DDS39RFS10: Device behavior

Hi Gil,

1: Yes

2: No

3: It depends on your environment whether a heat sink and fan are required. I think it would be good practice to have some sort of thermal management on your board design.

Regards,

Rob

Thank you Rob.

I want to rephrase my 2nd question, so I am sure you got my concern:

In another RF DAC that we use, DAC38RF82, when you change DACCLK without additional configuration inputs and JESD204 clock updates (SYSREF), the device "hangs" and stops outputting a signal.

This is understandable.

We are using the DDS39RFS10 in DDS mode with the NCO, thus, I want to make sure that after changing the DAC CLK and other than configuring the correct values to get the frequency we need, no additional calibration and/or synchronization processes are required.

thanks,

Gil

Hi Gil,

I understand your concern, the DDS version of the DAC39RFS10 does not have the JESD enabled, so therefore, it won't hang up when you change your sampling clock frequency. I would be best to purchase the EVM and try this out, as well as other features you might have questions about on the lab bench.

Regards,

Rob

Hi Rob,

Thank you.

On a matter of additive PN in the DDS39RFS10 (we have the TRF1108-DAC39RFEVM) and the DDS39RFS12, for Fo=997MHz, 10K offset, datasheet states:

DDS39RFS10

NRZ, 10.24GHz, -143dBc/Hz. Same value in DES2XL mode.

NRZ, 7.5GHz, -142dBc/HZ. Same value in DES2XL mode.

DDS39RFS12

NRZ, 12GHz, -143dBc/Hz.

DES2XL mode, 12GHz -148dBc/Hz, 8GHz -147dBc/Hz.

My questions:

1. Why is the RFS12 more superior than the RFS10 (6dB) in DES2XL mode? -148 in 8GHz versus -142 in 7.5GHz.

2. Why does the additive PN improve as the sampling clock increases? (you can see it in both devices).

3. Can you send the additive PN for the DDS39RFS12 in Fclk=6GHz? (setup as in the datasheet, p.23. DEM/Dither off, NRZ and DES2XL mode).

6GHz is the frequency we intend to use in our design.

4. Figure 8-5 (same on both datasheets) seems to relate to the DDS39RFS12 and not the RFS10.
Do the DEM/Dither have the same effect for both devices, in NRZ and DES2XL modes?

Thank you 

Gil

Hi Gil,

There is not a significant contribution from sampling rate to close-in phase noise.

There is not a significant difference between the RF12 and RF10 DACs, this must be an error in the datasheet. These are the exact same devices, just different speed grades.

DEM/DITHER will have the same effects in NRZ/DES2X modes, but DES2X will have better white noise floor.

I will see if we have any PN at 6GHz FCLK. Stay tuned.

Regards,

Rob

Hi Rob,

So, you say the additive PN of both devices, for the same Fclk, offset, mode, etc., is the same?

Can you find and send the accurate additive PN in 10K offset for these devices (DES2X mode and NRZ mode)?

So, it is not more advisable to use a higher Fclk for better additive PN?

Thanks,

Gol

Hi Gil,

Yes, that is what I am saying for the first question.

I will need to work on the second question. What signal output should we measure the PN for your application? and at 6GHz again?

Please advise.

Regards,

Rob

Hi Rob,

We'd appreciate the additive PN for 6GHz Fclk, NRZ and DESXL modes, 997MHZ Fout, with and without DEM/Dither.

If the RFS12 and RFS10 behave the same in that manner,  a measurement on a device without a RF AMP is preferable (only a balun and a load on the dds output), so we don't see any effect from an external RF AMP.

BTW, do you use a signal generator that has a superior PN in compare to the DDS39RFS, or you do calculate the additive PN such that the external clock contribution is subtracted out?

A question on what you wrote earlier:

You wrote "DEM/DITHER will have the same effects in NRZ/DES2X modes, but DES2X will have better white noise floor:

According to figure 8-5, DEM/Dither have stronger effect in DES mode, but has a similar noise floor (additive PN in 10M and higher offsets).

Thank you

Gil

Hi Rob,

Can you update on the above?

In addition:

Is there a way to run a script on the TRF1108-DAC39RFEVM?

Can we bypass the GUI and control the EVM via Python or MATLAB using the USB-to-SPI interface (via a provided DLL)?

thanks,

Gil

Gil, 

The latest version of the GUI for the DAC39RF12EVM has example scripts in Python to set things up. 

Regards, 

Matt

Hi Rob, Matt,

With regard to the FR bus:

It is powered by the VDDIO bus with typical voltage of 1.8V (1.71V to 1.89V).

I consider using a device that might output a voltage that can reach 2.1V (typical 1.8V but might get  higher).

So, I think of using a VDDIO voltage of 2V (abs max is 2.45V).

And with a 2.1V voltage on FRCLK/FRDI/FRCS I will be on the safe side (abs max for these is VDDIO+0.3V).

Do you see an issue here?

In addition, with VDDIO=2V, will VIH and VIL still answer to 0.7*VDDIO(min) and 0.3*VDDIO(max), respectively?

thanks,

Gil