phase locking two TSW3065EVM Standalone Local Oscillator Sources

Are you using the output divider of the TRF?

73,

Timothy

Hi,

I am using PA output of each LO. I am using different frequency for each LO, so I cannot use divider.

I am giving +13 dBm to external reference port to lock them to a common 10 MHz reference source.

Please find the attached picture, where I am showing improper locking when using external signal.

when I am using internal 10 MHz clock, I receive signal exactly at 2.1 GHz, which my LO output, however when I use 10 MHz reference from arbitrary waveform generator, I see a shift of 20 KHz.

Any suggestion, help is appreciated!

Thanks,

Meenakshi

Hi Meenakshi

Have you confirmed the accuracy of the 10 MHz reference output by the arbitrary waveform generator? A 10ppm error would result in about 20 kHz error in 2.1 GHz.

Can you capture the LO outputs from 2 or more of the TSW3065 units in a plot similar to the one above? Are the 4 frequencies all 20 kHz high or are there different errors from unit to unit?

Thanks,

Jim B

Hi Jim,

It seems that for different units this shift is similar. following figure shows the results for 2.1 GHz at 100 KHz span.

I guess higher noise is also due to external source.

thanks,

Hi Meenakshi

Here is one thing that you can try. It might provide a solution to your problem.

Configure the first of the 4 TSW3065 units for Internal reference. That will use the clean 10 MHz oscillator built-in. Then configure the other 3 TSW3065 units for external reference. Feed the Int_Ref_Out of the first unit to the Ext_Ref_In of the second unit. Connect each unit down the chain in a similar fashion, so all are using a regenerated (by the CDCV304PW clock buffer) version of the 10 MHz reference from the up-stream device.

I have not had a chance to try this, but it's pretty easy to do and may work well enough to resolve the issues you had getting the LO frequencies to all be the same and be accurate and low noise.

Let me know if this works.

Best regards,

Jim B

Hi Jim,

At first I intended to do the same to avoid the need for external signal generator.

But the signal at int_ref_out of the LO unit has many harmonics, please see the picture below and their power is less.

LO external trigger need power of +13 dBm at external in to lock it to external source.

With filter and additional PA for each int_ref_out , I had doubt that signal int_ref_out will remain clean.

But if you have any comments or suggestion on this, I would like to hear it.

Thanks,

Meenakshi