JG´s Nobrainer and Nobrainer Discrete

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is a classic paper about crossover distortion in a Class B amplifier :

http://hifisonix.com/wordpress/wp-c...oss-Over-Distortion-in-Class-B-Amplifiers.pdf

It is for the simple case of a parallel symmetric emitter follower.
The output stage of the Nobrainer is different. It is a CFB pair or Szikley.
This variety is best covered in Douglas Self´s book " Audio Power Amplifier Design ".
I recommend to buy the new 6th edition.

The original article is from a HP-Journal 1971-02 that can be downloaded here Index of /hpjournal/pdfs/IssuePDFs (you can find a table of content for the site here Index of /hpjournal/pdfs ).
 
You can alway add a coil external if you want and listen what sounds better.
Recently i had an oscillation problem with a VSSA. I really did not like my low inductance ribbon loudspeaker cable and produced an unpleasant high pitched tone. I took a 10 Ohm wire wound resistor and wound 12 windings of stiff solid core cable around it.
I dipped it in wax from a melted candle. That parallel combination of 10 Ohm plus coil solved the problem.
 
Member
Joined 2006
Paid Member
Thank you :)

I have been rearranging the layout (once you get the parts you can start figuring out how they fit together....)

Already made some breadboard tests regarding the opamp psu lines... cool.

Can I use a film cap in the input instead of the bipolar EL arrangement ? .... what would the minimum value required ?
 
What opamp are you using ?
If you use a Fet opamp a cap at the input is not necessarily required.
When you set the input impedance to 10kOhm a 4,7uF should be big enough.
You can alternatively rise the input impedance to 47kOhm and take a 1uF.
In recent years the input impedance of Power Amps has come down from 47kOhm to 10kOhm. So the old 47kOhm is the new 10kOhm. Theoretical 10kOhm is more robust to pick up noise.
 
Member
Joined 2006
Paid Member
I am going to use LME49710 opamp, so maybe I need the input cap.... strange that my Meridian 101 preamp uses a LME49720 in the input and does not need an input cap.... Anyway, 4u7 seems a reasonable value for the input cap.... I am following your schematic and am planning to use 10kohms for R11 (I believe that will determine input imp...)

Why did you choose two 100u caps resulting in a bipolar 50u ?
 
I am going to use LME49710 opamp, so maybe I need the input cap.... strange that my Meridian 101 preamp uses a LME49720 in the input and does not need an input cap.... Anyway, 4u7 seems a reasonable value for the input cap.... I am following your schematic and am planning to use 10kohms for R11 (I believe that will determine input imp...)

Why did you choose two 100u caps resulting in a bipolar 50u ?

I'm using these lovely lovely lovely ... 49710's all over the place and never need input caps (even in my phono) (please go for the metal can version :)).
It Ain't Half Hot Mum - My Lovely Boy 2-3 - YouTube
 
Well, that Opamp does have input bias cancelation, the NE5534 does not.
When you have a cap in the chain before the power amp you do not need a cap anyway.
A cap is praxis though in commercial amps for protection.
When you use elcaps it is better to use a bigger value or there is risk of rising distortion in the bass. The more voltage develops over the cap the more distortion, see Douglas Self.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.