[hpsdr] OCXO 10 MHz clock source
hartfuss
hartfuss at ipp.mpg.de
Fri Jun 12 04:48:40 PDT 2009
Hi Graham,
the external 10 MHz clock source is running very well. I appreciate your
hints very much. Thanks again!
As reported earlier I purchased a 10 MHz OCXO from the German AXTAL
company; it's their model AXIOM30-12-05 with very low phase noise. It
is operated from the 12V supply on Atlas and delivers 5V signals with
50% duty cycle.
As you might remember I built a compact hpSDR transceiver version with
the 3-slot Atlas-EU board with no space left for Excalibur.
Therefore I removed the ATX connector and put similar to your protoype
version a small printboard on its place carrying the OCXO and a driver
with the 74LVC244A operated with 3.3V supply voltage. This allows for
very short connections to C16, as well as to GND and the 12 supply
voltage on ATLAS.
The clock signals measured on C16 of ATLAS with all boards in place are
very nice: 3.3Vpp with a 10-90% rise time of 3 ns.
I am preparing a description with photos over the weekend.
With best 73s, Hans, DL2MDQ.
PS: to adjust the frequency as accurate as possible (a few Hz are
possible with a control voltage on the OCXO) I also purchased a used
Rb-standard (DATUM/EFRATOM, LPRO-101)cheaply available on the surplus
market. It will probably be ready in a week.
Hans:
Well, the latest software versions for Mercury and Penelope,
when not receiving a 10 MHz clock input will free run at about
the center of the frequency range of the 122.88 VCO. This means
that if you select ATLAS as the 10 MHz clock source, and there
is no 10 MHz signal on ATLAS bus line C16, then Mercury and
Penny will still operate and typically will stay within several
hundred Hertz of frequency. If you are real lucky they may stay
within several 10s of Hertz.
This was not true of the early software versions, and Mercury and
Penny would quit with the lack of a 10 MHz clock.
--- Graham
Phil Harman schrieb:
> ***** High Performance Software Defined Radio Discussion List *****
>
>
>
> ------------------------------------------------------------------------
>
> Hi Alberto,
>
> Let me see if I can shed some light on what is happening. If you use
> Mercury and Penelope you need a way to phase lock the 122.88MHz clocks
> on each board. There are a number of ways you can do this:
>
> 1. Use the 10MHz TCXO on Mercury
> 2. Use the 10MHz TCXO on Penelope
> 3. Use an external 10MHz reference on Atlas C16 e.g. Excalibur
>
> The TCXO's on Mercury and Penelope are not intended to be highly
> accurate - just good enough to lock the two 122.88MHz clocks together.
> If you would like higher accuracy then use an external reference.
>
> The 122.88MHz clocks on Mercury and Penelope are divided down to 80kHz
> as is the 10MHz reference from Penelope or Mercury or Atlas. The two
> 80kHz signals are presented to an OR gate which is used as a phase
> detector in a PLL.
>
> So if you select Atlas as your 10MHz source, and there is no 10MHz
> signal on it, then the output of the OR gate phase detector will be a
> 80kHz square wave. After this passes through the PLL filter the output
> is a DC voltage at 3.3/2 volts. This voltage is fed to the 122.88MHz
> VCXO which sets it to the mid point of its frequency range.
>
> If you are lucky, and have a good 122.88MHz clock, then this could be
> very close to 122.88MHz - closer than if you phase lock it to the
> relatively low accuracy 10MHz TCXOs.
>
> Hence you may find your frequency accuracy is actually better with no
> 10MHz clock on Atlas rather than using the 10MHz TCXO on Mercury/Penelope.
>
> If you want higher accuracy then use an external 10MHz reference. If
> you use Penelope and Mercury, and select Atlas as your 10MHz reference,
> then you MUST have an actual signal on it or - depending on how good you
> luck is - you will have a frequency difference between Tx and Rx.
>
> No mystery - just simple electronics!
>
> 73's Phil...VK6APH
>
>
>
>
> ----- Original Message -----
> *From:* Alberto I2PHD <mailto:i2phd at weaksignals.com>
> *To:* Keith <mailto:n3ick at cox.net>
> *Cc:* hpsdr at lists.openhpsdr.org <mailto:hpsdr at lists.openhpsdr.org>
> *Sent:* Monday, June 08, 2009 6:51 PM
> *Subject:* Re: [hpsdr] 10 MHz clock source
>
> ***** High Performance Software Defined Radio Discussion List *****
>
> ------------------------------------------------------------------------
>
> Keith wrote:
>
>>Hello Alberto
>> I noticed the frequency shift when working PSK31 before the latest updates
>>and was told that on Mercury if I shorted across R37 it would get it much
>>closer. I tried it and the frequency is very close now 10-15hz. I believe
>>there is an eco out for the change it has to do with the leading edge of the
>>clock pulse on the C1 bus of Atlas. Hope this helps
>>73 de N3ICK
>>
>
> Hello Keith,
>
> thanks for your answer. Will have a look at that EC.
> But still don't understand where the 10 MHz clock comes from, when
> Atlas is selected as source...
>
> 73 Alberto I2PHD
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help:
> http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/
>
> ------------------------------------------------------------------------
>
>
> No virus found in this incoming message.
> Checked by AVG - www.avg.com
> Version: 8.5.339 / Virus Database: 270.12.54/2157 - Release Date:
> 06/05/09 17:55:00
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help: http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/
1244807320.0
More information about the Hpsdr
mailing list