Audio Amp - Differential Stage Problem

Thread Starter

seanb

Joined Apr 21, 2011
6
I was given the job of designing an amplifier, there are not many constrictions other than the rail voltages (+/- 13.8 V) and load impedance (8 Ohms). I've spent hours trying to design this thing and am having trouble. I believe that the 2nd and 3rd stages of my amp are working fine but the differential stage not only inverts the signal, but it is also de-amplifying the signal instead of amplifying. The schematic that I am having trouble with is attached, any hints or suggestions would help.

Thanks!
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
Your output transistors are tiny little surface mounted low power devices. Maybe they can drive 32 ohm headphones but will melt trying to drive an 8 ohm speaker to 7W.

If the current sources are correct, R3 and R4 set the voltage gain at only 3.9. Then it will need a 5.6V p-p input signal for full output power.
 

designnut

Joined Apr 21, 2011
33
The diff amp operated into a base which is a low impedance load, you will not see a coltage gain, only a current gain. You might try reducing shunt feedvack resistor to force more gain,
 

Thread Starter

seanb

Joined Apr 21, 2011
6
The feedback is one of the things that I'm having a problem with, if I widen the gap between the two feedback resistors, the output will cease being sinusoidal which is an obvious problem.

Audioguru - in the real design, I will be using heatsunk Tip 41 and 42s, but I did not have a model to use with LTSpice, so I kind of chose some random transistors that seemed to be complimentary
 

Thread Starter

seanb

Joined Apr 21, 2011
6
So I figured out we're driving the transistors too hard with the input signal, it works if we drop it down, but we require a .5 volt peak input

EDIT: actually it seems like we can either decrease the input signal or increase the load resistance, but the load resistance is an 8 ohm speaker so that seems counterproductive
 
Last edited:

Audioguru

Joined Dec 20, 2007
11,248
The output impedance of your amplifier should be extremely low (0.04 ohms or less) so that the speaker impedance has nothing to do with voltage gain.

If you add emitter resistors to the differential transistors then they will be able to have a high signal level.
 

hgmjr

Joined Jan 28, 2005
9,027
I was given the job of designing an amplifier, there are not many constrictions other than the rail voltages (+/- 13.8 V) and load impedance (8 Ohms). I've spent hours trying to design this thing and am having trouble. I believe that the 2nd and 3rd stages of my amp are working fine but the differential stage not only inverts the signal, but it is also de-amplifying the signal instead of amplifying. The schematic that I am having trouble with is attached, any hints or suggestions would help.

Thanks!
Q9 in your design appears to be clamping the collector of your input differential pair to one diode drop away from the negative power rail. This is probably going to need a revisit.

hgmjr
 

Audioguru

Joined Dec 20, 2007
11,248
Q9 in your design appears to be clamping the collector of your input differential pair to one diode drop away from the negative power rail. This is probably going to need a revisit.
No.
Q9 is the VAS (voltage amplification stage) that has a very high open-loop voltage gain. The negative feedback from the output of the amplifier to the differential inverting onput transistor causes the voltage swing at the base of Q9 to be very low and linear.
 

Thread Starter

seanb

Joined Apr 21, 2011
6
Thanks for the modified schematic audioguru, it's much better. I still keep burning things out though. The Tip41s and tip42s seem to keep pulling too much current from the power supply, I'm still working at it
 

Audioguru

Joined Dec 20, 2007
11,248
The output transistors conduct more current when they get warm which makes them warmer which makes them conduct more current which makes them warmer which .... and is called "Thermal Runaway".
Thermal Runaway is prevented by adding emitter resistors (try 0.33 ohms) to each output transistor and by fastening the diodes to their heatsink. The current in the output transistors should be about 20mA to 50mA DC to avoid crossover distortion.
 

Thread Starter

seanb

Joined Apr 21, 2011
6
thanks, I'll try that, but we're kind of running out of time, I might have to end up doing something simple in order to get a grade. I would enjoy this if I wasn't so pressed for time.

Also the current into or out of the output transistors should be 20mA-50mA, I believe you are talking about the base current but I just want to make sure.

Another edit: It seems reasonable to me, but this circuit should pull about 700-800mA from the DC powersupply we're using correct?
 
Last edited:

PRS

Joined Aug 24, 2008
989
I ascribe to all of Audioguru's advice. You need bigger transistors at the output and even your drivers are too small; they should be medium power xistors. The voltage gain sould be at least 25 volts per volt. So your feedback resistors should have a ratio that makes for this gain.
 

Audioguru

Joined Dec 20, 2007
11,248
The DC collector to emitter current in the output transistors should be 20mA to 50mA making them operate in class-AB (look it up in google) to avoid crossover distortion.

With a power supply that is plus and minus 13.8V, the output will have a voltage swing of about 21V p-p which will create 7W at clipping in an 8 ohm speaker.
The amplifier will be about 60% efficient so its heating at clipping will be 4.7W.
The total power from the power supply at clipping is 11.7W so the power supply current is 11.7/27.6V= 424mA plus the idle current of 20mA to 50mA.
 

Thread Starter

seanb

Joined Apr 21, 2011
6
Okay that makes more sense to me, but everything I try always ends up pulling about 700mA collector to emitter in the output stage, I've even tried moving to op-amps as the differential input to guarantee that the first stage is set up correctly. I thought I understood class AB amps.
 

Audioguru

Joined Dec 20, 2007
11,248
The two diodes keep the output transistors from having too much current. If a diode is blown open or connected backwards then the output current will be very high.
 
Top