DAC39RF10: Spurious : 8388.608MHz clock output from LMX2615 to DAC39RF10

BH KIM

Part Number: DAC39RF10
Other Parts Discussed in Thread: ADC12DJ3200

Tool/software:

We are using LMX2615 for DAC39RF10 and ADC12DJ3200.

LMX2615's Output clocks are 8388.608MHz for DAC and 2097.152MHz for ADC.

At low temperatures around 0 degrees Celsius, there is below issue.
Our board has the spurious about 35dBc at +-2.08MHz from fundamental frequency of DAC. Of course, if DAC output frequency is changed, the spurious occurs around +/- 2.08MHz at that changed frequency.

We have two similar boards with DAC39RF10, and only one of them generates the spurious at low temperature.

Also, We measured the 8388.608MHz between LMX2615 and DAC39RF10.
There are the spurious about -80dBm at +-2.08MHz from fundamental frequency(8388.608MHz)
We controlled the power down of DAC39RF10(controlled by SPI : DAC Register 0x0002 0x03), then the spurious between LMX2615 and DAC39RF10 has been eliminated.

We are digging the DAC's input clock buffer power(VDDCLK10), core clock power power(VDDCLK18) noise and also pull up matching circuit between LMX2615 and DAC39RF10 for the spurious issue debugging.

Do you have any clue for this issue?

please refer the below schematic.

1 month ago

0 BH KIM 28 days ago

Prodigy 155 points

Spectrum between LMX2615 and DAC39RF10

(One of the differential lines is connected to the Spectrum Analyzer, and the other differential line path is connected between LMX2615 and DAC39RF10)

(at Ambient Temperature)

(Spurious (+/- 2.08MHz) at low Temperature)

0 Chase W 21 days ago in reply to BH KIM

TI__Mastermind 21415 points

Hello,

I would second looking into the clock rail supplies for the DAC. Is there a 2MHz/2.08MHz DC/DC switching frequency upstream of the 1V and 1.8V LDO for these DAC CLK supplies?

Thanks

0 BH KIM 21 days ago in reply to Chase W

Prodigy 155 points

Hello,

We use SVLTxxx to generate 5V from 50V and SVGAxxxx to generate 2.5V from 5V.
The switching frequency is 435~575kHz and 200~300kHz, respectively.

We used TPS7Hxxxx to generate 1.8V and 1.0V from 2.5V.

We also tested by inputting 5V from the power supply and there was no improvement.

Thank you.

0 Matthew Kramer 8 days ago in reply to BH KIM

TI__Intellectual 2550 points

Hey BH,

Just want to make sure I understand but you see a spur on the clock input going into the DAC? The DAC will mix any spurs that are on its clock input. We may need to loop in CTS to better understand where these spurs are coming from if they are originating from the clock.

Regards,

Matt

0 BH KIM 8 days ago in reply to Matthew Kramer

Prodigy 155 points

I posted the Spectrum between LMX2615 and DAC39RF10 20 days ago. The spurious showed up at cold temperature, but was not present at ambient temperature.

We found about -60dBm spurious at 2.08MHz of DAC Power supplies including VDDCLK18 and VDDCLK10.

As I described DAC output has approximately -35dBc relative to the fundamental at cold temperature, but has very low level spurious at ambient temperature.