Thank for your input Jerry. I may not have explained it very well. Yes, the ring mixer is lossy and the 200 mV sine coming from the AD9850 is pretty low at those higher frequencies. That is why, in the design I use, the ring mixer module has an extra 2N3904 buffer/amp stage which is not in Farhan's original design. Take a look at this schematic if you will. The 2N3904 seems to be used as a buffer stage before it enters the ring mixer. Maybe the signal is not amplified at all, only buffered. My knowledge is too limited. If so: it there an easy way to modify the circuit so it amplifies it a bit?
