analysis for 100w amplifier ciruit

Goto page Previous 1, 2

EDAboard.com Forum Index -> Analog Circuit Design -> analysis for 100w amplifier ciruit

Author

Message

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

07 Nov 2009 0:34 analysis for 100w amplifier ciruit

One transistor in an opamp or in an amplifier circuit has the most voltage gain. A capacitor from its collector to its base is amplified (Miller compensation) and reduces high frequencies which are less likely to cause oscillation due to phase-shift in the amplifier.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

07 Nov 2009 17:30 analysis for 100w amplifier ciruit

thanks for the reply
then how could I calculate the capacitance?
should I but the capacitor on the transistor connected to the input signal?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

07 Nov 2009 21:34 Re: analysis for 100w amplifier ciruit

mr_byte31 wrote:

thanks for the reply
then how could I calculate the capacitance?
should I but the capacitor on the transistor connected to the input signal?

The amplifier circuit is extremely simple so maybe it does not need a compensation capacitor.

To calculate the capacitor's value and where it should connect you must analyse the gains of the transistors and their phase shifts.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

07 Nov 2009 22:43 Re: analysis for 100w amplifier ciruit

pippone1987 wrote:

Of course, but they are not so cheap. For the drivers you can use 2SC2238 and 2SA968 (respectively for pull-up an pull-down net) while for the outputs you can use 2SA1216 and 2SC2922.

I couldn't find any of those transistors
how could i simulate them ?

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

08 Nov 2009 7:45 Re: analysis for 100w amplifier ciruit

pippone1987 wrote:

In my opinion the distortion is minimum when you use a couple of Sziklai pair in the output stage (or CFP complementary feedback pair) which guarantee a 100% feedback

I used TIP41 and TIP42 I know that it isn't logic to use them because they can't supply much current but it will be in simulation only till i find any substitutions
but the distortion is the same and i can't see any difference using Sziklai pair
any suggestions?

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

09 Nov 2009 8:19 analysis for 100w amplifier ciruit

ops

Sziklai pair and a capacitor(to stop oscillation)made the distortion reaches 0.15%
is there any way to improve more?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

09 Nov 2009 13:03 Re: analysis for 100w amplifier ciruit

Good opamps and good audio amps have low distortion circuits. They are careful to control the amount of current in their output transistors (for class AB) to avoid crossover distortion (class B) and avoid an idle current that is too high (class A).

Good opamps have an open-loop gain of 200,000 to one million at low frequencies. Then when negative feedback is added the distortion at low frequencies is extremely low.

Good quality audio amplifiers are the same.

The compensation capacitor (to stop oscillation) reduces the open-loop gain at high frequencies which results in less negative feedback at high frequencies which increases the distortion at high frequencies.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

10 Nov 2009 7:38 analysis for 100w amplifier ciruit

I tried to change the gain
by changing the resistance R4 and capacitor C2
the gain can reach 4500 but the distortion increases !!!

FvM

Joined: 22 Jan 2008
Posts: 5154
Helped: 766
Location: Bochum, Germany

10 Nov 2009 9:51 Re: analysis for 100w amplifier ciruit

Quote:

the gain can reach 4500 but the distortion increases

Yes, understandable. The suggestion was to increase the open loop gain, not the closed loop gain.
As also said, it wouldn't help much anyway without getting rid of the class B output behaviour.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

10 Nov 2009 11:02 analysis for 100w amplifier ciruit

do u mean the open loop gain of the diff pair?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

10 Nov 2009 13:45 analysis for 100w amplifier ciruit

The differential pair has a high emitter resistance and drives the low input resistance of the TIP41 so its gain is low or it has loss.

The TIP41 has the most gain of maybe 120.

The darlingtons are emitter-followers with no gain.

Therefore the total open-loop gain is low.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

10 Nov 2009 18:01 analysis for 100w amplifier ciruit

should I substitute TIP41 with a darlington transistor or another transistor?

Added after 2 hours 43 minutes:

anyway i modified the circuit again
now it gives me around 0.086% at 1mv
here is the sch
is there any improvement that can be made?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

11 Nov 2009 1:10 analysis for 100w amplifier ciruit

Replace R13 with a current source that has a very high impedance for a very high gain from the TIP41 transistor.

Remove R2.

Replace the three huge diodes with two normal 1N4148 diodes.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

11 Nov 2009 7:42 analysis for 100w amplifier ciruit

for the diodes I couldn't find 1N4148 in orcad so i used any diode
Ok i will remove R2 and see the effect
i tried simple current mirror using 2N2222 but it gives a very low current with a very bad performance (current = in nA)

Added after 7 minutes:

should i use current source like that
http://upload.wikimedia.org/wikipedia/en/1/10/Const_cur_src_112.gif

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

11 Nov 2009 15:37 Re: analysis for 100w amplifier ciruit

This amplifier is very similar to yours. http://www.sound.westhost.com/project3a.htm
It produces 60W into 8 ohms or 100W into 4 ohms. Its distortion is only 0.04% at 50W into 8 ohms.

1) It uses a current source at the emitters of the differential pair of input transistors.
2) It uses an adjustable Vbe multiplier transistor instead of diodes to bias the output transistors.
3) It uses a bootstrap capacitor C5 to increase the output power and increase the gain of Q4.
4) It has frequency/phase compensation capacitors.

Sorry, but you need login in to view this attachment

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

11 Nov 2009 18:55 analysis for 100w amplifier ciruit

I have a little question
in the diff amplifier why didn't he use a current mirror as a load?

Added after 1 hours 12 minutes:

another question plz Smile

could I use TIP41 and TIP42 for the last stage??
TIP41/42 can supply till 6A
I would use 3A or less to have 72W in 8 Ohm speaker
is that correct?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

11 Nov 2009 19:47 analysis for 100w amplifier ciruit

If the diff amp uses a current mirror as its load then the amp will have a higher open-loop gain which would be better because then its negative feedback will be more and its distortion would be less.
But it would be more complicated and cost more.

72W into 8 ohms is a signal of 24V RMS which is 34V peak. Then you need a plus and minus 36V to 38V power supply.

34V/8 ohms is a peak current of 4.25A.
The TIP41 and TIP42 have a minimum current gain of only about 10 at such a high current.
They might melt with such a high power dissipation.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

11 Nov 2009 20:24 analysis for 100w amplifier ciruit

I am sorry but I couldnot understand
If the collector current is 4.25A this is still Ok because the transistor can drive till 6A.!!!

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

11 Nov 2009 20:30 analysis for 100w amplifier ciruit

The datasheets for the TIP41 and TIP42 shows that they can survive 6A, they don't show their extremely low performance at such a high current.

Never ever use a transistor at its max rating.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

11 Nov 2009 21:11 analysis for 100w amplifier ciruit

but there is a 2A difference !!?
Ok i may use 2SC5200/A1943 or 2SA1216/2SC2922.?
whats ur idea?

Audioguru

Joined: 19 Jan 2008
Posts: 1145
Helped: 94
Location: Toronto area of Canada

12 Nov 2009 0:11 analysis for 100w amplifier ciruit

A 4 ohm load uses much more current than an 8 ohm load.

I have never used Japanese, Korean nor Chinese transistors and I probably never will. So I didn't look at their datasheets.

mr_byte31

Joined: 19 Oct 2005
Posts: 198
Helped: 1

12 Nov 2009 8:13 analysis for 100w amplifier ciruit

now I am finishing the PCB but I think I need a preamplifier for Mic
I am thinking about a popular IC that can amplify the mic signal and add some effects(bass, treble, echo, repeat)
do you have any experience about that?