• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 99 subscribers
  • Views 939 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

66AK2L06EVM Questions

William Tandy
Prodigy 20 points
Other Parts Discussed in Thread: ADS42JB49EVM, 66AK2L06

Hello,

I am looking to design a prototype of a laser signal processing system using your hardware. I want to directly measure laser pulses captured with a photodiode by integrating the area under the curve of each pulse.

I have determined that a 250 MSPS ADC will provide sufficient resolution. I see that the ADS42JB49EVM may be exactly the evaluation module I need. Now I need to find the board to hook it up to. I am interested in the 66AK2L06 SoC EVM, but have some questions:

1. Can the ADS42JB49EVM hook up to the 66AK2L06 EVM?

2. Can the 66AK2L06EVM capture raw, unprocessed data from the ADC so I can run data analysis offline? I would like to capture between 0.5 seconds and 1 second at a time. That would seem to be between about 220 and 440 MB of data.

2a. The TSW14J56EVM seems to be able to capture raw data, which is what I want. However, I'll also need to process data in real-time later down the road, so if the 66AK2L06EVM can both capture raw data as well as process it later in real-time, that would be the best.

3. The photodiode/gain electronics connect as single-end, 50 Ohm. I see that the ADS42JB49EVM has transformers so it's "flexible", but I want to make sure that I can use the board.

3a. Also, does the EVM convert the single end to differential with hardware? I'd like differential for better signal processing, if possible.

Is there anything else I should know about trying to capture raw data with your hardware? I'm only trying to capture one second, but that's still going to be a lot of data.

Is there anything else I should know about the SoC that might limit processing the signal in real time? I want to integrate the area under the curve from the pulse by reading a buffer and processing a group of points within it. I saw in another forum post that your chip may only be able to process ~186 MSPS in real time?

Thank you

  • 0
    TI__Guru* 81335 points

    Welcome to the TI E2E forum. I hope you will find many good answers here and in the TI.com documents and in the TI Wiki Pages (for processor issues). Be sure to search those for helpful information and to browse for the questions others may have asked on similar topics (e2e.ti.com). Please read all the links below my signature.

    We will get back to you on the above query shortly. Thank you for your patience.

  • 0 in reply to Raja
    TI__Mastermind 37500 points

    Hello

    There is a design #2 that is being developed, it uses a different ADC, the ADC 14x250.  The JB69 ADC needs a special JESD204B feature HD=1, which is not

    currently supported for > 153.6Msps.   

    1. Can the ADS42JB49EVM hook up to the 66AK2L06 EVM?

    We suggest that you consider the 14X250 ADC.  The ADC EVMs have two connection methods for the 66AK2L06 EVM:

         a) there is an FMC connector (the 66AK2L06 EVM has two FMC you use the 1st one), in this manner the JESD Serdes and Sync signals are connected over the FMC, the sampling clock may connect to the ADC through an external input,

         b) still using the FMC connectors, there is a DLC card, that is connected between the ADC EVM and the 66AK2L06, this provides the clock solution for the ADC sampling clock

    2. Can the 66AK2L06EVM capture raw, unprocessed data from the ADC so I can run data analysis offline? I would like to capture between 0.5 seconds and 1 second at a time. That would seem to be between about 220 and 440 MB of data.

    While the SOC can ping pong buffer a single ADC sampled at 245.76Msps for the design #2 example, this has signal processing performed.  The design #2 example,

    ADCin (real)245.76 -> Zero Interpolate to 491.52, Hilbert Transform ( -Fs/4 Frequency translation, filter, dec/2) to  245.76 (cplx), NCO frequency translate 61.44, Rx block Filter, Dec2 122.88 (cplx) -> Channel processing PFIR filter, 1 output channel 122.88Msps(cplx)

    2a. The TSW14J56EVM seems to be able to capture raw data, which is what I want. However, I'll also need to process data in real-time later down the road, so if the 66AK2L06EVM can both capture raw data as well as process it later in real-time, that would be the best.

