• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 590 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.

Question about DAC5688EVM : how to make mixing signal

Takumi Suzuki1
Mastermind 8420 points
Other Parts Discussed in Thread: DAC5688, DAC5688EVM

Hello,

I would like to ask you how to set/configure GUI of DAC5688 and "High Speed Data Converter Pro" to output mixing 63.86MHz signal from 102.5MHz(NCO)+38.64MHz(DAC input).

The hardware conditions/settings are as below.

 - 100MHz input to J20 connector of DAC5688EVM

 - J17 connector is connected with J7 connector of TSW1400EVM via cable

 - DAC5688 setting ---- PLL setting  : x8, NCO setting : 102.5MHz(Set : 0x20CCCCC0=550292672)

 - TSW1400 setting ---- DAC input : -38.64MHz (Complex), 800MHz sampling, 2's comp format, 65536 data length

                                      Firmware : "DAC CMOS" down load 

I could not get mixing 63.86MHz out from DAC5688 under these settings, got around 100-110MHz instead of 63.86MHz.

But in case of 100MHz Sampling data not 800MHz, I could get 63.86MHz.

Could you please advise me the reason why I could not succeed in case of 800MHz sampling?

I have attached GUI setting image and register setting file as below.

I would be very happy if you give some advice, thank you.

Best Regards,

Takumi 

 

 

 

2772.100MHzclkwithPLL (2).txt 
Texas Instruments Inc.
DAC5688 EVM Register Configuration

DAC5688 Registers
Address	Data
00		81
01		0B
02		03
03		00
04		19
05		10
06		00
07		00
08		C0
09		CC
0A		CC
0B		20
0C		A6
0D		A6
0E		00
0F		2D
10		00
11		00
12		00
13		00
14		00
15		00
16		AA
17		10
18		80
19		00
1A		0D
1B		FF
1C		00
1D		38
1E		39

CDCM7005 Registers
Address	Data
00		005FF1F0
01		02B282DD
02		D00000A2
03		00000027

  • 0
    TI__Genius 12965 points

    Hi Takumi,

    Everything you describe seems to be correct.  I think the only issue is that there may be some confusion with the data rate vs the DAC output sample rate.

    In your situation where the DAC is configured for 8x interp and the input clock is 100 MHz this implies that the DAC output rate is 800Msps and the input data rate is 100Msps.  The NCO setting is correct, it is calculated at the DAC output rate (800Msps, it will create a 102.4999...MHz mixing frequency.

    The issue is with how the input data is generated.  When you specify 800Msps for the data rate it will create a digital signal at -38.64MHz sampled at 800Msps.  The TSW1400 hardware feeds data to the DAC interface at the DAC input rate (this is the data clock coming from the DAC) - in this case 100Msps.  However the data was created at 800Msps, but it is now being played back at 100Msps, the actual frequency is 8 times smaller (frequency scaled) and you will get a tone at 38.64/8=-4.83MHz.  I would expect your output tone to be at NCO-4.83MHz or about 98MHz.

    If you set the TSW1400 data rate to 100M and recreate the tone, it should be ok.

    Ken.

  • 0 in reply to Ken C
    Mastermind 8420 points

    Hi Ken-san,

    Thank you very much for your kind reply!

    I studied your reply and I could understand what you mean, thank you.

    Let me confirm you one point please as below.

    If the input signal to DAC is 800MSPS sampled, we should prepare -309.12MHz(= -38.64MHz*8) signal due to 100MHz clock.

    Is my idea correct?

    Best Regards,

    Takumi

     

  • 0 in reply to Takumi Suzuki1
    TI__Genius 12965 points

    Takumi-san,

    Yes that would be correct - however to reduce confusion you should generate the data at the input data rate (before the interpolation) so in this case use 100Msps.  In this way if you have any IF frequencies they will still be correct at the final output stage.  For example if you wanted a -38.64MHz IF signal you would generate this with an input sample rate of 100Msps.  In this way you can see that the input IF signal and the output IF signal are the same and do not require some frequency scaling factor.  Interpolation doesn't change the frequency content, it just changes the size of the Nyquist band.  -38.64M at input rate of 100Msps with interpolation of 8x is the same -38.64M at the output rate of 800Msps.

    In general it would be less confusing to keep the signal relative to the input sampling rate and not the output sampling rate.

    Ken.