XO Questions

8 views
Skip to first unread message

TedS

unread,
May 11, 2007, 1:54:20 PM5/11/07
to EZDAC Builders
Hi Evan,

I've just about finished my EZDAC board. I'm waiting for another 8416
chip, since my first one was sacrificed to experimenting with
connecting for multiple inputs. Otherwise, the last item left is the
crystal oscillator. For another project I've been following, someone
suggested the following Crystek crystal oscillator from Mouser:

549-C3391-24.576

The data sheet for it has a jitter spec of 0.5ps Typ, 1.0ps Max. I've
been looking for, but haven't been able to find a jitter spec for your
Abracon oscillator. The only reference I've seen is in Appendix A of
the Abracon oscillator catalog, where they say they have typical
jitter of 20 - 60ps. Do you know what the jitter spec of your
oscillator is? If this part has better jitter performance, I'm
considering trying it out (I have one each of yours and this one). Do
you see any issues with substituting this XO? It is in a slightly
different package, so wiring it to your board would require short
jumper wires.

Secondly, I was looking at the data sheets and application notes for
both oscillators, and both seem to recommend a 0.01uF cap between pins
1 and 3 (VDD and GND). In your circuit this appears to be C21, which
has a value of 15pF. I was wondering why you chose that cap instead
of a 0.01uF, like C17?

Thanks,
Ted

ezkcdude

unread,
May 11, 2007, 2:40:16 PM5/11/07
to EZDAC Builders

Ted, that oscillator you refer to is a surface mount component.
Obviously, you wouldn't want to use that.

As for the capactior, have you read the Abracon application notes?
http://www.abracon.com/Support/appendix-a.pdf

In there it mentions using cap values of 30pF, but I saw some other
circuits that use 15pF for the XO. I think the reason that it is such
a small value is that the output of the XO is such a high frequency.
Anyway, no one has ever corrected me on that, so I will assume it's
o.k. until then.

TedS

unread,
May 11, 2007, 4:24:40 PM5/11/07
to EZDAC Builders
Hi Evan,

Yes, I was looking through the Abracon notes. I didn't see anything
applicable in Appendix A, but in Appendix B all the sample circuits
show 0.01 uF from pin 1 to GND. The Crystek circuits show the same
value cap from pin 1 to GND, too. Maybe I missed something?

I realize the Crystek oscillator is surface mount. I was considering
trying to attach it to your board with short jumper wires, since it
appears to have lower jitter (0.5ps rms). But I couldn't find a
jitter spec in the Abracon notes, either, other than the statement of


typical jitter of 20 - 60ps.

Did you find another jitter spec for the Abracon crystal somewhere?

Ted

> As for the capactior, have you read the Abracon application notes?http://www.abracon.com/Support/appendix-a.pdf


>
> In there it mentions using cap values of 30pF, but I saw some other
> circuits that use 15pF for the XO. I think the reason that it is such
> a small value is that the output of the XO is such a high frequency.
> Anyway, no one has ever corrected me on that, so I will assume it's

> o.k. until then.- Hide quoted text -
>
> - Show quoted text -

ezkcdude

unread,
May 11, 2007, 5:24:57 PM5/11/07
to EZDAC Builders
Ted, I remember seeing a value of around 50 ps for jitter for this
type of oscillator, but no specific measurements for this particular
model. I'd say try out the smt part, but it seems like simply
jumpering it in there isn't the greatest idea in the world, as it may
affect the impedance. Leo tried an external XO, however, and he liked
it. In the end, it's your board now, so do with it what you please, of
course. As for the cap, I think you should just try .01uF. Honestly, I
don't think it would make an audible difference either way, and you'll
probably rest easier at night ;) (how's that for diplomacy?)
-evan

TedS

unread,
May 11, 2007, 9:45:39 PM5/11/07
to EZDAC Builders
Hi Evan,

Thanks for your thoughts. I apologize if that came across wrong. The
anal part of me always wants the finer details! I agree that I
probably won't notice the difference regarding the cap. I already
have the 15pF part on the board, so I'm not going change it. I was
just curious about that and this other oscillator. I, too, was
wondering about how connecting the crystal with short wires to the
board would work, but I've seen it done in a couple other places. So
I thought maybe I'd give it a try, too.

Thanks again for your design and support of this kit!
Ted

> > > - Show quoted text -- Hide quoted text -

Hans Nielsen

unread,
May 12, 2007, 7:18:11 AM5/12/07
to ezdac-b...@googlegroups.com
 
Hi I vill use the new dclock from Dexa electronics, as u7. 
 
-------Original Message-------
FREE Emoticons for your email - By IncrediMail! Click Here!

Evan Zamir

unread,
May 12, 2007, 10:19:43 AM5/12/07
to ezdac-b...@googlegroups.com
Ted, no need to apologize. If you get the board working, you can always go back later and try out those other parts.
-evan

Evan Zamir

unread,
May 12, 2007, 10:21:10 AM5/12/07
to ezdac-b...@googlegroups.com
Hans, that's one expensive clock!
-evan

Hans Nielsen

unread,
May 12, 2007, 10:26:06 AM5/12/07
to ezdac-b...@googlegroups.com
 
Yes I now, but I think it is a god choice. 
 
-------Original Message-------

mult...@ntlworld.com

unread,
May 23, 2007, 9:19:34 PM5/23/07
to EZDAC Builders
I used the Crystek oscillator in the Asynchronous reclocker in my
Pedja dac, they are very good! I've not tried one in the Ezdac or
compared it directly against the Abracon, the Abracon works fine in
the Ezdac, build it standard first and upgrade once you know
everything is working ok;-)
Here you can see the Crystek hardwired to the pcb, its quite easy if
you have a steady hand http://pic20.picturetrail.com/VOL1293/4634025/9699445/138046909.jpg
I'm still using the modded Ezdac with my UCD amps:-)

Cheers,
Leo

mult...@ntlworld.com

unread,
May 23, 2007, 9:30:21 PM5/23/07
to EZDAC Builders
I forgot to mention I recently tried a Burson low jitter clock with
the Ezdac replacing the PFM Flea, I swapped the Burson fitted
oscillator to the Tentlabs XO, I also adjusted the output voltage of
the Burson low jitter clock to match.
Results so far are very good, I'm now in need of a bigger case, the
Pass D1 I/V circuit will hopefully be the next thing to try

Cheers,
Leo

Reply all
Reply to author
Forward
0 new messages