Sampling Rates and Evaluation Board
Hi,
I am developing an underwater acoustic localization system as part of a class project.
I am supposed to locate an acoustic tag that emits an acoustic wave underwater.
The acoustic tag that we have available on the laboratory emits a wave with 69 kHz.
The main focus of the project is the localization algorithms. So to process the signals, my professors want me to use off-the-shelf products, i.e. evaluation boards.
I have to sample the signal of 4 hydrophones that listen to the acoustic tag. I also was recommended to use ADCs with sampling rates at least 10x the frequency of the wave emitted by the acoustic tag.
But I am finding it hard to find an evaluation board that process signals with sampling rates above 690 kHz. Most of what I found are for the audio industry, giving me typically 22, 44, 96 up to 192 kHz sampling rates.
I found the EVAL-ADAU1787Z board that has an ADAU1787 audio codec. The ADAU1787 audio codec has 4 ADC with a sampling rate up to 768 kHz.
My first question is: can I use the 4 ADC channels of the ADAU1787 with the 768 kHz sampling frequency in each channel, or is this sampling frequency divided by the channels giving me most 192 kHz in each channel?
My second question is about the EVAL-ADAU1787Z board. This board has 2 mono audio inputs (AIN0 and AIN1) and one stereo input (AIN2 and AIN3). I wonder if there is a way to use all the channels, separately, as mono inputs so that I can sample all 4 hydrophones.
Thank you in advance.
ADAU1787
Integrated/Special Purpose A/D Converters
sampling_rate
eval-adau1787
Audio CODECs
Hello Murillo,
Sounds like a challenging assignment!
The four ADCs all have their own modulator and will all be sampling at the same rate and they are in phase with each other. No multiplexing. So it is good for this application.
The high sample rate has to stay in the fastDSP to stay at that rate. The fast DSP is basically a filter engine so you will have to filter the signals to get the levels but then the SigmaDSP can pick up the levels and possibly the phase difference and process that information at a slower rate. The SigmaDSP runs at the lower audio rates you have found. Then you can output via GPIO or serial audio port. There are many other details that I will not go into here. So this part looks like it is a good start for finding a solution.
Regarding the inputs. If you look at the schematic that is in the user guide, you will see that all four inputs are single ended and are similar. It was designed this way so that it is easy to connect a stereo source like a cell phone or computer using the stereo jack for channels 2 &3. Then for channels 0 and 1 it is setup for mono and for supplying power to bias up a microphone. So you may have to build some custom cables but it should be possible to do. That said, I know nothing of how hydrophones are usually connected and what kind of connectors they use. I think it will probably require to warm up a soldering iron and do a bit of custom cables.
Dave T
Hi Murillo and Dave,
I did some research on underwater acoustic communications a few years ago- it was lots of fun! We had built up a platform with hydrophones and transducers to try out different modulation and decoding schemes. We used audio parts to handle the conversion and sent the data back to the shore over a wireless link. Once, we lost our only working buoy during a bad storm- it's still out in the ocean somewhere. Anyway...
I would like to bring up the issue of frequency response. Specifically, the plots in the ADAU1787 datasheet only go up to 20kHz since this is an audio part. So, DaveThib, do we have any frequency response plots for the ADAU1787 converters that extend to 100kHz or beyond?
Murillo, are you planning to run the localization algorithm in the DSP or on a computer? If you are planning to offload processing we need to consider how the data will be transported. At 768kHz the I2S interface can only transport 2 channels. Also, you may have difficulty finding a USB to I2S interface that supports 768kHz and is also reasonably priced.
Joshua
Hi Murillo,
Dave ran a test on the ADAU1787 eval board. It was configured for passthrough, so the responses I am mentioning describe the combined transfer function of the ADC and DAC.
With the ADAU1787 sampling at 768kHz, the converters are 3dB down around 195kHz. At 69kHz, the response is only 0.5 dB down.
With the ADAU1787 set to 384kHz sample rate, the input-to-output transfer function has a -3dB point at 93kHz. At the 69kHz acoustic tag frequency, the response is roughly 2dB down.
I think either sample rate is just fine for your acoustic tag project. The signals will make it into the DSP no problem.
Joshua
Please log in to reply.