Help with class AB amp

Discussion in 'General Electronics Chat' started by CRY&PRAY, Jan 6, 2014.

  1. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    Hello guys, i'm new here. Here is my first class AB, it works pretty well, but it distort a lot when raising the volume giving me something about 4-5 watt clear on a 36v 1A PSU. The heat it's not excessive, the 20 wat power is there but sounds bad. here's the schematic: This is the first version, where the small transistor (pre) has on the base a 1 MOhm resitor (now is 100 KOhm and sounds better) and a 500 KOhm resistor that i've removed. The problem is "Where is my power??", any help would be apreciated a lot. :):)

    [​IMG]
     
    Last edited by a moderator: Jan 6, 2014
  2. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
    Hello,

    Looking at the schematic R5 is 1 mili Ohm in stead of the mentioned 1 mega Ohm.
    This would be like a short from output to input.

    Bertus
     
    CRY&PRAY likes this.
  3. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    Yeah my bad, that was 1 Mega Ohm, then i've put a 100 KOhm

    and i really don't know why everybody put's that resistor there. is that to bias the transistor?

    (i have the amplifier playing right now, it works pretty well)

    i'm building this amplifier to understand better how they works (i know why there are those 5W resistor, the decopuling cap and the diodes for the 0.7 v bias) but i still don't get why the scematics for 400 w amps on internet have all these transistors in the beginning, i know that are needed more Power transistor but all these small one?.
    i would like to build a 500 W x2 amp but without knowing how they works... what if i put more transistor (like 2sc5200..) to my amp and raise the voltage? will it works efficentily or just spreading a ton of heat and giving me 1 watt...
     
    Last edited: Jan 6, 2014
  4. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    so pratically i need these answers:
    1) why the amp that i build doesn't push at lest 20-25 w (36 w psu)
    2) why in other schematics of amplifiers there are all these transistor in the beginning?
    3) will my amp work like those (300-400w) by putting more powerfull transistors (mine is using a pair on 2sc5198 and 2sa1941, as pre a small bc548) and raising the current? maybe dual supply and how can i mod the amp tu run dual?

    4) this amp seems to push out 3-4 v AC to my speake which seems pretty low and it's consuming only 300 milli ampere why?

    Thanks all, Francesco
     
  5. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    R5 provides negative feedback to reduce distortion.

    The reason the other amps have all those parts is to better drive the output stage to give higher, low distortion power. Your circuit is inherently limited to relatively low power from limited drive current to the NPN output transistor (through the 1k resistor). You could likely increase the output power some by using a Darlingtion or a Sziklai pair for the two output transistors.
     
  6. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    Il try the Sziklai pair using other two power transistor, i'm pretty sure that the amp that i made could be done better, any opinion? it distort at 50-60 % and i can't get to some real power without that bad class B type sound.
    And i've seen a lot of amp with the Long tailed pair thing, how that thing can help the amplifier? comparing the output with the input?:confused:
    and the low power passing thru that 1k resistor? it get's really hot, how can i avoid that and have higher power?

    Thx to all.

    i also have a Omnitronic e900 amp, the power supply is perfectly working (95 v +-) howerver the amp is completely gone, all the 8 transistors need to be replaced, 4 transistors on the board are kinda xploded and 2 small one attached on the heatsink are gone too, i think those are for the bias, because they have a pot near them and both channels. I think that i'll maybe repair that amp... if not, i would like to re-use the power supply, That's why i'm building an AB amp that can handle that 1000W of power with a good efficency :)[​IMG]
     
    Last edited: Jan 6, 2014
  7. Veracohr

    Well-Known Member

    Jan 3, 2011
    552
    76
    Those transistor models have an ABSOLUTE maximum current rating of 200mA. You probably need different transistors for starters.
     
  8. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    I'm sorry, on the schematics is written 2n3904, that's a small one, i'm using a pair of 2sc5198 and it's Brother, as the pre using a bc548. I did mentioned that somewhere :) these transistor can handle 140 V 10A (seems too much, maybe i read wrong). That's why i'm thinking about raising the current to somewhere like 95+- V form the toroid of the Omnitronic E900 amp that's in the picture
     
    Last edited: Jan 7, 2014
  9. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    The math is pretty easy:

    P = I (squared) x R

    If you can get 5W, your output is driving roughly 0.8A (RMS) or about 1.2A peak. The output transistors are beta limiting.
     
