[Discuss-gnuradio] fpga halfband filter
Dear everyone: In the FPGA code of halfband_decim,v of USRP1, it has a input strobe_in, which indates a new data comes. Strobe_in is determined by the decimation rate. For example if we decimate by 6 we write to the register FR_DECIM_RATE 6/2-1=2 and strobe_in comes high every 3 clocks. My question is that if I want to use halfband filter only to get a decimation of 2 (for other purpose of my FPGA project, not used in USRP), how should I set strobe_in? I noticed that whether I set it equal to the clock or set it equal to 1, the strobe_out always comes high every two clocks and data_out is different. So can you tell me how I should set strobe_in? Besides my data are 8 bits wide, is it proper to pad 0 to my data to reach 16 bits wide? Best regards___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] fir filter parameters
Dear everyone: In ofdm_receiver.py there is chan_coeffs = gr.fihan_coeffs(1.0,1.0,tb+bw,tb,gr.firdes.WIN_HAMMING), which is to determine the parameters of fir low-pass filters. My question is about the third parameter, which denotes the center of transition band, and it is computed as tb= bw*0.08. So can you explain to me why we should set the center of transition band tb+bw and how tb is computed? Besides in benchmark_rx case, midpoint of trans. band is directly set 1 and width of trans. band 0.5. So in general how do we set the bandwith and midpoint of band? Best regards___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] Read daughterboard eeprom from python
Dear everyone: Is there a method that I can read daughterboard eeprom from python? Or is there anyway that I can know what stored in the daughterboard eeprom in a visible way? Best regards___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] Basic question about filter
Thank you To make it clear, I use USRP1. Does that mean in common the signals that come out of ADC chips (in USRP case it is the signals come out of AD9862 before processed in FPGA) is already filtered and aliasing prior is partly eliminated ? In the datasheet I do not see anything about that. If I am wrong please correct me. Best regards At 2012-06-19 09:54:11,"Marcus D. Leech" wrote: On 18/06/12 09:46 PM, signalswdm wrote: Hello everyone: You see in theory there should be a low-pass filter after ADC for eliminating the image. Also after DDC there should also be a filter. But in USRP there are not such things. So do the cic filter and half band filter implement that function besides decimation? Best regards ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.orghttps://lists.gnu.org/mailman/listinfo/discuss-gnuradio Actually, in typical architectures using I/Q, there are low-pass filters *in front* of the ADC to eliminate aliasing prior to sampling. So the CIC decimator "sees" a signal that has already been band-limited appropriate to the sample rate of the ADCs. The CIC decimation process *is* filtering, so out-of-band products appearing at the output are usually suppressed by at least 80dB, and usually more. -- Principal Investigator Shirleys Bay Radio Astronomy Consortium http://www.sbrac.org___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] Basic question about filter
Hello everyone: You see in theory there should be a low-pass filter after ADC for eliminating the image. Also after DDC there should also be a filter. But in USRP there are not such things. So do the cic filter and half band filter implement that function besides decimation? Best regards___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] RF and AD/DA
Dear everyone: Recently I have been doing my graduation design, wchich is the implemention of 802.11a ofdm PHY. My tutor required me to use FPGA so I implemented baseband OFDM modulation-demodulation using Virtex5 FPGA. My tutor decided not to use USRP because the RF bandwidth is too limited. Now I have to do up-conversion to 5.8GHz and I decided to use XCVR2450 board. Besides some configurations of the registers of MAX2829, such as bandsellect, divider ratio and so on, do I need to do anything else? Does AFC, AGC,eliminating dc offset and something like that need to be done in FPGA? I have been studying benchmark_tx and benchmark_rx for some time, but I don't find AFC or anyting like that. Any of your responses will be very appreciated.___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] About AD9862 latching data
Dear everyone: You see in the FPGA codec datas are interpolated into 32Msps and sended to AD9862. Then AD9862 interpolates the datas by 4. But AD9862 is configured to latch datas using CLKOUT1,and CLKOUT1 is 64M. Though CLKOUT1 is not linked to FPGA, it seems a little confilcting. So can anyone tell me the how exactly AD9862 latches datas? Besides,does the 64Msps A/D and 128Msps D/A in the datasheet of AD9862 mean the maximum achievable rates or the fixed working rates? If I change the VCTCXO,for example 40M,then will AD9862 work according to 40M? Any answers will be very appreciated.___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] Something about xcvr2450 daughterboard
Dear everyone: Recently I am studying the codes about daughterboard xcvr2450(mainly usrp\host\lib\db_xcvr2450.c) and there are some problems confuse me. 1. You see there are two steps in setting the frequency. First daughter board tunes as close as the target fequency and return the actual frequency . Second AD9862 perform the DUC procedure according to target fequency and actual frequency . But in function xcvr2450::set_freq() the so called actual_freq is set equal to target_freq and the computing of actual_freq is not used(there are // at the begining of the line). Why?? 2.Accoring to the table12 in the datasheet of MAX2829, we derive the divider ratio either using or .We should divide 20M.But it seems in function xcvr2450::set_freq() 64M/3 is used to be divided instead of 20M.Why?? Any of your answers will be very appreciated.Thank you.___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] A function named "usb_control_msg"
Dear everyone: You see in usrp_basic.h there is a function named _write_9862() and it uses function usrp_9862_write(),defined in usrp_prims_common.h. And usrp_9862_write() uses function usrp_spi_write(), also defined in usrp_prims_common.h. usrp_spi_write() uses function write_cmd(), defined in usrp_prims_libusb0.c and usrp_prims_libusb1.c. In usrp_prims_libusb0.c write_cmd() uses a function named usb_control_msg(). But I can not find where usb_control_msg() is. Can anyone tell me where it is???___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] The values written into AD9862 registers
Hello everyone： You see there are 63 registers in AD9862, which are configured by SPI. It seems in usrp_stand.cc and usrp_basic.cc there are codes about how to write and read AD9862 registers, but I still have not yet found the exact values written into all the 63 AD9862 registers. Can you point to me where the codes about the exact value written into AD9862 registers are?? Thank you!! ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
