0
\$\begingroup\$

I have many problems with TDA 2050.

  1. Why TDA2050 use high-pass filters?

  2. What is the bandwidth of it? Does the range of bandwidth is from the range of frequencies below the human hearing?(20-20KHz)

  3. I used Pspice to simulate this circuit [replaced TDA2050 by uA741(no TDA 2050 in Pspice)], and 8ohm resistor instead of the speaker. The bandwidth is 23Hz to 4.5KHz. If I increase the resistor of load, the graph of frequency response become strange with a jump value in 12.9KHz, how it occurs? Is there something wrong?

  4. If I remove the R6C6, the jump value disappeared, why? In addition, the bandwidth between R7=8ohm and R7=1K is difference, why?

R7=8 -->bandwidth<24-4.5KHz>

R7=1K-->bandwidth<11.5-20KHz>

TDA2050 R7=8 Remove R6C6 In this figure, R7(load)=1K Pspice

\$\endgroup\$
0

1 Answer 1

1
\$\begingroup\$

Despite not having a TDA2050 in your spice simulator, replacing it with a 741 tells you close to nothing about the TDA2050. Firstly the 741 isn't geared up (in reality or simulation) to driving an 8 ohm load and a zobel network comprising a 2.2 ohm resistor in series with a 0.47 uF capacitor.

To try and answer some of your questions

  1. High pass filters - it must have an output decoupling capacitor or your speaker will be dc grounding the output to 0V - the o/p from the device is centred between gnd and the positive supply - this shouldn't be attached to the speaker or you might damage the speaker or the TDA2050. The 2.2 uF input capacitor is needed because your input signal is ground referenced and the TDA2050 has bias voltages on its inputs that shouldn't be disturbed. It is not an op-amp despite looking like one.
  2. The bandwidth is in the data sheet. It says 20Hz to 80kHz on page 3.
  3. The strange jump (as you call it) at 12kHz is a 5dB peak in the response and is likely to do with the inappropriate use of a 741 in the circuit you've used and some peculiarity of its spice model. This is not-easy to explain but it's not likely to happen in a real 741 and the zobel network R6+C6 shouldn't be used with a 741 - this type of RC network is an attempt to linearize the impedance of the loudspeaker connected.
  4. Explained in (3)

Increasing the load from 8 ohm (inappropriate for a 741) to 1k ohm will make a difference - the 741 could expect to see a load like 1k but not 8 ohms.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks a lot. But I am still confused about the bandwidth of TDA2050. In experiment, I find the bandwidth is 27Hz-22KHz. why not 20Hz to 80kHz? How to get a TDA2050 in pspice? In addition, can I see the zobel circuit effect in Pspice? What should I do? \$\endgroup\$
    – user33593
    Commented Dec 4, 2013 at 13:41
  • \$\begingroup\$ @user33593 I doubt if pspice will have a model for the device. Regarding the 80kHz - if you look at the point where gain becomes unity, this should be 80kHz. \$\endgroup\$
    – Andy aka
    Commented Dec 4, 2013 at 14:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.