Variable Frequency Sine Wave Generator

[bgsteiner]

So for a project I am working on I would like to incorporate a variable frequency sine wave generator. The signal should be around line level (1.23V) or greater. The signal will be going to a 15W amplifier to drive a variety of speakers. I am trying to find the best or most cost effective way of doing this. My first Idea was to use a DAC like the MCP4725 to generate the signal. Really all I would need to do is generate 100Hz, 1kHz, 2kHz, 4kHz at best or just 100Hz and 4kHz. I may change the values in the future which made the DAC seems best.

So I guess what my question boils down to is is this a good option or would something else be simpler and more effective for what I plan on doing?

Thanks and have a nice day.

[suicidaleggroll]

A DAC would work, just watch out for the bandwidth of the data line.  How "clean" do you want your sine waves to be?  If you want, say, 50 steps per period, then a 4kHz sine wave would need new values sent to it at a rate of 200 kSPS.  With a 12-bit data resolution plus addressing, you're looking at a minimum of ~20 clock cycles per sample, which means you need an I2C clock speed of 4 MHz (these are all just back of the envelope calculations, you should look at the I2C interface details in the data sheet to confirm).

For AC signal generation, a parallel interface DAC would probably be a much better choice.

[Paul Moir]

There's a great little application note (AN-263) that TI has:
www.ti.com/lit/an/snoa665c/snoa665c.pdf

Table 1 is a great place to start.  You might find something on it that more fits your application than a DAC/ROM solution.