Out of curiosity, the MAX2769 EV Kit uses a 16.368 MHz Rakon TCXO. What would happen if a 16.369 MHz oscillator were used? I realize it would shift the IF, but it would still be functional, wouldn't it?


MAX2769 and MAX2769B can accommodate any reference frequency between 8 and 44 MHz. A PLL with 12 fractional bits, the sample rate and the IF bandpass filter are all configurable via SPI and you can realize a wide range of frequency plans. See datasheet.

16.368 MHz gives a very simple frequency plan, you can run the PLL in integer mode and sample at only 8MS/s, lowering computational load in the correlators.

Different frequency plans allow you to reduce beating from quantization noise, on the other hand they may introduce spurs in the IF from the fractional PLL. If you want to squeeze out the last decibel, I would try to avoid the prime factor 1023 in XTALS and go for a more elaborate setup.

Just to make sure I answer to your precise question: IF would be shifted by 96kHz, the complex bandpass would still let this pass. Taking the altered sample rate into consideration, the correlators would (for doppler 0Hz) see the phasor advance by 1.5338 rad per sample instead of by \$ \pi / 2 \$ per sample. This would be well outside the expected doppler range for ground based receivers and --- depending on software --- you would simply not acquire the signal ... although it is possible to receive it with software modifications.

  • \$\begingroup\$ I think what I'm asking is that if 1575.42 / 16.368 = 96.25, is the point to come out at effecitively zero IF? Would using 16.369 produce largely the same effect, or is 96.24412 sufficiently different such that it's a different animal altogether? What would be an ideal off-harmonic frequency plan to use? \$\endgroup\$
    – KyleG
    Jul 26 '20 at 2:28
  • \$\begingroup\$ @KyleG Zero IF is a bad idea with a 4-bit A/D. Remember you get a rounding error for every sample and for zero IF, this error will also have low frequency. This is what I call beating. For a 16.368 XTAL, I use RDIV=4, MDIV=191.453125 giving IF=8,567,625 Hz. I sample at 32MS/s and set the BPF to 8.58 MHz . This suppresses quantization noise by 16.7 dB. The reasoning behind this does not fit in a comment. You may ask: "How do I choose IF for a GPS-receiver" \$\endgroup\$
    – Andreas
    Jul 26 '20 at 9:08
  • \$\begingroup\$ This answer is fascinating. I have asked follow-up questions here and here. \$\endgroup\$
    – Damien
    Apr 4 '21 at 19:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.