Asok, it's impossible to say something useful as long as we do not know what the purpose of your circuit is!

I can only tell what goes through my head when looking at your scheme:

1. A R-2R-ladder DAC usulally must consist of very exact values of R and 2R. So, as Richard already mentioned, narrow tolerated 34k and 68k resistors would be a better choice.

2. The reason for such big R-2R-ladder resistors probably is that the standard 8051 cannot source relevant currents. A better alternative would be to use a 74HCMOS buffer to guarantee stable logic high levels. Only with such a buffer this R-2R-ladder gives rather exact results. Unfortunately, we don't know, whether your R-2R-ladder DAC must give precise values at all.

A circuit using a HCMOS-buffer can be seen here:



3. A R-2R-ladder usually must not be loaded by additional currents. So, the additional 68k resistor (behind the R-2R-ladder) should be omitted. There seems to be no reason for it anyway. The same is valid for the high pass filter formed by 22nF (22kpF = 22nF?) and 68k in series, in front of LM386. Such a load destroys the exact performance of R-2R-ladder DAC, by introducing uneven steps in the DAC stairway. In the lower scheme a FET-OPamp can be seen, which works as a high ohmic buffer, allowing the R-2R-ladder DAC to work properly.

4. The 22nF cap (22kpF = 22nF?) at output of R-2R-ladder is an allowed load forming a low pass filter in combination with the source resistance of R-2R-ladder (about 33kOhm). But do you really want a corner frequency of 1 / 2 / pi / 33kOhm / 22nF = 220Hz?

5. The 10µF cap arround LM386 gives a gain setting of 200. Do you really want such a high gain? There's no need for it, due to the rather high signal coming from the R-2R-ladder.

6. A R-2R-ladder DAC usually consists of a much higher resolution for tone generation. What you show is a 4bit DAC, which results in extreme noise and distortion. Why not using more bits?

Kai