Hi Guys,

I have a little question concerning the ADF5355 PLL. I used ADIsimmPLL to calculate my design, but I think there is something wrong, or at least I have a discrepancy with the programming of Register 6:

I want the PLL to produce a 8,5714 GHz LO for the HMC6300/01 . Therefore I use a very linear quarzt oscilattor of 20 MHz as a RefIN and use fPFD = 20 MHz.

ADIsimPLL gives me the follwoing programming assistance:

Ref Divider R Counter = 1

Main Divider INT = 214

Fractional Value FRAC = 4793490

Fractional Value FRAC2 = 571

This results in an output frequency of RFAout = 4,28571...GHz --> so RFBout = 8,5714..GHz

One thing that gives me a questionmark, is that ADIsimmPLL doesn't use the VCO divider allthough I am in the range of 3,4 GHz - 6,8 GHz . So for Register 6 programming, I program D13 = 1 . D12 is also 1.

Now with setting the CPbleed current I am mixed up. The equation states that :

4/N < Ibleed/Icp <10/N ;

Now N = 214 ,right? But then I never can fullfill the equation, because in ADIsimPLL I have 5mA of Icp. Even if I use the max. Ibleed, I am outside the suggested range of the equation.

Did I make a big mistake here, or what might be the problem?

best regards

Benjamin

Hi there,

The easiest way to validate your register writes is to use the ADF4355 GUI. For 8.5714GHz at RFoutB, a signal at 4.2857GHz will be fed back to the PFD (internally) - REG6 [DB24] should be set for fundamental feedback unless the output divider is being used. For fractional frequencies the algorithm used by the GUI will always choose a value that is closest to 4 * (Icp / N), I recommend that you use this equation for determining the charge pump offset current.

Generally we don't recommend use of high Icp values with the ADF5355 due to degraded spurious performance. Admittedly this is unique to this family of integrated VCO / PLL products. A good charge pump current value to start with for the ADF5355 is 0.9mA and is acceptable for most applications. For your case, if you set Icp = 0.9mA you will have 4 * (0.9mA / 214) = 16.8uA. Dividing 16.8uA / 3.75uA = 4.48 which rounds to a setting of "4" which matches the GUI calculation. Our 16.8uA is just short of the range recommended by the "optimal bleed setting" as recommended by the datasheet (4/N < Ibleed / Icp < 10/N) but is very close. Conversely, notice that when Icp = 2.5mA (max Icp) it equals the upper end of the range. I would consider this range as general guidance but use 4 * (Icp / N) for my actual setting.

Best Regards,

Marty