Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Reasons Of Oscillations In Audio Amplifiers And Best Ways Of Elimination
Reasons Of Oscillations In Audio Amplifiers And Best Ways Of Elimination
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 14th January 2018, 11:33 AM   #261
AndriyOL is offline AndriyOL  Ukraine
diyAudio Member
 
Join Date: Jul 2008
Location: Lviv
Quote:
Originally Posted by mjona View Post
I have been looking at square wave testing. The dc offset in the attached .asc file is +14 mV. The circuit does require an output coil to be stable.
This is SE, no balanced input?
Perhaps we need to connect signal source like this? But cold input isn't a return for hot input. Than there should be return from differential signal source?
Maybe that version distorts if the signal to the cold input is in opposite phase to the hot input because of absence of return?
Attached Files
File Type: asc SSAHP kplus dual siggen sim3.asc (26.9 KB, 7 views)
__________________
If for pristine and pure adoration - it's diyAudio dedication!
  Reply With Quote
Old 14th January 2018, 10:03 PM   #262
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Quote:
Originally Posted by AndriyOL View Post
This is SE, no balanced input?
Perhaps we need to connect signal source like this? But cold input isn't a return for hot input. Than there should be return from differential signal source?
Maybe that version distorts if the signal to the cold input is in opposite phase to the hot input because of absence of return?
It is clear from my simulations that this circuit performs better when the hot and cold input signals are in the same phase. The reasons for the deficiency with opposite phase were given by keantoken in post 182 on page 19 - re impacts on the nfb due to R2 and R4 forming an alternative decoupling route to earth.

This increases the closed loop gain significantly and reduces the feedback factor a situation that does not arise with identical input phase signals on hot and cold.

With signals on hot and cold inputs with the same voltage but opposite in phase the THD is about 2%.

Looking at the circuit from a different viewpoint - in the positive rail half R2, R3, and R17 all share a common supply rail with Q1 and Q12 sharing the same configuration.

In the case of identical phase input signals on the hot and cold terminals an increase in a negative going signal will cause Q1 and Q12 to draw more current.

On the opposite side of the circuit this will cause Q2 and Q13 to draw less current. The cycle reverses when the input signal goes negative.

Importantly the current flow through all these transistors does not cut off in their push pull action.

When the signal inputs to hot and cold are in opposite phase the action becomes partial push-pull with zones where the above transistors will work in opposition to the point of becoming cut off.

This has been illustrated before in some of my earlier posts - you can refresh some of these images in SPICE to see this again.

I have looked at a method to change the circuit so it will operate with balanced inputs - but the simulation says THD is 0.3% which is senseless in view of the performance indicated with the same phase signals on hot and cold inputs. The figure 0.3% raises the question how stable is it?

At some point one has to call a halt in doing simulations - I have to return to my care obligations so I will be taking a break from DIY for a while.

In the meantime good luck with your endeavours.
  Reply With Quote
Old 14th January 2018, 11:59 PM   #263
AndriyOL is offline AndriyOL  Ukraine
diyAudio Member
 
Join Date: Jul 2008
Location: Lviv
Maybe someone on this forum will find the solution to use balanced inputs with decent THD.
In case of SE, this one should perform better.

Many thanks Mjona for your help and support. Take care of your wife.

Best Regards.
Attached Files
File Type: asc SSAHP kplus siggen sim4.asc (25.3 KB, 5 views)
__________________
If for pristine and pure adoration - it's diyAudio dedication!

Last edited by AndriyOL; 15th January 2018 at 12:03 AM.
  Reply With Quote
Old 16th January 2018, 06:10 AM   #264
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Quote:
Originally Posted by AndriyOL View Post
Maybe someone on this forum will find the solution to use balanced inputs with decent THD.
In case of SE, this one should perform better.

Many thanks Mjona for your help and support. Take care of your wife.

Best Regards.
Yes I will be doing that for six weeks - unfortunately your simulation does not have a load at the output - with 8R in place it is unstable.

I have gone back to the original circuit to see why you had problems with the dc offset adjustment knowing also that a lot of the circuit has been assumed to be correct.

The values for R10 and R11 were 2.2k in the version of the circuit running 60 volt rails in association with these. A value of 750R was substituted for 45 volt rails.

That does not help the coarse adjustment of 100R trim pot pairs to accuracy of an Ohm or less. If the resistance of either is set too high there will be a kink in the form of a sine wave.

It is possible this finicky business is responsible for what you were seeing on your 'scope .

In fact the circuit will work whichever way those CCS transistors are oriented on the pcb - I have done some further simulations - now attached.

One of these is a 10 kHz square wave test with the merged section in the nfb line with a step response capacitor in parallel. In this alternative there is no overshoot in the wave before the output coil.

The ripples after the coil represent the selection of capacitors from 10n to 2uF in one shot.

With the original network as used in the 1kHz sine wave test simulation - there is a single overshoot which many would overlook - but overall the amplifier should be stable with reactive loads and it will be convenient not to alter your pcb.

