Why source resistor?i understand it

[Mordor]

In the AB complementary amps generally are used very small value resistors ( 0.1 ohm ) between source and the virtual mass ( where is connected the speaker ). WHy? this resistance increse the value of Rout, decreasing the current... Please help me to understand

[MikeB]

Hi !

The mainreason is to improve thermal stability, without these
resistors the thermal runaway can get big. And if paralleling
outputdevices, these are a must, its the only way to ensure
proper currentsharing. Otherwise the transistor with the
lowest Vbe takes all the work, get's hotter than the others, and
the Vbe get even lower... Bang
And these resistors can add some stability to capacitive loads.

Michael

[Mordor]

what are the value of the resistor ? can i use 0,1 ohm 5 w ?

[markp]

Originally posted by Mordor
what are the value of the resistor ? can i use 0,1 ohm 5 w ?

Yes, .1ohm 5w is in the range. For more stability you can go up to as much as .5ohm but you increase you source resistance and lower damping ratio. If you use lots of output pairs this is fine as they are all in parallel.

[MikeB]

Originally posted by Mordor
what are the value of the resistor ? can i use 0,1 ohm 5 w ?

This depends on the power of the amp. You can use this formula:

W = I*I*R/2

This means, if the max current is 5A, R is 0.1ohm, you need at
least 2.5Watt. This formula is for sinus. To be save, skip the /2...
If you double the outputdevices, you only need 1/4 as the current
halfes.

Michael

[Mordor]

I'm building an AB power amp with a dual opamp for gate polarization, with a feedback and lowpass filter for adj the virtual 0. If work i'll paste the schematic and pcb... stay tuned

[destroyer X]

When the current increase, lets think about one amplifier without signal... to be easy to understand.

Lets imagine that the bias control became crazy... and sometimes big current, and sometimes normal, some bad solder....bad contact.

When bias (polarization current, stand by current, iddle current, corrente de repouso, polarizzazione) is low..... small current passes througth this resistor.... if 1 ohm resistor and 100 mA current passing.... using ohms law, you will discover that you will have a diference of potencial (voltage), that you can measure over the extremes resistor points.... with a voltmeter measuring those two resistor leads....you will find 0.1 volts... 100 milivolts...OK!.... this will disturb about nothing, because amplifier use a enormous voltage related this small one, loosen over this resistor.

But lets think now the crazy regulator start to make confusions.... and he made now a 5 ampere current passing over this resistor, because transistor is completely "on"... and full supply energy is passing from colector to emitter and them flowing througth this resistor...... using ohm law again.... 5A passing inside this poor resistor....now 5 volts will be on the resistor.... and the power on it will be 5 watts (hot!), if this resistor burns, good, will protect transistor, sometimes create more problems..better not to burn.

This way, 5 volts will be loosen over resistor.... this voltage will reduce the entire voltage that was over its transistor... and will reduce the power over the transistor.... dividing the problem.

Lets go on..... now!.... damn crazy bias multiplier potentiometer fails again.... and now 15 ampéres passing!!!....MY GOD!... My transistor can hold maximum 20 amperes.... and maximum 250 Watts dissipation unit!...... when 15 amperes is passing inside the transistor, and in consequence througth his emitter resistor, will create a big potencial difference (voltage...because some electrons will be alike too many people trying to enter a small door...one side of the cinema door full of people (electrons) and outside the cinema doors only few people that goes here, there and everywhere.... not too electrons the outside cinema door... difference of people here, is alike difference of electrons... we call this voltage) ..... a big diference of potential over the resistor extremes again..... 15 amperes over 1 ohm, will be 15 volts..... this way, over resistor, the power will be 15 volts and 15 watts power overheating it, it will became red, burn or explode, at least will smell a lot... those, are advises to the poor amplifier's owner.....the board became black, sometimes a hole, sometimes fire.....but normally a fuse blows and all thing stop!.....if not..... lets imagine your amplifier are using plus 30 volts and minus 30 volts..... we gonna study one side.... the top side..... the plus side....only 30 volts.

If this voltage is keept in 30 volts when 15 amperes are passing into transistor (what a supply!).....15 volts and 15 amperes will be over the transistor and 15 volts and the same current over the resistor...HOHOHOHOHOHÔ.... You get it "fratello".....this way, alike good friends, they divided the problem..... voltage reduced... 225 watts power over transistor, and 225 watts power over resistor..... your 250 Watts maximum transistor will be protected by the resistor.....if no resistor.... 450 watts will melt the transistor!

I made this study static... when dinamic, working, this will work too... resistor is wonderfull, and love is wonderfull too.

I piacere fratello, siamo tutti amicci, My pleasure brother, we are all friends.

regards,

Carlos

[markp]

Dont forget that a 5 watt power resistor will take huge peaks without any problem. You can get well over 50 watt transients and not burn the 5 watter. It will not take that power steady-state though.

[sam9]

And if your amp includes a zobel and you anticpate a 50k-100khz test signal, be sure your zobel resistor can take it. In actual operation it gets almost no current at all, but during initial setup and testing it can be a drag to keep smoking it!

[anatech]

I have found through testing that most resistors R33 and lower have a strong positive tempco. R10 went up to R50! These were ceramic "plate" non-inductive resistors comonly used for emitter resistors. "Bathtub" types suffer from this as well. This will throw your bias readings off and cause you to set bias to low when setting hot.
I used an HP 34401A in 4 wire mode (kelvin connection), nulled. My heatsource was a soldering iron under the body. High tech ;-). I didn't care about the numbers, just the trends and types of resistors. No list as it's easy to test yourself.
I'm thinking of paralleling some flameproofs. 1R0 seem pretty stable.
-Chris