Re: [time-nuts] SI532X Chips Close In Spurs (Somewhat Long)

[Mark Goldberg]

On Mon, Jan 8, 2018 at 11:37 PM, Hal Murray  wrote:

> If you look at the binary of the word that goes into a DDS adder, the ones
> with more low order bits have closer in spurs.


Which of the setting is the binary word going into the DDS adder? I think
this uses a SI5328, and it is not described as a DDS chip in the data
sheet, but I am far from knowledgeable about this device.

I don't know what the output of the GPS is that goes into the SI5328.

Thanks,

Mark
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] SI532X Chips Close In Spurs (Somewhat Long)

[Hal Murray]

marklgoldb...@gmail.com said:
> I was hoping someone who knows these devices might have some insight into
> what the issue is and how to determine which plans will not provide close in
> spurs, without requiring testing. Any suggestions are appreciated. 

If you look at the binary of the word that goes into a DDS adder, the ones 
with more low order bits have closer in spurs.  Think of that word as a 
fraction, binary point on the left rather than right.

If your adder word is 0.001 (binary) it will divide by 8, no spurs.
If you have 0.0011, it will divide by a little less than 8 and have spurs.



-- 
These are my opinions.  I hate spam.



___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

[time-nuts] SI532X Chips Close In Spurs (Somewhat Long)

[Mark Goldberg]

I am using the Leo Bodnar Single output Mini GPSDO to generate 38.4 MHz as
an input to a Wenzel Oscillator that locks to it and produces a low phase
noise 80 MHz to drive my Perseus SDR. The Wenzel has a low loop bandwidth
so the phase noise from the Bodnar above the 10-100 Hz range does not seem
to get through.

The issue is that there are sometimes some close in spurs from the Bodnar
that do get through, about 31 and 62 Hz on either side of the 38.4 MHz
output. I believe the GPSDO is based on a SI532X chip, as reported for the
two output version. The spurs are only about 70 dB down, maybe 15-20 dB
above the phase noise in the area.

I tried several frequency plans as the SI532X can generate the same output
with several different sets of dividers. I did find one that worked very
well, with phase noise results and Allan Deviation that appear better than
the reported phase noise for the two output version. I basically used the
Bodnar software, SI's DSPLLSIM, some spreadsheet calculations and trial and
error to determine the working frequency plan.

I was hoping someone who knows these devices might have some insight into
what the issue is and how to determine which plans will not provide close
in spurs, without requiring testing. Any suggestions are appreciated.

Below is the long part. I have listed the frequency plans I tried and the
results below. Maybe someone can see a pattern.

Thanks

Mark


*My First Try Plan *


GPS 2,560,000
N31 1
N2_HS 7
N2_LS 300
N1_HS 7
NC1_LS 20
BW 15


F3 2,560,000
Fosc 5,376,000,000


Fout 38,400,000
Fout/F3 15


*No Spurs, F3 above data sheet spec*






*My Plan with in spec F3*


GPS 2,560,000
N31 2
N2_HS 7
N2_LS 600
N1_HS 7
NC1_LS 20
BW 15


F3 1,280,000
Fosc 5,376,000,000


Fout 38,400,000
Fout/F3 30


*Spurs*






*Another Plan*


GPS 2,560,000
N31 1
N2_HS 11
N2_LS 150
N1_HS 11
NC1_LS 10
BW 15


F3 2,560,000
Fosc 4,224,000,000


Fout 38,400,000
Fout/F3 15


*No Spurs, F3 above data sheet spec*






*DSPLLSIMs Plan*


GPS 2,560,000
N31 2
N2_HS 10
N2_LS 396
N1_HS 11
NC1_LS 12
BW 15


F3 1,280,000
Fosc 5,068,800,000


Fout 38,400,000
Fout/F3 30


*Spurs*






*DSPLLSIMs Plan 2*


GPS 1,920,000
N31 1
N2_HS 5
N2_LS 576
N1_HS 8
NC1_LS 18
BW 15


F3 1,920,000
Fosc 5,529,600,000


Fout 38,400,000
Fout/F3 20


*Bad Spurs*






*DSPLLSIMs Plan 3 Using This One*


GPS 176
N31 1
N2_HS 9
N2_LS 320
N1_HS 6
NC1_LS 22
BW 15


F3 1,760,000
Fosc 5,068,800,000


Fout 38,400,000
Fout/F3 21.8181818181818


*No Spurs*
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.