• State Resolved
  • Locked Locked
  • Replies 10 replies
  • Subscribers 45 subscribers
  • Views 2350 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

AFE5809EVM: Dissatisfaction: AFE5809EVM + TSW1400EVM not working after 3 months of try

Sajad Rahmdel
Expert 1705 points
Part Number: AFE5809EVM
Other Parts Discussed in Thread: AFE5809, , ALLIGATOR

Dear TI Employees and future customers,

I'm Sajad Rahmdel, I'm a master's student at Korea Maritime University.
I'm writing to you to inform you how dissatisfied I am from using TI products. Hopefully, you could find the cause of this problem in your company and improve your customer satisfaction in the future.
Our research center, AITASC at Korea Maritime University, has recently purchased an AFE5809 Evaluation Module and TSW1400 Evaluation Module from TI.
Sadly, we have been struggling to start using these devices. I'm going to break down the issue:
1. The AFE5809EVM and TSW1400EVM user manuals are NOT up-to-date. This causes a lot of confusion because some of the data provided in these user manuals don't exist anymore, or some new functionalities are not explained.
2. The user manual is very brief, not enough to fully understand the device. The user manual is just a few pages.
3. Despite having this issue for months and creating 7 threads (as shown in the links below), TI employees could not solve the issue. How well-trained are your employees? Why can't they properly debug a simple issue (I'm just trying to start using these devices with factory default configuration).
4. The provided software (AFE5809 EVM GUI and high speed data converter pro) are not working properly. We get error despite following exact process explained in the user manual. There is no explanation why we are getting these errors, and TI employees failed to solve the issue. 
List of threads regarding this issue:
I never expected a well-reputed company, such as TI, would provide such a poor experience for us. I hope to hear a satisfactory respond from you.

Thank you.
Regards,
Sajad Rahmdel
  • 0
    TI__Expert 4145 points

    Hi Sajad,

    We apologize for the delay in getting a complete resolution regarding the issues you are seeing with AFE5809EVM and TSW1400. 

    1. Besides the HSDC Pro disparities due to the latest GUI software updates, what other functionalities or portions of the EVM's user guide do you believe are outdated?

    2. The user guide is over 60 pages and is deliberately written to help with just initial device setup. In order to fully understand the device, you will have to read the datasheet also.

    3. We have had a lot of back and forth with you over the last couple of months regarding this issue and even replaced a $600+ TSW1400 board with the hopes of helping you resolve this problem. 

    4. I believe there is some disconnect somewhere concerning this  issue you are currently seeing where no clocks are provided by the AFE5809EVM. The problem either lies with the GUI installation or the AFE5809EVM. As we have had many customers successfully order the EVMs, install our GUI(from the TI website) and configure the AFE5809EVM with no issues, I am inclined to believe the problem lies with your computer system, user error or the AFE5809EVM unit you currently have in your possession.

    Is there anything special about your environment/handling of these boards that could provide some illumination? 

    Sincerely,

    Olu

  • 0 in reply to Olu Sodipe
    Expert 1705 points
    Dear Olu,
    1. Please check page 7, AFE5809_09_14b doesn't exit. Figure 6 is outdated. Figure 7 outdated. Figure 8 outdated.
    2. The user guide is over 60 pages, right. However, about 40 pages are really useful. I didn't count the terms of use attached at the end of every document. My personal suggestion is that if the user guide elaborates in more details, it would help us better understand the device and avoid making mistakes.
    3. I truly appreciate the support team. Despite all the efforts, the problem is not solved yet.
    4. The boards are handled with extra care. I have even screwed the boards to a wooden plate to avoid any static electrics. I have tested the device on 3 different PCs using Win 10, Win 8.1, and Win 7 (all 64bit), so I don't suspect the problem is with our operating system or the PCs.
    The first post on below thread explains my environment is details, please check it out:
    e2e.ti.com/.../563722

    Thank you and looking forward to hearing from you.
  • 0 in reply to Sajad Rahmdel
    TI__Expert 4145 points

    Sajad,

    1. Figures 6, 7, and 8 are different due to a recent HSDC Pro GUI software update--the user guide will be modified to reflect this soon.

    Please try to keep all your inquiries in one thread so they can be answered in one post.

    Regarding the AFE5809EVM's power supply, the picture shows the required voltages of +/-5V being provided to the board--the color of the wires has no significance.

    DCLK and FCLK are generated by the AFE5809 device and can only be probed at pins P10-58 and P10-66 as stated earlier.

    Your post about having to swap the oscilloscope probe's polarity to see the 40MHz signal is definitely not expected behavior for the AFE5809EVM plus FCLK's magnitude is too low at ~350mV pk-pk.

    Based on your responses so far, I would like you to check the power draw for the AFE5809EVM--disconnect all other devices connected to the power source you are using and measure the current draw of the EVM for the following cases;

    • At startup before connecting the USB cable to your PC.
    • After startup and connecting the USB cable but before opening the AFE5809EVM GUI.
    • Immediately after opening the GUI, during device initialization, note peak current draw.
    • Post-initialization SPI commands by GUI note steady state current draw.

    Sincerely,

    Olu

  • 0 in reply to Olu Sodipe
    Expert 1705 points

    Olu,

    Thank you for your prompt reply.

    I'll post all the issues here, as you suggested.

    I've disconnected all devices to our power, which provides +5, ground, and -5 voltages to  the AFE5809EVM. The following image shows our power supply. I have already tested the power input using a multi-meter. The power supply provides is +5 and -5 volts with 3A and 0.5A currents, respectively.

