• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 47 subscribers
  • Views 1792 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.

DAC38RF82: Can't modulate a simple tone generated in HSDC Pro GUI (running TSW14J56EVM) using the DAC38RF82EVM

Steve Krupa
Intellectual 300 points
Part Number: DAC38RF82

Hi Jim Seton and/ or Eben Dwobeng:

I'm having a problem very similar to that reported by Ivan Komsic on 25 April 2017.  I'm following the example contained in the DAC38RF8x User's Guide (slau671a.pdf), except that I'm trying to used the on-board VCXO (running at 122.88 MHz). 

I'm running:

>Windows 10 (region settings = United States, Language=English)

>Labview RTE 2014 SP1

>RF38RF82EVM GUI

>HSDC Pro Gui (w/ a TSW14J56EVM board)

>2 separate power supplies (each capable of putting out 3 Amps or more)

>all the app notes associated with the TI Hardware and software GUIs mentioned above.

Here's what I *CAN* do:

>Set up the DAC38RF82EVM using the instructions in the User's Guide and the DAC GUI interface

>Output a constant sinusoid tone using the Constant Input mode (3FFF) and the NCO Oscillator

....HOWEVER:  the frequency of the tone seems a bit off (using a Signal Hound spectrum analyzer)  AND when I ONLY use the Mixer (without NCO) I get a series of tones separated by ~ 400 MHz (just as Ivan Komsic described)

The DAC clock is set to 6144 MHz, just like in the User's Guide example.  In using the 122.88 MHz VCXO, I have enabled the PLL (on  the overview screen of  the DAC38RF82 GUI) and set M=25, N=2 (giving the 6144 MHz DAC clock rate used in DAC38RF8x User's Guide).  With some  effort, I can tune the PLL to get a PLL LF voltage of 3 or 4, by manually changing the M and N divider values on the "Clocking" tab of the DAC38RF8xEVM GUI.

It seems like I am able to pass information from the TSW14J56EVM board to the DAC38RF82EVM board (no persistent bus errors when I hit "Send" in the HSDC Pro GUI), but I am not able to get the Tx CLK LED to flash (indicator D2, on the TSW14J56EVM board).  

If someone could provide some instructions from how to move from the (successful) Constant Input/ NCO  tone generation to the (unsuccessful) mixing of a pattern generated by the HSDC Pro GUI/ TSW14J56EVM, I would be extremely grateful.   We are trying to evaluate both the DAC38RF82  and a high speed ADC board (maybe the ADC2DJ3200) for a future product, but the difficulties we've encountered so far have made this process very time consuming.  Any help you could provide is greatly appreciated.

Thanks,

Steve Krupa

  • 0
    TI__Intellectual 2750 points
    Steve, I have sent this over to Jim to see if he can help
    Regards,
    Brian
  • 0 in reply to Brian Kincaid
    TI__Guru* 91947 points

    Steve,

    I got your setup working using an external 10MHz reference clock to lock the on-board VCXO to 122.88MHz. Give this a try (see attachment).

    Regards,

    Jim

    DAC38RF82_6144M_PLL_3072M_ref_841.pptx

  • 0 in reply to jim s
    Intellectual 300 points

    Hi Jim,

    Much thanks for the prompt response, and the screenshots of your setup in the PowerPoint presentation you provided.

    I applied a 10 MHz clock signal (@ -12 dBm) to SMA connector J4 on the DAC38RF82EVM board, and tried to duplicate your setup by making the needed changes in the DAC38RF8x EVM and HSDCpro GUIs.  On the positive side, this was the first time I was able to get the PLL AutoTune function in the DAC GUI to work (bringing the PLL LF Voltage on the "Clocking" tab up to 4).  This was also the first time I was able to connect the TSW14J56EVM board to the DAC board with a confirmed synchronization:  LED D2/ TX CLK on TSW14J56EVM blinking after "Send"ing the 300 MHz tone pattern, with LED D1/ TX SYNC~ remaining UNlit.  This is good progress.

