(I forgot to include Flexedge in the reply.)

On Sun, Mar 21, 2010 at 9:04 AM, Brian Lloyd <[email protected]> wrote:
> On Sun, Mar 21, 2010 at 8:06 AM, Mack <[email protected]> wrote:
>> I'd appreciate the thoughts of anyone who has tried to use the Flex 5000A as 
>> an exceptionally accurate frequency measuring standard. As many of you know 
>> the ARRL Frequency Measuring Test is upcoming on April 22nd. It will be a 
>> constant carrier. The test is described in this months QST on page 52.
>>
>> I would use a 10 MHz GPS Disciplined external frequency reference on the 
>> Flex. I could also use a Rubidium 10 MHz house standard as a reference. I 
>> suppose I would use 48 KHz bandwidth and a small number of buffers. Is this 
>> correct and are there other settings that would optimize resolution? I'd 
>> just like to show off how accurate the Flex is if this is feasible.
>>
>> I have other frequency measuring devices that can measure to 31 mili Hz but 
>> really want to use the Flex in this test.
>>
>> All thoughts and input are most welcome.
>
> You probably saw my post about using the Datum/Euratom LPRO-101 Rb
> reference on the main Flex reflector. Jerry Flanders, W4UK, sent me a
> very thoughtful post about F5K accuracy. He had been trying to come up
> with a calibration curve for the F5K and was finding it to be less
> smooth than for his Kenwood. He was therefore concerned that, perhaps
> the F5K is not as good for the FMT as other radios. That caused the
> light-bulb to go on in my head (only a 4W night-light so don't expect
> too much :-).
>
> The DDS determines frequency by the ratio of a multiplier and a
> divisor value. Basically you multiply the reference by a fraction. Now
> the multiplier and divisor values are very large but they are
> discrete. The result is, you can get very close to any frequency you
> want but you are unlikely to land spot-on a frequency. Two things to
> remember:
>
> 1. the steps are not an integer relation to 1Hz;
>
> 2. the step size is not uniform.
>
> The way to use the F5K as a high-accuracy reference receiver is to
> know the exact multiplier and divisor values being used with the DDS
> LO. This will allow you to calculate the exact LO frequency since it
> will probably be several mHz (millihertz) different from the PSDR
> display value. Once you know the exact LO injection frequency you can
> calculate the proper offset to use.
>
> Now there are probably some fractional values that land spot-on a 1Hz
> boundary. Knowing how to figure those out might be helpful. Still,
> knowing the exact algorithm used to select multiplier and divisor
> would be very useful. It is probably buried in the PSDR source.
> Perhaps we can impose on someone at Flex to either provide the
> algorithm used or a pointer to the particular module that does the
> calculation. I bet that module can be modified to actually report the
> error between frequency requested and actual LO frequency produced.
>
> --
> 73 de Brian, WB6RQN/J79BPL
>



-- 
73 de Brian, WB6RQN/J79BPL

_______________________________________________
Flexedge mailing list
[email protected]
http://mail.flex-radio.biz/mailman/listinfo/flexedge_flex-radio.biz
This is the FlexRadio Systems e-mail Reflector called FlexEdge.  It is used for 
posting topics related to SDR software development and experimentalist who are 
using alpha and beta versions of the software.

Reply via email to