JG´s Nobrainer and Nobrainer Discrete

[FdW]

The 180 ohm resistors give about 3mA that times 2 gives your 7 (give or take one) mA's. That's all functioning. Maybe it is a startup problem? the Vce of the low voltage parts could start conducting before the high voltage parts (Vce breakdown). Try the same with all high voltage parts and see. You can leave the resistor in place it will not be of a great influence. Also try to get the resistor down and see what the lowest 'working' value is. I can do some test myself, but not before the weekend

 

[FdW]

Now this is cool! Low rising voltage (green) different resistors, all currents are equal. Now you may ask what is cool? I show you (next post )

 

[FdW]

You see way cool Fast rising voltage (green) only purple is stable (with the 1k resistor).

 

[FdW]

And the sim-file. Never seen this (and I did use a few of these )

 

[Joachim Gerhard]

Very high impedance circuits can have problems.
The extra resistor looks like emitter degeneration to me.
Have not seen that before but it seems to lower " distortion ".
So Ricardo found something usefull by " playing around ".

 

[FdW]

And here is my final result.

 

[RCruz]

Ok..... thank you so much.

Now back to the breadboard

 

[RCruz]

according to this sim, your new ccs version seems much stiffer than the "usual" two NPN arrangement (that I built just for Learning purposes)

The "usual" one works flawlessly but it's current highly dépends on Vcc.... going from 35v to 70v almost doubles ccs current.

According to the sim, FdW ccs does not suffer from this "illness". It stays put whatever Vcc I use

 

[RCruz]

Green line is the "usual" ccs..... purple line is the improved FdW design.

PS: I modified the resistor values because I am searching for 7mA ccs (the same I get with the actual 10k resistor in my stock build.)

 

[FdW]

... FdW ccs does not suffer from this "illness". It stays put whatever Vcc I use

That's what I like calling it then 'FdW ccs' like I invented it. Actually I did do a lot of Google-ing to see if there is any pre-existing art and I could find nothing.

My question here is; 'Is there anyone here (or somewhere else ) that knows of any prior existence if this CCS?', let me know!

 

[RCruz]

Good news..... just finished breadboard example and it works perfectly with 35v psu producing 7mA (mouche)

Now I will try it with 70v

 

[RCruz]

It goes without saying.... of course it works with 70v.

Now I will build the two ccs to implement in the amp

 

[Joachim Gerhard]

Very nice. Good luck and i hope it sounds better.

 

[RCruz]

Well, I will now record some measurements on the stock build and post them here.

Will do the same after each mod.

First mod is the ccs, second one will be reducing miller cap and finally the bias circuit.

Lots of work, loads of fun.... in the end I will have collected a lot of knowledge and hopefully get a fine sounding power amp.

 

[Joachim Gerhard]

Sure, learning by doing. I work the same way and when it works it is very satisfactory.

 

[RCruz]

My measurements done with 100mv input and a "real" 8ohm speaker load. (Do not know if using a resistive load would change things significantly)

From 200Hz to 5KHz things look quite good

 

[RCruz]

At 5Hz 10Hz and 20Hz I notice Low frequency loss and phase shift, no amplitude distortion but out of phase.

 

[RCruz]

At 10khz and 15khz gain is still max but we start to have high frequency loss and phase shift , without amplitude distortion but out of phase.

 

[RCruz]

At 20kHz gain starts to lower from top 20dB (19dB at 20k) and the trend is accentuated (high frequency loss and phase shift , without amplitude distortion but out of phase)

 

[RCruz]

At very high frequencies we notice high frequency loss and low frequency phase shift (no amplitude distortion but out of phase).