biff44
Advanced Member level 6
- Joined
- Dec 24, 2004
- Messages
- 5,059
- Helped
- 1,376
- Reputation
- 2,748
- Reaction score
- 1,060
- Trophy points
- 1,393
- Location
- New England, USA
- Activity points
- 37,988
I have a system demonstration project, and I need to do some DSP processing on microwave signals. I am not a DSP engineer, so I need some recommendations! I am looking for off the shelf ADC/DSP board solution for the demonstration. Here is what I am trying to do:
I have 1 transmit and 3 receiver channels. I downconvert all 4 channels to a lower frequency range of 10 MHz to 109 MHz. In any one channel I need to find two tones, and computer their amplitude and phase. I need the phase information very accurate, maybe to the 0.01 degree range (certainly better than 0.1 degree). The idea was that we would sample the data for 3 widths of 100 microseconds, do 3 ffts, and average the result--doing this regimen two times per second.
I was thinking I needed 4 cards, one for each channel. I have an external 300 MHz clock with low phase noise that I would like the boards to run off of. A 300 MHz clock sounded good nyquist-wise for my highest frequency 109 MHz tone. I was thinking I needed a 12 bit ADC input on each board (8 bits to achieve my desired phase accuracy for a 22 dB carrier to noise ratio, and another 4 bits to handle some dynamic range in the received signals). But 14 bits would be even better. We were thinking that the dsp would just do a 30,000 point fft, and output data in 10 KHz wide bin widths.
That is about all I know so far.
Does someone make an off the shelf DSP/ADC data acquistion board that can handle that 300 MHz clock rate with at least 12 bits. I would need to hook up 4 such boards into a van-portable system using pci express or some other connection means to a central computer motherboard.
Suggestions?
I have 1 transmit and 3 receiver channels. I downconvert all 4 channels to a lower frequency range of 10 MHz to 109 MHz. In any one channel I need to find two tones, and computer their amplitude and phase. I need the phase information very accurate, maybe to the 0.01 degree range (certainly better than 0.1 degree). The idea was that we would sample the data for 3 widths of 100 microseconds, do 3 ffts, and average the result--doing this regimen two times per second.
I was thinking I needed 4 cards, one for each channel. I have an external 300 MHz clock with low phase noise that I would like the boards to run off of. A 300 MHz clock sounded good nyquist-wise for my highest frequency 109 MHz tone. I was thinking I needed a 12 bit ADC input on each board (8 bits to achieve my desired phase accuracy for a 22 dB carrier to noise ratio, and another 4 bits to handle some dynamic range in the received signals). But 14 bits would be even better. We were thinking that the dsp would just do a 30,000 point fft, and output data in 10 KHz wide bin widths.
That is about all I know so far.
Does someone make an off the shelf DSP/ADC data acquistion board that can handle that 300 MHz clock rate with at least 12 bits. I would need to hook up 4 such boards into a van-portable system using pci express or some other connection means to a central computer motherboard.
Suggestions?
Last edited: