Post Go back to editing

HMC505 Low Output Power

Category: Hardware
Product Number: HMC505


I am using an AD HMC505 VCO (spec. 6.8-7.4 GHz) as the VCO in a PLL (ADF4108) to generate a LO tone.

The desired tone frequency is at 7.5 GHz.  While this is out of the spec'ced operating range for the VCO, the tuning curves and output power in the datasheet indicate that the device is able to operate at this tone with an output power of ~11 dBm with a tuning voltage of 7V. Originally the HMC532 VCO was spec'ced but unavailable to purchase so the HMC505 was used in place.

Problem Statement

In a test of the VCO (unlocked / PLL disabled), the required tuning voltage to output 7.5 GHz was found to be 9.5V and the output power was ~1 dBm, or ~7 dB lower than minimum from the datasheet.

With the PLL enabled, the ADF4108 is pulling Vtune of the HMC505 to the rail limit of 11V, which results in a tone at ~7.6 GHz.  The output power is ~ -2dBm.  This results in too low power at the RFin of the ADF4108 and it does not lock.

I am concerned by the low output power at the HMC505 output.

Existing Debugging

The measurements are taken at J18 using a Keysight FieldFox SA and conformable coax with ~ -1dB of loss which I have accounted for in the above power measurements.

I have validated the SA using a signal source to check that the measured power is valid.

The power to the VCO is good (3V0) but it is drawing only 73 mA of current, compared to the nominal 80mA at a 3V3 supply.

There is no short from the RF output to GND (in case of a short under the package).   Can you provide any advice on how to proceed - whether this is likely a faulty IC or other potential issues.

Any advice on additional troubleshooting appreciated.


Kind regards, Jonathan  

  • Hello Jonathan,

    Thank you for reaching out. 

    Since you are using the part beyond its specification, we can't guarantee its performance as the Vtune can vary from part to part specially at frequency outside the specification An example of this would be the Vtune shift you found at 7.5GHz resulting to 9.5V.

    As a debug step, could you try to output a frequency that is within the specified operating range of the VCO to see if the part then locks and yield a high power output?

    Kind Regards,


  • Hi Michelle,

    At lower tuning voltages than this, the power output is similar to the values I've descried above in the range of 1-2 dBm.

    Attached below is the frequency vs. tuning curve from the HMC505 datasheet which I was basing operation at 7.5 GHz.  To tune to 7.5 GHz at 9.5V would look to require a pretty serious deviation in the tuning curve from that over the stated 6.8-7.4 GHz operating band.

    Kind regards,


  • Hi,

    The issue seems like it is to do with the tuning voltage curve variation. The datasheet Vtune vs Frequency curves are only typical curves that may be seen. This curve can be change from part to part. The part is still considered to be in spec as long as the the max/min frequency is achieved within the max/min Vtune. 

    Below is an example of typical variation expected from the datasheet curve. 

    As shown in the plot from the datasheet that I’ve annotated, blue intersect at the Vtune and Frequency operating limits. At the intersection of 1V and 6.8GHz , the Vtune curve can provide a typical upper limit (light blue). Similarly, the intersection of 7.4GHz and 11v provides a typical lower limit. But to be clear, these are not guaranteed limits, as Kvco variation does not have defined limits in the test stage. However, we expect parts to be within these boundaries within a certain tolerance level. 



  • Hi Kieran,

    Thanks for your answer here.

    Would it be possible to update the datasheet to reflect this variation?



  • Hi Jonathan,

    Unfortunately for most of the HMC format datasheets, we are unable to make edits unless the specs are incorrect. The Kvco variation can be expected on the majority of our VCO parts unless the Kvco max/min is stated within the specs.