transformer.

Your second diagram is thought-provoking but it doesn't correspond with any measurement I've ever seen so I'm suspicious of the context and meaning wrt. audio output transformers. I'll need to study the reference.

Thanks!

Michael
  Reply With Quote
Old 8th May 2012, 09:32 PM   #47
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
The DC field in a SE xfmr causes the highest permeability domains in the material to flip one way permanently, leaving the lower permeability domains to switch back and forth for the signal, plus the air gap to lower the perm. even further. For a given signal voltage, and a similar core/winding design, the same number of domains must flip back and forth for the same voltage swing (voltage proportional to flux swing) whether P-P or SE, so the SE xfmr has to be made much bigger to compensate for the loss of domains. The total hysteresis would be similar for both cases (P-P, SE) if the core length stayed the same, but the SE with it's bigger core will end up with more material switching along the longer magnetic path.

The hysteresis shows up as a non-linear magnetizing current, ie. in the primary inductance, not as direct distortion in the voltage conversion. Lower primary inductance means higher magnetizing current. Again, both SE and P-P could be similar if made to the same primary inductance spec, but the expense and winding difficulties for HF response on the bigger core usually limit the turns for SE, giving lower primary inductance and higher magnetizing current.

The hysteresis current becomes a distortion issue for high output impedance stages since the magnetizing current influences their output voltage. A lower output impedance stage (low Mu triode or local feedback pentode) can minimize the problem quite well. And it is not just the hysteresis component of the magnetizing current that causes distortion, the far bigger "linear" part of the magnetizing current is 90 degrees out of phase with the signal as well. P-P OTs typically have such high primary inductance that both issues are non-existant except for the poorest designed output stages. SE OTs typically have issues.

There is also the argument that SE OTs have more constant primary inductance due to the air gap and this is true. But it is a constant lower inductance, which is not good. Higher inductance simply removes the issue. Also, both types have a small linear (low hysteresis) range near zero AC signal, where the operating domains are not completely flipping, but rather are in a reversible "spring" like mode. This is helpful for low level signal quality, but the permeability is lower in this limited range. The P-P will show a large upswing in inductance as the signal grows above this (up until sat. effects start to bring it back down). While the SE will not, due to the air gap keeping it low. This might be the sonic origin of the zero crossing tales, but a low Z output stage will readily fix this. Nickel alloy pinstriping is another fix.

A parafeed OT operates in the same mode as a P-P OT, with no DC flux in the OT. Using a SE OT in parafeed mode will sound even better without its gap (ie. remove the gap spacer). A properly made P-P OT will then extend the upper bandwidth further by using fewer turns and less core for the same inductance. The SE OT in parafeed mode will also have huge voltage headroom, since it has twice as much flux capability without the DC present, this may well improve the sound some, but would simply be equivalent to using a 3X or 4X big P-P OT per the required Watts.

The usual simplistic talk about avoiding zero crossing in SE OTs is just salespeople talk. Who wouldn't love myths that sell more expensive OTs.
__________________
Without MS, how would we live without all those hackers?

Last edited by smoking-amp; 8th May 2012 at 09:52 PM.
  Reply With Quote
Old 8th May 2012, 10:05 PM   #48
MelB is offline MelB  Canada
diyAudio Member
 
Join Date: Feb 2005
Location: BC Canada
An article on transformers and measurement:
http://www.classicaudio.ru/articles/...ion_p2_a57.pdf
  Reply With Quote
Old 9th May 2012, 02:25 AM   #49
popilin is offline popilin  Argentina
diyAudio Member
 
popilin's Avatar
 
Join Date: Oct 2010
Location: Traslasierra
Quote:
Originally Posted by Michael Koster View Post
Hi Johann,

I meant to say rate/reversal of voltage change rather than magnetization. And english is my first language
Hi Michael
Spanish is my first language and the words confuse me all the time, imagine in English, and thinking so slowly, I find it hard.

Quote:
Originally Posted by Michael Koster View Post
So I find that the magnetic hysteresis occurs when the direction of the voltage change reverses, i.e. increasing to decreasing, rather than the polarity change (plus to minus) of the applied voltage.

This means to me that there isn't any additional source of hysteresis or distortion in the case of PP vs. SE. In fact, the typical B-H curve shown in your first plate has a polarity reversal in the middle and there is no bump
I clarified that I quote the link in post # 17, not to look like boy star, but was very helpful to me, I was took some effort to make them, helped me design my own transformers, but can be useful to other, the formulas that appear in the books are not as accurate.And some are not listed anywhere.
For an Ideal loop where an AC current i(AC) flows with a voltage U(AC), from equation (20), the AC magnetic field is given by

B(AC) = [ U(AC) x 10^8 ] / [ sqr (2) * (Pi) * S * f ]

What you say is true in the saturation zone, but it is unavoidable.
However, both effects can be minimized, if view equation (32), causes the hysteresis loop is moved toward the first quadrant, reducing the area of the curve, then increase linearity, ofcourse we are talking about SE.

Quote:
Originally Posted by Michael Koster View Post
It seems to have more of an impact on small signals because the magnetization energy is small compared with signal level in a good transformer.
I prefer not to speak of magnetization, but that's the idea, you're right, and it is one of the secrets of the SE.

Quote:
Originally Posted by Michael Koster View Post
Your second diagram is thought-provoking but it doesn't correspond with any measurement I've ever seen so I'm suspicious of the context and meaning wrt. audio output transformers. I'll need to study the reference.

Thanks!

Michael
I apologize, as I have not how to draw, put the diagrams found on the web, almost mean nothing.
I have nothing against PP, I'm making a hybrid with BJT at the output.
The deeper truth is that... I hate transformers.

Best regards
Johann

Last edited by popilin; 9th May 2012 at 02:33 AM.
  Reply With Quote
Old 9th May 2012, 01:48 PM   #50
TheGimp is offline TheGimp  United States
diyAudio Member
 
TheGimp's Avatar
 
Join Date: Aug 2009
Location: Johnson City, TN
Here is part 1 of the Crowhurst article.

http://classicaudio.ru/articles/Outp...former_a57.pdf
  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
OPC’s The WIRE – Latest Generation - PCBs for sale (4x BAL-SE, 1x SE-SE, 1x PSU KIT) fualcr Swap Meet 0 16th February 2012 12:39 AM
INCREDIBLE 12" woofer-midrange SE friendly, lowest distortion-2khz...MUST READ!!!! Spiritzly Multi-Way 17 29th August 2010 02:05 AM
A question about distortion in SE vs. PP designs: BHD Tubes / Valves 8 18th November 2005 07:36 AM
Non Linear Distortion testing - Harmonic Distortion JMB Multi-Way 0 20th July 2004 03:54 PM


New To Site? Need Help?

All times are GMT. The time now is 12:17 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2