Modeling Clock Filters

[NZ0I]

Below is a screenshot of the TINA modeling tool analyzing a revised filter design for the ~10 MHz BFO. It shows the schematic being analyzed, a Bode plot of the filter response, and an oscilloscope trace showing the input signal (a square wave) and the filtered signal going into pin 10 of the SA605 (sine wave). You can see that the ~2V pp square wave coming out of the Si5351 gets attenuated and filtered into a sinusoid with an amplitude of ~350 mV pp. 

The modeling revealed an error in my earlier calculation of the attenuator R values. We will probably want more attenuation than I previously calculated. Note, however, that all the modeling is approximate. Using the previous attenuator R values should still be a big improvement over feeding the LO and BFO signals directly into the SA605 inputs. But we shouldn't be surprised if we find that we still need to knock the signal down more than previously thought.

Here are the assumptions under which these resistor attenuator networks were calculated:

Si5351 Zout = 50Ω, Vout = 1.5Vrms (2.1V p-p)

SA605 Zin = 10kΩ, Vin = 0.2 Vrms (0.28V p-p)

In the Rev B.4 schematic the attenuator values are as follows:

R203 = R204 = 51Ω

R205 = R206 = 34.8kΩ

R207 = R208 = 14kΩ

The values are not critical. Just having decoupling between the SA605 and the oscillator might be all that is needed in order to eliminate much of the de-sense seen in earlier testing/RSSI values. Leaving out the L and C filter components also should have minimal impact... but cleaning up the clock signals might result in nominal improvements to an otherwise sensitive receiver, since it should reduce some of the higher-order mixer products. So I'll put pads for them into the new board design so we can easily experiment.

In the Rev B.4 schematic the LC values are as follows:

133 MHz: L202 = 15 nH; C209 = 30pF

10 MHz: L203 = 470 nH; C210 = 680pF

The LC values are not critical, and the filtering they provide might have only a tiny/negligible impact on receiver performance.

[Gerald Boyd]

Will try the resistance values below

Sent from my iPad

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/0d6d9014-28e6-490c-aafb-3780a8d8a52f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.