    Last edited: Jan 7, 2014
    CRY&PRAY likes this.
  10. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    but to get that 10A collector current you probably have to stuff at least 0.5A into the base.
     
  11. crutschow

    Expert

    Mar 14, 2008
    13,016
    3,235
    That should certainly help.
    The first transistor of the Sziklai pair can be a lower current device since it only has to provide base current to the second transistor, which is likely no more that about 1/20 of the output current..
     
    CRY&PRAY likes this.
  12. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    Yeah, i know, but i have only other 2 of the power transistors, so i'm gonna use them, i have to wire the NPN to the PNP and vice versa right?

    I'm currently running the amp for all the day listening to my pc, and sometimes i push it just to see if doesn't xplode and it seams to be something around 10 watt or maybe the speaker is really sensitive (i think not, there are 2 8" woofers and a plastic tweeter)

    if i add the long paired thing (http://www.radio-electronics.com/info/circuits/transistor/long-tailed-pair.php) will work better with more gain? and if yes, how can i modify the scheme that i did to be working with that? and is better to use a dual supply? if yes how can i add that?

    Thx everybody here is nice and kind.
    i'm italian so i'm sorry for my bad english, and i'm 15 now :)
     
  13. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    The amp is using only 0.3 A (measured with a multimeter) so i don't think that at the output i have 0.8 A :confused:. i get 3-4 volts AC at the output with no load. So i did the math W=V*A and i got 1 watt.

    but there is a waste of 35 watt of power!!! the psu is a bose switching psu, giving 18 v 1 A (*2) and i've wired that in series to obtain 36 v.
    those 35 watt should dissipate somewhere, but the heatsink is normal, maybe at 45-50 °C max
     
  14. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0

    i've tried to wire the skisomething pair... i've tried PNP and NPN, nothing, tried to invert the positive and negative.. nothing.. tried NPN with NPN and PNP with PNP... nothing... i'm kinda retarded or something.. how can i wire them?

    i'm curently trying to make a long tailed pair to see if something changes
     
  15. tubeguy

    Well-Known Member

    Nov 3, 2012
    1,157
    197
    Don't feel bad. A cup of coffee always helps me, but maybe you don't drink coffee (yet). :)

    Understand that a Sziklai pair would replace each of your single output transistors.

    EDIT: It would help to see some actual circuits you have tried.
     
    Last edited: Jan 7, 2014
  16. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    hahaa :) still not drinking coffee ;)
    yeah, i'll post some pictures here, the circuit is the first schematic that i made, without R4, and R5 is somewhere 500 KΩ, the transistors are 2sc5198 and a 2sd688 (too lazy to put a 2sa1941) and of course a little bc548.
    by doing a sziklai pair i'm kinda making a NNN and PPP transistor isn't it? how can i drive them after?

    [​IMG]

    [​IMG]

    [​IMG]
     
    Last edited: Jan 7, 2014
  17. donpetru

    Active Member

    Nov 14, 2008
    186
    25
    I attached a circuit diagram of audio amplifier. Analyze the diagram and see what you've done wrong. Good luck!
     
    CRY&PRAY likes this.
  18. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    there are the final transistors with the resistors, 2 diodes for the bias that's NOT attached to the PSU but at the speaker, after the cap, so it's taking some sort of current or it's just for a negative feedback... Yep the bias is taking current from a 3rd diode and a 4th transistor that i don't get what is it doing. There is the pre transistor that's driving the big guys and it's taking current from the output before the cap so it's taking Vcc/2. R6 and C4 are for some sort of power adjustement as that transistor is running in class A. Why 8 Ω 5W at 22 V?? is that so unefficent? still don't get what i did wrong, it's playing pretty well, but it's giving 10-20% max (if lucky) of the vcc.
     
  19. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    The current has to be going somewhere. Probably the idling current. You should be able to measure the drop across the series resistors.
     
  20. CRY&PRAY

    Thread Starter Member

    Jan 6, 2014
    84
    0
    The only part that's getting hot is the 1k resistor on the collector and the base of the NPN Transistor (across this resistor there are 25.8V). The voltage on the bases are: 8.6v on the pnp and 10 v on the npn (these voltages are by tapping the bases and the negative). if i tap the output before the cap and the bases i got 0.6 on both. If i measure te voltage across the 2 diodes i got 1.3-1.4 v. However that resistor is a 1/4 w, which if it's getting that 30w of power waste would explode.
     
Loading...