I'm trying to get 10 pounds of processing in a 5-pound radio. I have a homemade signal conditioning block that (should) allow for integration and decimation suitable for low-rate reporting. I've been placing my block between the radio and the stock DDC but the DDC does more than I need and takes a lot of space. There's no room left to add anything else. I tried both the fir and moving average blocks with keep-one-in-n block but those combinations are even bigger than the ddc and don't leave room for my block. I'm guessing that there's some overhead in building a stand-alone block and the DDC has been refined quite a bit. Can anyone suggest a lighter-weight integrator+decimator? I'm looking to get 2 channels at full native rate (30 MHz) compressed down to somewhere around 1 MHz or a little less for transport over Ethernet.
Also, is there any chance that the compatibility mode is somehow overriding the block order? I've noticed that in Gnuradio I can ask for radio sample rates lower than I should be able to even though my block is between the radio and the ddc and the output data looks suspiciously as though the ddc has been placed before rather than after my block despite UHD_LOG_CONSOLE_LEVEL=trace detailed reporting that the blocks are connected in the right order. It's difficult to debug since I can't save full-rate data to be sure exactly what is coming in. I probably need to test with an rfnoc replay block to feed it known data. Thanks, as always. _______________________________________________ USRP-users mailing list USRP-users@lists.ettus.com http://lists.ettus.com/mailman/listinfo/usrp-users_lists.ettus.com
