Post Go back to editing

ADAR2004 IF frequency range and performance

Category: Datasheet/Specs
Product Number: ADAR2004

ADAR2004 specifies IF of 0 to 800 MHz. What happens if you were to try and mix a signal down beyond that range? For example, if I tried to mix a signal down to 1 GHz IF, would there just be reduced performance in terms of gain/linearity/noise? or is there some filtering that happens at 800 MHz that would prevent me from using this chip in that way?

Parents
  • The IF output has a natural roll-off that begins at around 800 MHz.  But you can see from the plot below that the roll-off is gain dependent, that is, it rolls off earlier at higher gains. I would expect that IP3 would show similar behavior. So I think that the part can definitely be used at 1GHz but try to keep the gain low. Please note that I'm including the plot below because it shows performance out to 2 GHz. It was taken early in the development. The gain is showing lower because we were only looking at one of the two differential outputs (so it's 6 dB lower). 

Reply
  • The IF output has a natural roll-off that begins at around 800 MHz.  But you can see from the plot below that the roll-off is gain dependent, that is, it rolls off earlier at higher gains. I would expect that IP3 would show similar behavior. So I think that the part can definitely be used at 1GHz but try to keep the gain low. Please note that I'm including the plot below because it shows performance out to 2 GHz. It was taken early in the development. The gain is showing lower because we were only looking at one of the two differential outputs (so it's 6 dB lower). 

Children
  • This is great data. Thanks! I purchased an eval board to characterize this for our application, but do you expect the performance past 800 MHz to vary much part to part? Trying to determine whether the performance outside 800 MHz is something we can rely on, and whether ADI would ever modify the part such that the performance past 800 MHz could look totally different without our prior knowledge.

  • Unfortunately, no part-to-part distribution data to share and to confirm the performance. I don't think there was a part revision that could result in a big performance difference outside 800MHz.