    On the negative side, the output of the DAC board seems very low, with signals typically in the neighbourhood of -60 to -30 dBm.  I observed this with both the UNmixed 300 MHz tone, using a 1 MHz NCO mixer frequency, and a 2500 MHz NCO mixer frequency.  Please see attached PowerPoint presentation for my DAC mixer settings and the output from our spectrum analyzer (SignalHound USB-SA124B).  There are also some unwanted spurs which come up, at around the same level as the desired output (mixed 300 MHz tone).  Do you have any idea what might be causing the low output levels, or the unwanted spurs?  With your particular setup, the TSW14J56EVM board is drawing around 2.3 Amps (from 5A supply)  and the DAC38RF82EVM board is drawing about 1.65 Amps of current (from a 3A supply).  Do these numbers seem about right, or too low?

    We ultimately want to load an External Pattern file and mix it with the DAC.  It will be a high bandwidth signal, mixed with a carrier frequency between 2 and 4.5 GHz.   The wideband nature of our signal makes us nervous about the spurs, and the low output level is going to be very problematic.    I would like to solve these issues before we worry about moving forward with the External Pattern file.

    I'm grateful for the help you provided, and look forward to working with you on the issues cited above.

    Much Thanks,

    SteveMixingTrials_DAC38RF82EVM__TSW14J56EVM__VCXO122p88MHz_10MHzRefClk_JSexample_V0.pptx

  • 0 in reply to Steve Krupa
    TI__Guru* 91947 points

    Steve,

    I think you may have missed a setting. Please double check every parameter shown in my screen shots. I have also saved all of the register settings to the file attached. Verify that the D4 and D3 are illuminated on the DAC EVM indicating both LMK PLL's are locked. If you go the Clock Output tab of the DAC GUI and disable the Group Powerdown on CLKout 4 and 5, and set the DCLK divider to 10, you should now see a 307.2MHz tone on SMA J24. This is a must before doing anything else. If you are still having problems, send me screen shots of every GUI page you configured for both boards.

    Regards,

    Jim    

    6144MSPS_8x_10M_ref_LMF_841.cfg

  • 0 in reply to jim s
    Intellectual 300 points

    Hi Jim,

    I used the .cfg file you provided for the DAC, applied a 10 MHz clock signal (6dBm output level) to SMA J4, went through all the GUI panels to ensure that both LMK PLL's got locked (DAC LEDs D3 & D4 lit),  detected the 307.2 MHz tone on SMA J24...  and got a clean, single tone signal on SMA J7 (IOUTA_SE) at 301 MHz, output level around -40 dBm.  When I turned off the NCO mixer for DAC A and used the NCO mixer for DAC B, I saw the signal level jump to ~ -4 dBM (looking at output from SMA J2 (IOUTB_SE).   The output level from DAC B is pretty much the same as what you showed in your .PPT document, so I'm really happy about that.  But is there supposed to be such a huge difference in output levels for DAC A (IOUTA_SE) and DAC B (IOUTB_SE)?  Or do you think a component in the terminal output chain for DAC A might be damaged?  

    I loaded an example pattern file from the HSDC Pro "Test Files" directory, and verified that I saw the signal (~ 30 MHz BW) centered about the frequency I selected for the NCO mixer (2.5 GHz).  Happily, that part was easy.  We are going to load some of our own external wideband patterns into the HSDC Pro software:  what kind of output level (on J2/ IOUTB_SE) should I expect for a signal with a flat frequency response, say 500 MHz of bandwidth?  Is there any way to control the (maximum) output level on J2 (IOUTB_SE) besides playing around with the mixer gain (0 or 6 dB) and/ or the GainAB_EN field ( both on the DAC38RF8x>Digital(DAC B) sub-panel )?

    Thanks so much for all your great help and detailed troubleshooting recommendations in getting the DAC and the TSW EVM's to output a clean tone with decent signal strength (above -10 dBm).  I'll let you know how things go when we  load our own patterns, and try to fire the DAC with some kind of external triggering.

    Best Regards,

    Steve Krupa

  • 0 in reply to Steve Krupa
    TI__Guru* 91947 points
    Steve,

    Use SMA J6 for CHA out.

    Regards,

    Jim
  • 0 in reply to jim s
    Intellectual 300 points
    Hi Jim,

    That did the trick. Thanks for getting this week started on the right foot.

    Cheers,

    -steve