[time-nuts] new WWVB BPSK dev board

[Bob kb8tq]

Hi

The “good sound card” in this case is indeed not quite as easy to find
as one might think. Most people playing with audio rarely are concerned 
once you get past 20KHz or so. Finding a card that stays quiet past 
40 KHz can be tough. Even on a 192 KHz card, one can find anti-alias 
filters cutting in way before the hoped for ~90 KHz passband is reached. 

Bob


> On Dec 5, 2018, at 1:49 PM, Alex Pummer <alex at pcscons.com> wrote:
> 
> There is a problem with most pre-amplifiers and also particularly with anything up font of the sampling; to get  good data the amplitude of the sampled signal has to have relative high levels, comparable to microphone output level -- up to 50mV -- if one wanted to use the sound cart input. Additionally to that the sound card has it's internal amplifier before the A to D converter. Unfortunately the "vicinity" of the 60kHz is not a quite area, but it is full with noise of house hold and industrial origin. I spent substantial time to investigate that noises and I found many unsuspected sources. I may own household I have on tooth cleaning jet for every day use, which was replaced recently wit a new one since the previous one was more than fifteen years old and started to show old age problems. The new unit looking from the outside was very similar to the old one except that the old had one AC motor the new one has a DC motor with a switching mode power supply, which is with the motor's commutator generating RF noise up to the TV bands disturbing the reception and of course generating enough interference to overload any sound card with pre -amplifier. I am very curious how did they get that device -- the water jet-- trough the FCC test  or did not ever border to do it? But that water jet  is not alone, just look your own computer's including it's power supply. You no not need any very sophisticated instrument just some old spectrum analyzer, which is running on it's own without any computer connection, and you will what is out there, sweep so between 35kHz to 100kHz use one coils as antenna with a diameter of 2' to 3' and at least thirty turns. With that set up you will be able to see WWVB to. If you turn the coil's axis perpendicular to the direction from your position to Forth Collins in Colorado, you will see the --AM also-- modulated WWVB at 60kHz. These illegal carriers with very large amplitude are able to overload the input of the sound card's -- or any other --A to D converter. The old WWVB receivers for the at the time just AM modulated signal used to use crystal filter for good reason, which is not usable any more since the frequency spreading of the phase modulation will not fit into the crystal filter's band width, and the settling time of the filter's output will cause one additional AM modulation.   By using some selection and very linear amplification I was able to get did I get a 120mVp_p signal from WWVB and compared it's phase stability  to the phase of one other 60kHz signal derived from a GPS receiver with one u-blox board. If there is interest  I could describe how did I do it.
> 
> 73
> 
> KJ6UHN
> 
> Alex
> 
> On 12/4/2018 8:09 PM, David G. McGaw wrote:
>> Actually, an RTL-SDR can because there is direct access to the ADC
>> available by soldering to internal pads:
>> www.rtl-sdr.com/rtl-sdr-direct-sampling-mode/  That will give you 8-bit,
>> 14.4Msps.
>> 
>> But as has also been said, a good sound card sampling 24 bits at 192kHz
>> can be used.
>> 
>> David N1HAC
>> 
>> 
>> On 12/4/18 6:54 PM, Poul-Henning Kamp wrote:
>>> --------
>>> In message <2e7cf0ff-4094-2750-4874-96dfe2efe485 at earthlink.net>, jimlux writes:
>>> 
>>>> I'm going to bet that the 8 bit RTL-SDR isn't going to work on 60kHz.
>>> I don't know about the RTL-SDR, but 8 bits will get you quite far with
>>> slow moving time signals like WWVB because you can average for minutes
>>> if you want - provided you feed the ADC a good stable clock.
>>> 
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>> 
>> 
>> ---
>> This email has been checked for viruses by AVG.
>> https://www.avg.com
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.