[hpsdr] OCXO's for Griffin

Mon Aug 30 15:06:15 PDT 2010

  Hi Phil,

Comments inline....

73's,

Henry.

On 8/30/2010 3:43 AM, Phil Harman wrote:
> Hi Henry,
>
> Given the high cost of a new OCXO my expectation is that TAPR will provide
> a version of the board with or without the OCXO.
This what I was hoping for Excalibur, but it didn't happen that way.
Remember that TAPR's (very legal) interest is a higher number of boards 
to make a good deal on  manifacturing them.
This also goes for buying higher quantities of components.
Manufacturing some with or some without OXCO could also be more 
expensive as could be testing of the different types of boards.

Since no bare PCB boards are made availlable and the Gerbers are not 
released to the public until all the boards are sold, there is no option 
but to buy a more expensive board with partly (expensive) hardware on it 
you do not want/use.

> Given the low cost of the GPS receiver module then I would expect this to
> be included as standard.
>
> The board will have the ability to accept an external GPS locked 10MHz
> reference or an external OCXO that can use the internal GPS to lock it.
> The board will also accept an external 1 PPS signal.
>
> Whilst the board will also have an RS232 input, which could be connected
> to an external GPS receiver, I'm not sure how compatible the protocol is
> between different receivers.

Consider running a serial cable down to the shack from an outside active 
GPS unit (Antenna plus integrated receiver) instead of running a (long) 
UHF signal cable (coax, RF losses) from a passive antenna.

> Since there will always be a GPS receiver on board then we could use that
> for determining the current time.

You are right, I forgot the "requirement" to have an accurate time 
source for WSPR.

 >>> Entering "Time Nuts" territory..... ;-)

External GPS timing solutions are also relatively cheap and widely 
available these days and offer a wide range of options at the choice of 
a particular user, including 1 PPS and 10 MHz options.

The defacto protocol for the majority of GPS units  is the NMEA protocol 
I think , or can be selected in the setup of many GPS units.
Different common protocols are not that different and should be easy to 
impelent as option in Griffin.
Maybe wise to include a small, PIC-like processor for this task instead 
of doing this in the FPGA softcore processor.

This a particular area where a lot of personal preferences are present, 
look at the "Time Nuts" mailing list.....
For these reasons II still would vote for an external time / clock 
reference solution
> Since the board is intended to be powered on at all times we will use the
> latest SMPSU techniques to reduce the power consumption/heat. In which
> case it will make a nice 10MHz reference for the shack as well.

My general point is to give the individual buyer/user the option to 
begin simple and (relatively) cheap and "grow" and expand to a more 
elaborate setup if he wants or use already present equipment in the 
shack without having to invest in hardware he does not use or want.

> Plenty of options to consider whilst we are still in the design phase.
>
> 73 Phil...VK6APH
>
>
>
>
>
>
>> ***** High Performance Software Defined Radio Discussion List *****
>>
>>    Hi Kevin, All,
>>
>> Sorry for having quite a different opinion on OXCO's and the like for
>> Griffin, but I  would leave out all the OXCO and GPS hardware.
>> Just implement the external OXCO header and the external 10 MHz
>> reference clock input.
>>
>> My reasoning for this is that if a particular user wants to have a high
>> precision clock in a stand alone setup he can easily add an external
>> (small, could be simple or more complex design)  PCB with an OXCO or add
>> an external 10 MHz or GPS reference clock (maybe already present in the
>> shack eg. from Excalibur or a Thunderbolt and the like, (maybe even
>> Gibraltar in the future?))
>> I think one of the reasons of having a standard reference clock in the
>> shack is that all equipment  uses this one single clock reference
>> instead of keeping up with a number of different single "references".
>> (I would for this reason prefer a board without any OXCO)
>>
>> I see there is a big advantage of having everything in one standalone
>> package of course, but I see advantages of leaving these out: besides a
>> much simpler design, a lot lower initial cost for a buyer.
>>
>> You maybe also want to avoid the testing/manufacturing difficulties for
>> TAPR (if these are going to be "mass produced" by TAPR) having multiple
>> OXCO types and hence also different populated board types.
>>   From the past experience with eg. Excalibur this ends up in TAPR only
>> making one version, with a particular fixed type of OXCO (for all good
>> reasons for them)  but with unnecessary higher costs for a user who
>> doesn't want this type of OXCO or no OXCO at all (and so save the money
>> for a rather expensive high quality OXCO part)
>>
>> Besides that I believe that from the experiences with Mercury and
>> Penelope, the used !25 MHz (10 MHz) clock XO hardware on these boards
>> are not that bad at all wrt. frequency stability.
>> Maybe even so good that without an external high precision and stable
>> clock reference WSPR should be possible?
>> How does the WSPR (by K1JT) software when used with a standard, analog,
>> (non-HPSDR) transmitter handle (compensate) this aspect?
>>
>>
>> Just my 2 Eurocents ......
>>
>> This is meant as positive comment, no offense intended.
>> Thank you for taking up this new wonderful project for the cool world of
>> HPSDR.
>>
>> In fact so cool that I re-registed my old call "PA0HJA"; effective as of
>> last Saturday 28-AUG-2010 !!!
>>
>> 73's,
>>
>> PA0HJA  Henry, The Netherlands.   ( Changed nickname on Teamspeak ....
>> ;-)   )
>>
>>
>> On 8/29/2010 11:14 AM, Kevin Wheatley wrote:
>>> ***** High Performance Software Defined Radio Discussion List *****
>>>
>>> I've started the schematics for Griffin based on the block diagram
>>> posted on the Wiki, the aim is to provide a number of OCXO footprints on
>>> the PCB, currently I have three:
>>>
>>> http://www.vectron.com/products/ocxo/co714718s.htm (available via
>>> surplus from time to time)
>>> http://www.doveonline.com/item.php?ITEM=DOV013501&DESC=TCO-6730&DESC2=10.000MHZ+12V+2TTL+OCXO
>>> (available via surplus from time to time)
>>> and the VFOV300 from Valpey Fisher
>>> http://www.valpeyfisher.com/Products.aspx?ProductCode=OCXO
>>>
>>> are there any others I should consider?
>>>
>>> 73's
>>> Kevin
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> 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/
>> _______________________________________________
>> 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/
>>
>>
>

 1283205975.0