AXM0F343_ PLL/vco issues

Hey,
Im having issues with my vco at lower frequency’s. I was going for 40.68MHz with an external loop filter and an external LVCO (VCO2).

image

My filter values are 82pF, 12K, and 10nF with respect to
https://www.onsemi.com/pub/collateral/and9858-d.pdf

The device only transmits when the VCO calibration bit is turned off, however the carrier frequency varies by up to a 1MHz.

When the VCO calibration is turned on I can see the sine wave lock at 4 times my carrier frequency on the VCO with around 750mV DC offset, however the device has no output.

Not quite sure how to make it lock more constantly at 40.68. (Made three boards and they are all off a slightly different amount).

Hi Alex, please note that the VCO Calibration applies only when using the internal VCO of the AXM0F343. In your case you are using an external VCO so this calibration is not applicable.
In practice, you see no output when enabling the VCO Calibration likely because when running the calibration PLLVCODIV gets re-written thus selecting the internal VCO which PLL has not been previously autoranged.

1 Like

So the slight changes I’m getting in frequency are probably due to slight variations in the LVCO inductor tolerance?

Doesn’t the auto ranging adjust the PLL current to accommodate for the small frequency changes?
Also is the process for adjusting the PLL still the same as
https://www.onsemi.com/pub/collateral/and9858-d.pdf.
It seems as though a few of the registers and names might have changed.

Should I still be getting a clean dc offset sine wave on my external VCO on a lock?
If I turn off the calibration it gets real choppy.

Hi Alex,

So the slight changes I’m getting in frequency are probably due to slight variations in the LVCO inductor tolerance?

Yes this is likely the explanation.

Doesn’t the auto ranging adjust the PLL current to accommodate for the small frequency changes?

Yes it does but note that there are two different calibration routines:

  1. PLL RANGING which is triggered by the PLLRANGING register and its an internal calibration where the PLL locks itself at a desired frequency (with an external or internal VCO).
  2. VCO Calibration as described in the app note you cited, is an optimization of the internal VCO current, where by monitoring the voltage at the FLT pin it is possible to calibrate the best VCOI (in PLLVCOI).

Please read the following post answer where I discuss briefly the differences between the two AX5043 & STM32 Res - #11 by georgi.gorine - Sub-GHz - onsemi Community Forums

These are the specs for the VCO inductor that I am using.
image

What do you recommend to get more accurate frequency across prototypes?
Do you recommend a ferrite inductor or a wire wound one?

The prototypes that are having this issue were samples, is there a tolerance with the internal VCO with external inductor? I ask because I removed the inductor from one of my boards and used it on a different one and it still resulted in the slight frequency discrepancy.

Hi Alex, I don’t have any specific inductor manufacturer/technology to suggest you. Normally both wire-wound and ferrite core inductors are available with high values and high Q factor. What you should focus is in selecting an inductor with low tolerance (±2% vs 5%) making it a better choice in designs like this one, where precise values are very important.

As my colleague wrote in the post FAQ: How to operate AX5043 at 27 MHz - (custom matching networks for AX5043) selecting a single inductor can be a challenge as there is going to be a large board2board variation:

One issue we’ve seen is that unless extremely tight tolerances are used on this tuning inductor, a test bench design at 27 MHz may experience issues in production. This is because the slight variations in inductance place the entire PLL tuning range outside of the desired RF operating frequency band.