I cannot give you any good news about achieving balanced operation.
I see some place in the original authors commentary mention of bridging stereo pairs to increase power output?

Anyway, if the dc adjustment is correctly set the original configuration has the lower THD result.

I have changed the emitter resistor values to those given by Lazy Cat and set the current to give close to 27 mV across these.
  Reply With Quote
Old 16th January 2018, 10:13 AM   #265
AndriyOL is offline AndriyOL  Ukraine
diyAudio Member
 
Join Date: Jul 2008
Location: Lviv
It's stable with 8R load if to increase Cdom caps on drivers to 150pF. However the most important to know why the schematic require so big cdom to be stable? I think in SE config we can remove small VAS transistors, maybe it will become more stable with lower value compensation.
DC offset I had before was because of different characteristics of NPN PNP transistors causing asymmetry in operation. That's why values on R8 R9 are not equal.
2.2K for R10 R11 was a mistake i think. If to check other designer's schematic, as well as other people similar ssa projects, these current set resistors were different values (1-1.5K for SE) which should provide enough current for ssa bridge to maintain stable 15V potential on zeners.
With these pots I can set the DC offset within 1-5mV range easily.
I also tried to simulate it with different ccs orientation, the best opt thd-wise for SE use like the original orientation.
What do you mean by "I see some place in the original authors commentary mention of bridging stereo pairs to increase power output?" I don't need bridging or increase power output. 3 output pairs should provide plenty of power I guess.
__________________
If for pristine and pure adoration - it's diyAudio dedication!

Last edited by AndriyOL; 16th January 2018 at 10:15 AM.
  Reply With Quote
Old 16th January 2018, 02:54 PM   #266
AndriyOL is offline AndriyOL  Ukraine
diyAudio Member
 
Join Date: Jul 2008
Location: Lviv
I have omitted R39 R40 (2R2) as they are not required. thd is unbelievable.
Attached Images
File Type: jpg fft.jpg (102.9 KB, 26 views)
Attached Files
File Type: asc SSAHP kplus siggen sim4.asc (25.2 KB, 2 views)
__________________
If for pristine and pure adoration - it's diyAudio dedication!

Last edited by AndriyOL; 16th January 2018 at 03:02 PM.
  Reply With Quote
Old 17th January 2018, 02:28 AM   #267
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Quote:
Originally Posted by AndriyOL View Post
It's stable with 8R load if to increase Cdom caps on drivers to 150pF. However the most important to know why the schematic require so big cdom to be stable? I think in SE config we can remove small VAS transistors, maybe it will become more stable with lower value compensation.
DC offset I had before was because of different characteristics of NPN PNP transistors causing asymmetry in operation. That's why values on R8 R9 are not equal.
2.2K for R10 R11 was a mistake i think. If to check other designer's schematic, as well as other people similar ssa projects, these current set resistors were different values (1-1.5K for SE) which should provide enough current for ssa bridge to maintain stable 15V potential on zeners.
With these pots I can set the DC offset within 1-5mV range easily.
I also tried to simulate it with different ccs orientation, the best opt thd-wise for SE use like the original orientation.
What do you mean by "I see some place in the original authors commentary mention of bridging stereo pairs to increase power output?"
This simulation is of the earlier circuit which does not purport to be balanced so the last question is not relevant.

More to the point there is a problem with the output stage current consumption in this.

Some small changes in this circuit can have a disproportionate affect on performance results.

For instance a small change could upset the dc offset, or increase the THD, or reduce the circuit stability.

Anyway I have done a pair of simulations one of THD and one of square wave performance.

By this means I was able to cross check the results on two fronts by making the same change in each simulation - taking things one step at a time.

I have checked the square wave simulation with a variety of capacitors in parallel with 8R with varying signal input levels monitoring the result at node n20 - I have seen circuit simulations where a malfunction occurs at a low level input signal level.

I don't have the time to fine tune at length - the dc offset is 22 mV

The file is attached for reference purposes.
Attached Files
File Type: asc SSAHP LCshortsquare version.asc (25.6 KB, 0 views)
  Reply With Quote
Old 17th January 2018, 02:43 AM   #268
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Default LC Sinewave simulation

This shows fft THD with various capacitor parallel loads
Attached Images
File Type: png LCShortsine fft THD.PNG (216.6 KB, 20 views)
  Reply With Quote

Reply


Reasons Of Oscillations In Audio Amplifiers And Best Ways Of EliminationHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
bandpass port resonance elimination scott68 Subwoofers 41 15th July 2017 08:09 PM
crossover elimination hewo Multi-Way 17 1st August 2015 06:41 PM
Looking for good ways to learn more about audio. robotinc Construction Tips 3 14th December 2012 06:26 PM
2 ways or 3 ways horn BMS coax or Raidian 950..etc..a newbie with too much questions hahayan Multi-Way 35 27th December 2011 12:13 PM
pro amplifiers vs. home audio amplifiers bonsai171 Solid State 26 25th February 2004 09:24 PM


New To Site? Need Help?

All times are GMT. The time now is 12:08 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki