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.

ADS5282 EVM Problems

Other Parts Discussed in Thread: ADS5282, ADS5545, ADS5281

I am trying to connect the ADS5282 evaluation module to the DE2-115 FPGA dev kit.  I followed the install on windows 7 instructions.  I've disabled VCP.  Every time I try to send a command I receive an error (see below).  I've also unplugged other USB serial devices to where only the ADC's comm port is available.  I don't know what else to do.

My test setup is as follows: ADC Input CLK is a 1.5 V signal from the FPGA.  ADC Channel 1 receives a 1.5 Vpp 3 MHz signal from a function generator.  All grounds are connected (Power supply, FPGA, ADC, Function generator).  The FPGA takes in the ADC's output signals and put them through a differential to single ended buffer then to GPIO to be viewed by a logic analyzer.  Nothing happens on the logic analyzer.  

Error 1 occurred at FT_SetBitMode:
Set Bit Bang

Possible reason(s):
LabVIEW:  An input parameter is invalid. For example if the input is a path, the path might contain a character not allowed by the OS such as ? or @.

My OS is Windows 7 x64.

  • Hi,

    Are the instructions you followed the same as are atteched to this posting?   If so, could you see if the SPI GUI is working at all by choosing the powerdown button and seeing if the amount of current from a bench supply falls sharply?  Or does it still report the 'bit bang' error no matter what button you use?

    is there a Windows XP machine available anywhere that you could use to run the TI ADC SPI Interface GUI and make sure that the EVM is responding properly in the first place? 

    If you are using an FPGA to take the high speed serial LVDS and convert to single ended output, keep in mind that at the higher sample rates the FPGA output simply will not be able to switch that fast.  For example, at 50Msps the serial data would be at 600 Mbps and the bit clock at 300 MHz - and an LVCMOS output from the FPGA may not even be toggling at that rate.  But the Frame Clock would be at the 50MHz rate and *that* should show up on a single ended output.

    I seem to remember that the default optin for the ADC sample clock after the device on the EVM powers up is for single ended clock input, and I have to use the SPI GUI to set it for differential clocking before it works with my TSW1400 or TSW1200 capture cards - so if you don't have the SPI GUI working then this default setting for the ADC might be the problem.    If so, then i think we need to get the SPI GUI working.

    Richard P.

    TI ADC SPI 5p2 installation for Windows 7.docx
  • The instructions you attached are the instructions I followed.  The power on the PSU displays 0.06 Amps with the FPGA unconfigured.  When the FPGA is configured, current increases to 0.12-0.14 Amps, and stays there regardless of what button I press.  I do not have easy access to an XP machine.  I do however have a virtual XP machine on this windows 7 laptop.  I will attempt to see if that will work.  In the meantime, what else do you suggest?

  • Tested with an XP VM with control over the comm port.  Same error.  The drivers are the latest from FTDI 2.10.  Current didn't change on the PSU display. 

    :-/

  • Hi,

    I don't think using a virtual or emulated XP environment would be sufficient.  The USB chip on the EVM was programmed at manufacture with some text strings in the USB device eeprom to identify the EVM, and under Windows XP if you were to open Hardware Device Manager to look at how the EVM showed up it would show up as a "TI ADC SPI Interface" and the device description that the Labview GUI source would use to get a handle to the EVM was "USB ADC Interface".  Under Windows7 the EVM is reported to be something different.  I am looking into it again.  I will discuss the 'instructions' document you followed with the guy who wrote it, and I am asking the software team that does our new GUIs to find a way to connect to the old EVMs without having to reprogram the USB chip. I will look into this and report back what i can come up with.

    Regards,

    Richard P.

  • Just to be sure, I went over the instructions again to verify that I didn't miss a step.  Since I've been working with USB debugging, I have a very simple tool to view USB device descriptors.  The chip doesn't show any strings that identify it as an ADC device,  Only "USB <-> Serial".   Could my chip's EEprom be broken or programmed incorrectly?  Maybe the USB View program isn't capturing the right information.

    Either way, I'll be very happy to have an up-to-date GUI that will work on Windows 7 for this device.  It is very important that I move ahead with these projects.  I hope the fix is as easy as enabling a drop down menu to select which virtual port to use.

    I appreciate all of your help. 

    Thank you,

    -Joe

  • Hi,

    I just went back to this EVM on my bench and got it working.   In hardware device manager, I right clicked on the device and chose uninstall (since I had previously installed the wrong driver for it.)  Then I unconnected and reconnected the USB cable.  Then I chose a new driver for it manually, and on my computer I chose the driver under FTDI under USB Controlelrs as in the attached screenshot.  Then the GUI worked just fine.

    It is possible that the USB chip on your EVM is not properly programmed with the text descriptors, although on this EVM if the USB chip wasn't programmed then the EVM should not have passed testing.  Can you get a copy of FTprog from the FTDI web site and use it to scan for FTDI devices attached to the computer and see if the device descriptors show programming?  For example, under manufacturer it would say TI and the device type is ADC SPI Interface.

    If the USB chip still has default prrogramming then we can supply a program that will use the template to program the USB device.  It would be an older program than FTProg - it used to be Mprog that was used and the template we have for prgramming the device is an Mprog template.

    Regards,

    Richard P.

  • Hi,

    I've attached a zip folder of the Mprog utility that we use to program the USB device on the older EVMs.  I renamed some of the file types so that they would not be stripped out as possible malicious code.  exe and dll file types are sometimes striped out of even zip files for protection.  You would have to unzip the contents and then rename the file extensions.  It should be obvious - what should be an .exe is named as .eexxee with every letter doubled and you would just have to change it back to .exe.

    There is a screenshot of the mprog window marked up as to what to do.  It is pretty simple.  First use the folder icon to choose the template.  Almost all the old EVMs use the same template name what was made for first EVM to use this USB chip - the ADS5545 EVM - so choose that template.  Then press the lightning bolt icon to program.  There should be a serial number appear in the window if successful.  That should be it.  We could always program it here if you wanted to ship it back to us, but that would take time.

    In the meantime i am looking into whether some EVMs got out without programming.

    Regards,

    Richard P.

    mprog 3.0a.zip
  • It works!  The EEprom was indeed blank!  If I read the program correctly, it contained nothing but 0xFF's.  I must warn anyone else that reads this, unplug all other equipment.  My LogicPort device runs a similar FTDI chip and it's EEprom could have been wiped.  But yes, the EEprom was empty, I flashed it and rebooted the device, installed the correct drivers, disabled  VCP and it worked.  I was able to send commands without error and watched the power toggle between 0.06 and 0.05 Amps by pressing the power buttons.  So, tomorrow I will begin (trying to) use this ADC.

    Thank you so much for your help!  Problem Solved! 

    -Joe

  • Good.  Thanks for letting me know. 

    We are pulling the rest of these particular EVMs from the stock so we can check them all for programming, and program if necessary and restock them.   But if any others already got out that were not programmed, now the instructions are documented here for anybody who searches the TI forum for it.

    Regards,

    Richard P.

  • If you have found any devices that have a cleared or corrupted EEprom, could you compare the original programing dates with those that have a correct EEprom?  I'm interested in the shelf life of the EEprom.  

  • Hi,

    I have never actually caught any devices that have *been* programmed that lost the programming - all the ones that I could track down the actual chain of events were ones that we simply skipped the progrmming step in preparing the EVM for one reason or another.  But I'll check when these come back from stock that they have the default FTDI factory programming.

    Regards,

    Richard P. 

  • Hi, Richard.

    I did all procedures posted here.
    But still our EVM (ADS5281) got out of device manager list on my desktop.

    Anyway, we ordered a new same EVM.
    Here is a test setup list.
    1. Power up ADC EVM with 5 V supply and its ground.
    2. Connect the EVM to desktop (Win 7, 64 bit)
    3. Install driver (D2XX from FTDI) and SPI software
    4. Reset ADC EVM through SPI software
    5. Feed ADC input clock (50MHz) from FPGA (Virtex5 SX95T) to ADC EVM
    6. Connect input analog signal to ADC EVM
    7. Capture ADC signals (bit & frame clocks, data) on chipscope

    Please look at it and give some comments on how to start to use correctly the EVM.
    It would be great to get SPI 5p2 software installer and verilog code for ADC interface.
    My email is smeekim@kaeri.re.kr.

    Best regards,

    Soo Mee.