Can I Use a DAC to Control the Amplitude of a 2V Square Wave with AC Coupling?

[Mindscrape]

So here is the deal: I am driving a polarization rotator to attenuate light. The polarization rotator takes a 3-5kHz AC signal, they suggest a square wave. I have a frequency generator chip that can generate a 2Vpp signal, but I want something a signal of 0Vpp-10Vpp (though most of the extinction happens around 5V). At my disposal, I have DAC that I can run with labview, and it will generate any DC voltage from 0-10V.

If this were a DC problem then I would simply configure the a differential amplifier to use a 0-2V DAC voltage to subtract the other signal (so the total signal can be anywhere from 0-2V), and then amplify with a gain of 5. However, I am not quite sure what to do with the mixed DC/AC signals.

So in summary, I am given a 2V square wave, and I want to control its amplitude with a DAC. Could I AC couple the DC signal, such that I could still use the idea of the differential amplifier to amplify the square wave to the desired value? If so, I don't really know anything about AC coupling, do you just put a capacitor to ground such that the impedence matches the impedance of the resistor next to it (what it looks like from Horrowitz and Hill)?

 

[berkeman]

Can you output the 4kHz square wave directly from the DAC?

If not, the easiest way to make the variable amplitude square wave would be to make a clipper that limits at the DAC output voltage. Have the 2Vpp input square wave drive through a resistor into the base of an NPN transistor with grounded emitter, and pull the collector up to the DAC voltage through a collector resistor (you may need to buffer the DAC output voltage, depending on the output current available). That will give you an output square wave from Vce,sat up to about the DAC output voltage, and yes, just use a capacitor to couple that waveform into your polarization rotator element. Make the DC blocking capacitor at least 10x the capacitance of the rotator element, and check all the waveforms with an oscilloscope to see if you want to tune anything up.

 

[Mindscrape]

I figured out a way that doesn't use transistors (since temperature and stuff will cause it to oscillate), and I'm not really sure why I didn't think of it earlier. I just used a multiplier, and that was that.

Thanks for the help though.