AFE7950EVM: SPUR mirrored around REF_CLK_HIGH/2

David Hendry

Part Number: AFE7950EVM
Other Parts Discussed in Thread: TSW14J58EVM

Using the TSW14J58EVM DC182_A7 to connect to the AFE7950EVM DC135 REV A

Default bringup scripts for external clocks with both 204B and 204C from TI-AFE79xx-Latte_V2p4p1.exe

I'm seeing a TX SPUR mirrored about the REF_CLK_HIGH/2 frequency. So for 204C I'm getting the results in the table below. Clock source phase noise is less than -55 dBc/Hz 3Hz away.

Are you aware of any settings I can fiddle with to try and remove these SPURS?

Fund (GHz dBm) SPUR (GHz dBm)
1.5 -12 10.3 -21
2.5 -11 9.3 -18
3.5 -5 8.3 -14
4.5 -1.5 7.3 -9
5.5 -2.4 6.3 -6
6.5 -5 5.3 -4
7.5 -6 4.3 -5
8.5 -7 3.3 -15
9.5 -7 2.3 -24

Thanks, David.

over 1 year ago

0 Camilo Garcia Trujillo over 1 year ago

TI__Expert 5035 points

Hi David,

We will set this up on the lab to verify the spurs.

Could you clarify which script you were using? Specifically to know the sampling rate and interpolation factor?

Best,

Camilo

0 David Hendry over 1 year ago in reply to Camilo Garcia Trujillo

Prodigy 30 points

Hi Camilo,

Scripts used are:

S8_ExternalClk_Rx_Tx_500M_FB_disabled.py for 204B, and

J58_64b66b_S8_ExternalClk_Rx_Tx_1000M_FB_disabled.py for 204C.

Using the default settings, which I believe are:

External clock (Fdac): 12 GHz for 204B, 11796.48MHz for 204C.

Interpolation factor (ducFactorTx): 24 for 204B, 12 for 204C.

I've tried many combinations of settings but keep getting similar SPURS.

Thanks for your help,

David.

0 Camilo Garcia Trujillo over 1 year ago in reply to David Hendry

TI__Expert 5035 points

Hi David,

Based on the sampling rates and the fundamental frequencies you are outputting you are looking at your image. This is in no way related to your JESD mode.

For example, with Fdac = 11796.12MHz and an output frequency of 2.5GHz, the image of your fundamental tone in your second Nyquist will be at Fimg = 11796.12-2500 = 9296.48MHz.

In order to deal with these images, you can either frequency plan such that they appear outside your band of interest and then you can use filters to attenuate them if needed. But this is something that you will see on all RF sampling DACs.

For some extra information please refer to the application report below:

https://www.ti.com/lit/an/slaa523a/slaa523a.pdf?ts=1707757975652&ref_url=https%253A%252F%252Fwww.google.com%252F

I will still ask internally to see if there is a way inside the AFE to reduce the level of the image. But I would not be very hopeful.

Best,

Camilo

0 David Hendry over 1 year ago in reply to Camilo Garcia Trujillo

Prodigy 30 points

ok - thanks for your feedback Camilo

0 Camilo Garcia Trujillo over 1 year ago in reply to Camilo Garcia Trujillo

TI__Expert 5035 points

Hi David,

I have heard back, and it has been confirmed that nothing can be done from the device configuration to improve this.

The only options are to frequency plan such that they appear outside your band of interest and then tune your output circuit to filter out the image.

So when looking at some of these cases in the AFE7950EVM, the image might be worse if it is falling on the X-band since that is what the matching circuit is tuned for.

Best,

Camilo

RF & microwave

RF & microwave forum

AFE7950EVM: SPUR mirrored around REF_CLK_HIGH/2