[hpsdr] Angelia and Hermes External Clock problem

[John Sager]

The TCXO used in Hermes is a Fox 924. Although there is no rise and fall
time specification on the data sheet, it does say that the output is
HCMOS, which implies a fast risetime. This is borne out by there being
no PLL unlock glitches, though the device does exhibit small steps in
frequency with temperature changes, as discussed previously on this list.
For many modes, neither the specified frequency error nor the frequency
steps are of any consequence but there are uses for Hermes which require
good frequency stability and/or high frequency accuracy.

The external reference input on Hermes has not performed well thus far
and the addition of a schmitt trigger buffer as a piggyback board
restores its performance in this area.

73 John G8ONH

On 20/08/13 15:21, Helmut, DC6NY wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> Hi Stijn,
>
> no disagreement! I also don't know why the designer saved such a hex
> inverter or similar on Hermes/Angelia. Most other hardware details and
> components are a 100% copy of Mercury, Penny and Metis. On all of these
> cards the same 'simple' TCXO is used to feed the FPGA with a clipped sine
> signal. May be that's an issue in frequency critical applications. In that
> case Mike, KF4BQ, found an appropriate solution. It seems also possible to
> feed an adjusted external square wave signal directly via the SMA jack and a
> short piece of thin coax to J20.
>
> 73, Helmut, DC6NY
>
>
> -----Ursprüngliche Nachricht-----
> Von: hpsdr-bounces at lists.openhpsdr.org
> [mailto:hpsdr-bounces at lists.openhpsdr.org] Im Auftrag von Stijn Nestra
> Gesendet: Dienstag, 20. August 2013 15:38
> An: hpsdr at lists.openhpsdr.org
> Betreff: Re: [hpsdr] Angelia and Hermes External Clock problem
>
> ***** High Performance Software Defined Radio Discussion List *****
>
> Hello Helmut ea.
>
> Indeed Atlas is a solid platform in this regards and Hermes has some parts
> loaned from this system. But in the external clock circuitry some parts have
> been omitted. Excalibur in the Atlas system contains except the clock shaper
> also a 74ac04 buffer. This takes care of the fast rice-times that we seem to
> be missing on Hermes. The second thing this
> 74ac04 does is: it protects the fpga input a bit against overvoltage since
> the 74ac04 is powered by 3.3v and not 5v.
>
> Sincerely,
>
> Stijn PE1RKS
>


 1377021751.0