Fractional PLL selection, ADF41513SD2Z vs ADF41513SD3Z vs ADF4372

Our newer frac N PLLs, like the ADF41513, are capable of sub-hertz resolution in fractional mode even with high PFD frequencies. You shouldn't require any hardware modifications to the eval board to accomplish this. We also have another version of the ADF41513 board, SD3Z, which uses the ADF5709 VCO.

The one thing I would say is that with the SD2Z board you are using a wideband 10 - 20GHz VCO (HMC733) but you would only using a portion of the frequency range. 

What about the ADF4372? It does use an internal x2 stage to get the 10-12GHz frequency which might be a concern for you, but it has an inbuilt tracking filter to help minimise harmonics. The ADF4372 has an integrated VCO and uses a passive loop filter and you can get multiple SE or differential outputs from it so it might be better as a bench LO generator.

Thank you for your reply Andrews!

I would have a few follow up questions about these three boards.

1. What are the arguments for choosing the board with the ADF5709 VCO over the HMC733 VCO?
   I have the feeling that the HMC has slightly better performances overall compared to the ADF (output power, SSB phase noise).
   1. Are the voltage regulators (LT3045/LT3042) included with the ADF5709 to compensate its larger frequency pushing? How do the performances compare in the end?
2. Regarding phase noise, the datasheet seem to indicate that the ADF4372 board would be the obvious choice. Is that right?

Thank you for your help.

The main reason for the the ADF5709 was to fix an issue we found post-release on the HMC733 (detailed here and here)

A side-effect of fix was a minor hit taken on the PN as you mentioned. The change to the output buffer also meant lower output power in general but the ADF5709 does have a flatter response over frequency. For an application where you know the harmonic issue detailed above will not affect you, it's probably better to use HMC733. This is why we ended up keeping both parts.

The ADF41513 could be susceptible to the harmonic issue so it made sense to respin the board with the ADF5709. Rather than just copying the SD2Z design completely I decided to also make a minor change to the LDO setup as I've had good experiences with the LT3045/LT3042 regulators and I wanted to test what the performance looked like with a reduced LDO count. The SD2Z had the supplies paired, whereas on the SD3Z I separated them into what I considered noisier supplies and supplies more susceptible to noise. So I was predominately looking at the PLL performance with the new LDOs rather than anything related to the VCO.
In my tests I found I was getting a modest in-band PN performance increase with the SD3Z board and the new regulator setup. I've attached a plot with 2 x SD2Z boards that both had the VCO replaced with an ADF5709 vs 2 x ADF5709 boards.

We plan on continuing production of both SD2Z and SD3Z boards.

The PLL portion of the ADF4372 and the ADF41513 are quite similar they both have all the same features. The choice between them will mostly come down to the usual considerations of standalone PLL vs PLL with integrated VCO (e.g. wideband VCO vs multi band, spurious performance, footprint, supply requirements). I think for what you need it for the ADF4372 would be a good choice.

Regards,

Alex

The main reason for the the ADF5709 was to fix an issue we found post-release on the HMC733 (detailed here and here)

A side-effect of fix was a minor hit taken on the PN as you mentioned. The change to the output buffer also meant lower output power in general but the ADF5709 does have a flatter response over frequency. For an application where you know the harmonic issue detailed above will not affect you, it's probably better to use HMC733. This is why we ended up keeping both parts.

The ADF41513 could be susceptible to the harmonic issue so it made sense to respin the board with the ADF5709. Rather than just copying the SD2Z design completely I decided to also make a minor change to the LDO setup as I've had good experiences with the LT3045/LT3042 regulators and I wanted to test what the performance looked like with a reduced LDO count. The SD2Z had the supplies paired, whereas on the SD3Z I separated them into what I considered noisier supplies and supplies more susceptible to noise. So I was predominately looking at the PLL performance with the new LDOs rather than anything related to the VCO.
In my tests I found I was getting a modest in-band PN performance increase with the SD3Z board and the new regulator setup. I've attached a plot with 2 x SD2Z boards that both had the VCO replaced with an ADF5709 vs 2 x ADF5709 boards.

We plan on continuing production of both SD2Z and SD3Z boards.

The PLL portion of the ADF4372 and the ADF41513 are quite similar they both have all the same features. The choice between them will mostly come down to the usual considerations of standalone PLL vs PLL with integrated VCO (e.g. wideband VCO vs multi band, spurious performance, footprint, supply requirements). I think for what you need it for the ADF4372 would be a good choice.

Regards,

Alex

Thank you very much for this answer. We received the ADF4372 eval board a few days ago and we are happy with the performances.