[DSP-10] 10MHz Crystal osc question

Fri, 22 Feb 2002 22:28:43 -0600

Steve:

I am, and quite aware, about the Pierce Oscillator in addition of several
other types since I have been using them for more that 38 years as an
Electronic Design Engineer in areas of R.F.  --

I wanted you to do the clarification of your statement.

Regards

Nando

-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Steven Bible
Sent: Friday, February 22, 2002 9:33 PM
To: [email protected]
Subject: RE: [DSP-10] 10MHz Crystal osc question

Hello Nando,

You are correct in stating if the capacitors are in parallel you add them.
I made a misleading statement.  Please allow me to clarify,

In a Pierce oscillator, two load capacitors are placed on each side of the
crystal.  They have a dual purpose, they provide reactance and thus phase
shift and they provide impedance matching between the amplifier (which is
onboard the LMX1501A) and the crystal.  Recall that the Barkhausen criteria
states that for an oscillator to start, you must have gain exceeding unity
and N x 2 X Pi radian phase shift in the loop.  The two load capacitors
provide part of the necessary phase shift.  The best explaination of Pierce
oscillator operation I have read is in R.J. Matthys, "Crystal Oscillator
Circuits."

I made the mistake of saying the load capacitors are in parallel with the
crystal.  Physically maybe, but electroly they are in series.  Thus the
formula I gave (capacitors in series) is correct, my description was not.

Does this help clear things up?

Sorry about the mistyping :-)

- Steve, N7HPR
  ([email protected])

> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of H.G.
> Sent: Friday, February 22, 2002 11:42 AM
> To: [email protected]
> Subject: RE: [DSP-10] 10MHz Crystal osc question
>
>
> STEVE:
>  >If C108 is 22 Pf and the 37 Pf you added to C110, in parallel
> that's 13.8
> Pf.
>
> If the capacitors are in parallel, the values need to be added, if the
> capacitors are in series the capacitors need to be added with the inverse
> value and the result the inverse value.
>
>  59 Pf if in parallel  or 1/22 + 1/37 = 13.79 Pf if in series,.
>
> Regards
>
> Nando
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of Steven Bible
> Sent: Thursday, February 21, 2002 10:22 PM
> To: [email protected]
> Subject: RE: [DSP-10] 10MHz Crystal osc question
>
>
> Hello Lee,
>
> The 10 MHz crystal supplied with the TAPR kit should be an Epson
> CA-301 Type
> (DK SE3420).  There's information about how Bob selected this
> crystal on the
> DSP-10 web pages.
>
> In looking at the data sheet, the load capacitance is 18 pF.  I
> assume that
> when you say you are adding 37 pF you are stating the total parallel
> capacitance around the crystal.  If that's the case, then it
> sounds like you
> are in the ball park.  If C108 is 22 pF and the 37pF you added to C110, in
> parallel that's 13.8 pF.  Add 5 pF for stray capacitance and the total is
> 18.8 pF.  So you see, you are there.  Since U104 (LMX1501A) is using a
> Pierce Oscillator, it would be best to balance the load capacitor values.
>
> If I made a bad assumption about your parts placement, can you let us know
> what capacitors you added where?  What values you have for C108, C109 and
> C110?
>
> - Steve, N7HPR
>   ([email protected])
>
>
>
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]]On Behalf Of Lee Scott
> > Sent: Thursday, February 21, 2002 12:02 PM
> > To: [email protected]
> > Subject: [DSP-10] 10MHz Crystal osc question
> >
> >
> > I have my DSP-10 up and running but have had to pad quite a bit on the
> > 10MHz crystal to get it to the right frequency.  As originally
> > constructed,
> > the DSP-10 was transmitting 10KHz high so I padded C109 first
> with a 5pf,
> > then a 10pf cap.  I was still too high and I had to add a 27pf
> > NPO disk cap
> > from C110 to gnd before the low end of the C110 would tune the
> > osc down low
> > enough to bring it on freq.
> >
> > Am wondering if the 10MHz crystal TAPR is kitting is the right one as I
> > have had to pad a total of 37pf to pull the osc down.
> >
> > Lee Scott - AA1YN
> >
> > _______________________________________________
> > DSP-10 mailing list
> > [email protected]
> > http://mailman.qth.net/mailman/listinfo/dsp-10
>
> _______________________________________________
> DSP-10 mailing list
> [email protected]
> http://mailman.qth.net/mailman/listinfo/dsp-10
>
> _______________________________________________
> DSP-10 mailing list
> [email protected]
> http://mailman.qth.net/mailman/listinfo/dsp-10

_______________________________________________
DSP-10 mailing list
[email protected]
http://mailman.qth.net/mailman/listinfo/dsp-10