Crystal trimming and phase noise measurement of CYW4373E

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
cmpxchg_
Level 1
Level 1
First question asked Welcome!

1) trimming

The CYW3273E datasheet mentions the requirement of a (by default) 37.4 MHz crystal or RF source of 20 ppm accuracy, and a comment 'without trimming'.

Is there more documentation on how I should perform trimming of the oscillator in-situ, (i.e. by tuning on-chip additional crystal loading caps) ?

2) phase noise measurements

The CYW3273E datasheet lists the maximum phase noise requirements of the 37.4 MHz crystal, ranging from -129 dBc/Hz to -155 dBc/Hz when using IEEE 802.11ac, 5 GHz modes.

I do not see a buffered clock output among the WRF_* pins to measure the 37.4 MHz phase-noise using external RF equipment, and without loading the oscillator circuit itself. Is there a way to do this @ 37.4 MHz ?

3) RF based trimming / phase noise measurements

Another way of measuring the phasenoise (via the on-chip PLL) would be to correlate with a very long repetitive RF signal generated by a testmode. (This to avoid the complexity/uncertainty of silence-spaced bursts of wifi-sequences of BPSK and OFDM symbols).

The mode would then need to have a limited power to not heat up the VCO too much, but be long enough to get enough samples to measure the phase-noise on both 10 kHz and 100 kHz offset, but then in the 2.4 or 5 GHz range. Then an SDR can be used, validate the SDRs LO using another phase-noise reference, and then determine the phase noise of the crystal.

Does the CYW4373E have such an RF-testmode that can be entered ?

0 Likes
1 Solution
Murali_R
Moderator
Moderator
Moderator
250 sign-ins 250 replies posted 100 solutions authored

1. Typically the frequency offset is done adjusting the external XTAL loading caps. Increasing the loading cap values will shift the offset to the left, while decreasing the loading cap values will shift the offset to the right. This is mostly a trial and error process by adjusting the external loading cap values 1 step at a time. The frequency offset also from board to board because it is also dependent on the PCB layout.

2. We can measure this by simply attaching a SMA connector to the 37.4MHz XTAL pin and measure it on a spectrum analyser that can support phase noise measurement. However, this is not an accurate way to measure the phase noise because attaching a SMA connector can change the impedance of the XTAL. Normally the XTAL vendor should be able to provide the phase noise measurement of the XTAL itself.

3. Measuring the phase noise of the XTAL through the RF signal is not a reliable method, because in the ideal situation, we do not want any XTAL noise to be mixed into the RF signal and we typically have internal filter to supress the XTAL noise as much as possible.

View solution in original post

0 Likes
2 Replies
Murali_R
Moderator
Moderator
Moderator
250 sign-ins 250 replies posted 100 solutions authored

Not sure of this. Will check with the team internally and update soon.

0 Likes
Murali_R
Moderator
Moderator
Moderator
250 sign-ins 250 replies posted 100 solutions authored

1. Typically the frequency offset is done adjusting the external XTAL loading caps. Increasing the loading cap values will shift the offset to the left, while decreasing the loading cap values will shift the offset to the right. This is mostly a trial and error process by adjusting the external loading cap values 1 step at a time. The frequency offset also from board to board because it is also dependent on the PCB layout.

2. We can measure this by simply attaching a SMA connector to the 37.4MHz XTAL pin and measure it on a spectrum analyser that can support phase noise measurement. However, this is not an accurate way to measure the phase noise because attaching a SMA connector can change the impedance of the XTAL. Normally the XTAL vendor should be able to provide the phase noise measurement of the XTAL itself.

3. Measuring the phase noise of the XTAL through the RF signal is not a reliable method, because in the ideal situation, we do not want any XTAL noise to be mixed into the RF signal and we typically have internal filter to supress the XTAL noise as much as possible.

0 Likes