Simple Symetrical Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am simulating something in this way. ;)

Red trace is amp freq response with 10uF input cap. Green trace is servo response. As simmed, the servo with the shown values would work at about 0.2Hz and quickly fall in response in both sides. I used this with my DOZ amps and it worked pretty well, the thing to keep in mind is not to provide voltage gain in the VLF zone where the servo will work.

:2c:

Nice variation of SSA Shaan.
 
Nice variation of SSA Shaan.

Thanks Nico. I hope it'll work for me.

Nice one.
May-be you can chose an OP Amp with low offset, better than the poor TL072 (Just my two cents) ?

Hi Christophe. :wave:

Thanks for the tip.

However, TL072 is INR12(about $0.25) here. Although poor(and noisy when used as buffers), did a nice job as servos with +-1mV offset in my last tests with DOZ. And even-lower-offset opamps seem to cost more than a dollar :sigh:. Of course if one can afford the premium devices, then there is No reason not to use them.
 
I noticed in spice that when I added a cap in the FB path it raised the H10 - 20 noise floor.

I also noticed also in spice when I added a servo it raised the H10 - H20 noise floor.

In both cases lowering the lower -3db frequency reduces this problem into insigfificance but the servo needed a lower -3db frequency than the the FB cap to achieve same noise floor.

The servo spoiled the sound of the amp until the -3db frequency was lowered.

This might suggest that a FB cap is a better solution than a servo but with a servo once the frequency has been lowered it is effectively out of the audio band and sonically invisible whereas the FB cap - whatever the value - is always present throughout the audioband and will influence the character of the sound to some extent.

My take home message:

Avoid both Caps & Servos if possible and, as Christophe says, with SSA it is possible :)
 
Last edited:
can you design circuits which do not require a dc servo? -RNM

As my dear and respected friends already explained, SSA doesn't require DC servo for proper operation. It is integrated in the nature of an input stage that only small thermally caused Vbe voltage differences offset the SSA from zero, usually within +/- 50 mV range but not more than +/- 100 mV.
This offset can be set and compensated to close to zero mV by some steps-measures:
- input pair thermally coupled
- NTC to monitor input bias curent via mid FB DC voltage level
- high precision CCS to set input bias

SSA BIGBT HP I made in June 2012 is working without any DC servo, it is completely DC coupled and it works in recording studio on daily basis since then. :yes:
 
can you design circuits which do not require a dc servo? -RNM

Yes. Would you like some LTP, with maybe little flavor of mirrors? :rofl:

With a servo once the frequency has been lowered it is effectively out of the audio band and sonically invisible whereas the FB cap - whatever the value - is always present throughout the audioband and will influence the character of the sound to some extent.

Agree.

My take home message:

Avoid both Caps & Servos if possible and, as Christophe says, with SSA it is possible :)

Totally agree. However, a servo is being added for the freshly-into-SSA diy-ers for whom I like the great SSA to be one of their first circuits which works with minimal "setting" and is reliable and safe and won't make a hole in their pockets. Every time I log in I see a lot of new people online in this thread, often more than 15. Won't it be great if some of them become interested in SSA and could enjoy it in a friendly and easy way? I can bet no one can leave the SSA once he/she hears its magic, and will inevitably be pushed to learn more about it, and maybe develop their own "suitable" or "complex" versions of SSA. :2c:

A circuit can be build not to use servo or caps, but servo is needed if your previous source sends out DC.

For commercial i would not run without a cap or servo.

If you have a gain of 21x and your previous source outputs +/-50mV => +/-1050mV at the speakers.....

Very true. I have lost a fine speaker this way, and I had only one of them and now they are out of production.

As my dear and respected friends already explained, SSA doesn't require DC servo for proper operation. It is integrated in the nature of an input stage that only small thermally caused Vbe voltage differences offset the SSA from zero, usually within +/- 50 mV range but not more than +/- 100 mV.

Of course.

This offset can be set and compensated to close to zero mV by some steps-measures:
- input pair thermally coupled
- NTC to monitor input bias curent via mid FB DC voltage level
- high precision CCS to set input bias

IMHO, second and third options are a lot more complicated in implementation than the first one which requires only a drop of heatsink paste and a zip tie and also limit the total number of parts to a minimum, preferable to guys who like the KIS approach. However, the three steps, when taken together, will definitely make the servo unnecessary.



Choice... the problem is choice. Thomas A. Anderson :D
 
Oh you thought the circuit was new too.....:D:D:D

Yeah, audio amplifier electronics is not rocket science. There's nothing new under the sun, but...

I have had simple solution for this certain SSA issue. I couldn't believe that it was an invention, because I have never seen such solution before. But later on I saw a CFB amp implementing the same thing.

So, I think it is not about which circuits, old or new, but why you choose it.

CFB amps has been around since year 1. But even today everyone is arguing about CFB versus VFB.
 
Disabled Account
Joined 2006
Yeah, audio amplifier electronics is not rocket science. There's nothing new under the sun, but...

I have had simple solution for this certain SSA issue. I couldn't believe that it was an invention, because I have never seen such solution before. But later on I saw a CFB amp implementing the same thing.

So, I think it is not about which circuits, old or new, but why you choose it.

CFB amps has been around since year 1. But even today everyone is arguing about CFB versus VFB.

Do you mean solution for the temp compensation of the input pair or the Dc offset, why not post it if it doesnt change the rest of the design.
 
I never claimed I "designed" a certain circuit, or SSA, or any other amp for that matter. So I don't care who's the person that posts such sarcastic comment like Mr. RNMarsh did. If it was the prime minister of India then he would be treated the same way. People should look at homemodder and learn how not to behave in diyAudio.
 
Disabled Account
Joined 2006
I never claimed I "designed" a certain circuit, or SSA, or any other amp for that matter. So I don't care who's the person that posts such sarcastic comment like Mr. RNMarsh did. If it was the prime minister of India then he would be treated the same way. People should look at homemodder and learn how not to behave in diyAudio.

Youre showing it perfectly, never heard of treat others as youd want them to treat you. :D
 
......So I don't care who's the person that posts such sarcastic comment like Mr. RNMarsh did.......

One day you might feel sorry for having said such things.
You know what a " GURU " is don't you ? In our country we are taught to give the utmost respect to our Guru's.

RN Marsh is one of our Guru's !

Did you ever Google to check who he is ?
Here is one.
Marsh Sound Design ~ Experience the Sound with Marsh Sound Design Products !!!

Grow up young man .It's never wrong to make a mistake provided you learn from it ........and make up for it !

Cheers.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.