    The TSW can capture the raw data, the host PC can collect the data over USB.  You would need to write an application to use the 66AK2L06 to receive a data file (probably over Ethernet), and do the post processing.

    3 and other ADC questions, need to be posted to the data converter E2E forum. 

    In the next month, there should be a design folder for a design #2 radar demonstration, which is the base project.  There is currently a design #1 demonstration you could reference to start with

    Regards,

    Radio Joe

  • 0 in reply to Radio-Joe
    Prodigy 20 points

    Thank you, Joe. I appreciate the responses. I should say I'm a graduate student trying to learn these things so some of what you said shot way over my head. I'm trying to learn. I am hoping to invest both time and money in learning TI specific hardware. Hopefully TI can get me where I need to be.

    Follow-up questions:

    1. So I should go with the 14x250 because the 66AK2L06 can't actually run at 250 MSPS for the JB49EVM? If I use the 14x250, does that mean the 66AK2L06 can read at 250 MSPS?

    2. Even if the 66AK2L06 can read at 250 MSPS, it sounds like it can't actually store a half second or so of data. It sounds like I have to buy one board (the TSW14J56EVM) to store data and one board (the 66AK2L06) to process data. Is that correct?

    3. I liked the JB49EVM board as it had the clocking solution built in. What is the "DLC" board you're saying I should go with in order to connect the 14x250to the 66AK2L06?

    4. Is it possible to just connect the JB49EVM to the TSW14J56EVM so I can save raw data with no real-time processing? Is there a better, more simple way to read and store 250 MSPS data so I can look at the raw data?

    5. Is there a more powerful version of the 66AK2L06 coming that would let me process data in real time? If I follow your list of functions, it sounds like it's not fast enough to process pulses in real time if I'm not interested in frequency domain math.

    Thank you,

    -Bill

  • 0 in reply to William Tandy
    TI__Mastermind 37500 points

    1) the JB49 ADC has a serdes limit of 3.125Msps.  So in order to run at a higher sample rate than 156.25Msps, the JESD parameters change from F=2, to F=1.

    The F=1 JESD mode does not work with the 66AK2L06.   The 14x250 can operate at the 245.76Msps rate, with F=2.

    2) The data capture, is based on the memory size.  The 66AK2L06 demonstration software is not designed to capture .5seconds in the standard software.   As long as the processing of the packets of data can fit in memory, you would modify the standard PktDMA interface to have a destination in DDR memory.  

    122.88Msps complex, 16bit data is 491.52Mbytes /sec.  As long as you can save that amount of storage, the software design would have 2 buffers on to collect data from DFE and IQN and push a packet, the destination is local SRAM (it can't store 1/2 second) usually a 16K or 32Kbyte chunk, 4K or 8K complex samples. 

    The 2nd operation would take the first buffer, and send it to a large file in DDR. 

    There is a planned demonstration release this quarter with the 14x250 ADC.  Here is the first demonstration web section http://www.ti.com/tool/TIDEP0034, the red card in the middle is the DLC card.

    3) In order to work with the 66AK2L06, some clocking solution that involves BOTH the ADC and the 66aK2L06 has to be used.     If you only have the 66AK2L06 and an ADC EVM, they can be directly connected, but then BOTH EVMs have to be referenced to an external clock source.  The DLC card takes the clock from the 66AK2L06 and synchronizes it to another clock solution which can be used by the ADC and DAC EVMS.

    note: although its more complicated to have the 66AK2L06 work with BOTH a DAC and ADC.  It is easier to test without as much external lab equipment. 

    4) You could use the TSW14J56 to store the data today.   Then export the data over USB.   You then have a choice of other interfaces to push the data towards the Keystone processor (USB, Ethernet, PCIe)

    5) The 66AK2L06 is the only Keystone device with DFE and JESD.   If you instead use something like the TSW14J56, you could investigate the other Keystone two devices, that have more DSP cores for processing.  

    Regards,

    Joe QUintal