    We did current flow test and here is the result:

    Stage ground to plus (Amp) ground and negative (Amp)
    1 0.351 0.06
    2 0.421 0.06
    3 0.63 0.06
    4 0.63 0.06

    Stage 1: AFE5809EVM connected to power (before connecting to USB). 

    Stage 2: After USB connected to the board (before opening GUI).

    Stage 3: Immediately after Opening GUI

    Stage 4: A few seconds after opening GUI

    Note: all above measurements are done by Multi-meter.

    As you see in the table above, the negative amplitude is pretty low (60mA). We are sure our power supply provides 0.5Amp to the negative voltage.

    Side note: I though this might help.

    The following image shows the 40Mhz sampling frequency generated by the crystal. This signal is obtained by checking the JP9 Pins 3(positive) and 4(ground). As you can see the signal has 2.3 volts peak-to-peak (which is good I suppose). However, the FCLK(P10-66) pk-pk voltage was 330mV, which was pretty low. 

    Looking forward to hearing from you :)

  • 0 in reply to Sajad Rahmdel
    TI__Expert 4145 points
    Sajad,

    That again is unexpected behavior. Can you also check the current draw when you use the AFE5809GUI to reset the board?
    Finally, after step 4, leave the AFE5809EVM on with the USB connected but close the GUI--what is he current draw now?

    When you press reset on the GUI, is there any activity at the bottom of the GUI?

    Sincerely,
    Olu
  • 0 in reply to Olu Sodipe
    Expert 1705 points

    Olu,

    Thank you again for helping me debug this issue.

    Here is the updated table:

    stage positive(A) negative(A)
    1 0.361 0.047
    2 0.584 0.047
    3 0.71 0.047
    4 0.641 0.047
    5 0.74 0.047
    6 0.641 0.047

    Note: all data are newly obtained, including the data from stage 1 to 4

    Stage 1: AFE5809EVM connected to power (before connecting to USB).

    Stage 2: After USB connected to the board (before opening GUI).

    Stage 3: Immediately after Opening GUI

    Stage 4: A few seconds after opening GUI

    Stage 5: Use the AFE5809GUI to reset the board

    Stage 6: leave the AFE5809EVM on with the USB connected but close the GUI

    When you press reset on the GUI, I the text on the button of the GUI changing rapidly and the final text is "Ready for input command"

    Thanks again :)

  • 0 in reply to Sajad Rahmdel
    TI__Expert 4145 points

    Hi Sajad,

    That looks more like it--the current draw is within expected limits and seems like the GUI is communicating with the AFE5809EVM just fine.

    Which means that the AFE5809 is receiving the 40MHz clock and able to process SPI commands. The device should also be generating DCLK and FCLK.

    How exactly are you probing the P10-66 and P10-58 pins again? Are you using a naked probe or an attachment+probe?

    Sincerely,

    Olu

  • 0 in reply to Olu Sodipe
    Expert 1705 points

    Dear Olu,

    The following photos show how I probe the P10-66 and P10-58 pins.

    I'm using alligator probe with a piece of cable. I connected the red cable of probe to P10-58 and the black one to ground (P10-60). Another time, I also tried red to P10-58 (DCLKP) and black to P10-56 (DCLKM). In addition, I probled FCLK by connecting Red cable to P10-66 (FCLKP) and black to P10-60 (ground). 

    Note: Previously I mentioned that I saw FCLK only by changing the polarity. Because you said that was strange behavior from AFE5809EVM, I did the experiment again. I noticed that I can see the FCLK with and without changing the polarity(changing red with black cable) . The only difference was that the voltage becomes negative after changing the polarity.

     

    Here is another photo:

  • 0 in reply to Sajad Rahmdel
    TI__Expert 4145 points

    Hi Sajad,

    Are you using BNC-to-alligator cables there? Please use a naked voltage probe to avoid loading the clock source (probes generally have lower capacitance).

    Also, use one of the GNDx test points on the EVM for your negative connection. Now that you can confirm that you are seeing the proper FCLK (what amplitude?), you should be able to see DCLK also--the AFE produces both. 

    Also, please try to mate the AFE5809EVM to the TSW1400 and follow the user guide's instructions on configuration and attempting a capture now that you have resolved the GUI problems.

    Sincerely,

    Olu

  • 0 in reply to Olu Sodipe
    Expert 1705 points
    Dear Future User and TI Employees,
    Well, we finally found the solution, thanks to Mr. Ken C (e2e.ti.com/.../42894).
    If you are receiving any of the following errors, continue with the solution provided bellow:
    Errors:
    1. "Read DDR to file TIMED_OUT_ERROR"
    2. "no firmware please select a device to load firmware into the board"
    3. Any error during the initialization of TSW1400EVM

    Solution:
    During the initialization of TSW1400EVM, the FPGA requires at least 3-4A of current with exactly 5 volts (according to Mr. SM (e2e.ti.com/.../3387639), they got this error because they were using 4.9 volts instead of 5 volts). If this current is not provided, the voltage will drop and the TSW1400EVM will reset itself. This will cause the above mentioned errors.
    We reccomend that the user uses two separate power switches, one for the TSW1400EVM and the other for the board being attached to TSW1400EVM.

    Suggestions for TI employees:
    1. Please clearly mention in the user's guide that TSW1400EVM will not work properly if the current is less than 3A. It was unexpected for us to see that this board requires such a high current. We would never expect that the issue could be caused by lack of current.