Years and years ago, I was involved in an area that had to do with
propagation and circuit path analysis. I was in a room full of academics
when the hot discussion began. The issue was what is the instantaneous
dynamic range of the environment. There were two schools of thought. One
said that the environment had a dynamic range of 120db and the other
said 144db. What is missed by outsiders to this is the question for what
period of time? As path reliability is quoted as a probability, time
matters. As it turns out all of these folk were fighting over a period
represented by less than 5%. And, most of the published environmental
data is for the median, 50% of the time.
There was a study that someone wrote and published a number of years ago
in QST that stated 80db was all you need as a ham. And, the author gave
his reasons. I think it should be a bit more, but we could start another
thread on that issue.
I have done direct down conversion SDRs using 16 bit 100MHz A/Ds.
However, you really don't get 100MHz bandwidth as Nyguist says you need
to sample at least 2X the bandwidth of interest. And adding a little
practicality to that, you really don't get 1/2 the bandwidth either as
you would like some over-sampling. If you want to play with this, Analog
Devices makes the A/D that I used. However they ain't cheap.
I can assure you that direct sampling works. There are some real tricks
involved known to the DSP engineers, of which I am not one. But, it
works and works well.
To account for what you are concerned with, what you really want is an
inverse taper of receiver gain, less gain at the bottom and more at the
top. By doing this, environmental noise will less likely overwhelm your
A/D, yet total system noise figure can be maintained such that you will
always be externally noise limited. And secondarily, Don't use too much
gain up front. Distribute the gain along the receiver chain to keep
dynamic range. This latter point is how you get quiet, high dynamic
range, sensitive receivers.
Spending more than say $150 for a really good sound card probably
doesn't make a whole lot of sense when using a KX3. But, I admit I do
like the idea of the capability to mess with the thing as I do. If
you're saner than I, use the sound card in you computer. It will work
quite well when using the line out. You just miss the fun of an
additional play toy.
73,
Barry
K3NDM
On 8/21/2012 8:37 PM, Joe Subich, W4TV wrote:
>> I am not using demoded line out audio from my KX3. I use the raw I &
>> Q data. I think that is what we missed in our past exchanges. That
>> means I can really use the dynamic range of a 24 bit card because I
>> set things so that noise doesn't control my A/D any more than a few
>> bits.
> The hype by certain manufacturers of direct down conversion I/Q rigs
> aside, I do not believe that direct I/Q to a 24 bit soundcard provides
> the necessary dynamic range. First as K9YC has alluded to, far too
> many pf the low level bits are "advertising bits" - they are either
> lost in the low level noise or exceed the clipping level in prior
> stages. Secondly AGC is always an issue leading to significant
> issues with blocking dynamic range problems ("window" pumping).
>
> Direct sampling architecture may provide the ultimate solution in
> those areas but I still feel it will be necessary to wait on the
> development of cost effective a 28 or 32 bit ADC in order to obtain
> the performance necessary for a true 150 dB (MDS= -145 dBm) spurious
> free dynamic range in the presence of multiple strong signals. Until
> then, well designed 16 bit ADC sound devices fed with demodulated audio
> will continue to be the most cost effective high performance solution
> for mainstream applications.
>
> 73,
>
> ... Joe, W4TV
>
>
> On 8/21/2012 6:17 PM, Barry LaZar wrote:
>> Joe,
>> I think I just figured out we are not necessarily comparing apples
>> to apples. I just responded to a note you wrote on this reflector that
>> caused me to review what we are saying. You are correct of course if we
>> allow our basic radios to do their thing without our involvement and
>> then we do an A/D afterward.
>>
>> You are also correct in pointing out that all of our radios are
>> externally noise limited today. However, the issue of USEFUL dynamic
>> range is only eluded to by you. Antenna noise is the vector sum of all
>> of the noise in the environment, regardless of how it got there. It's
>> true living where I do, in an urban environment with power lines and
>> gasoline engines, I'll never see antenna noise that looks like the
>> atmospheric noise that can be seen down on the farm. And yes, I'm a
>> little idealistic in my designs, but during a contest or near the BC
>> stations your electronics and software are going to be tested. And
>> remember that the international BC guys design to hit urban
>> environments, meaning lots pf power that will assure high S/N ratios on
>> you communications receiver, not necessarily that cheapy SWL radio from
>> Radio Shack.
>>
>> I am not using demoded line out audio from my KX3. I use the raw I
>> & Q data. I think that is what we missed in our past exchanges. That
>> means I can really use the dynamic range of a 24 bit card because I set
>> things so that noise doesn't control my A/D any more than a few bits. I
>> believe you need to so it this way just so you don't miss anything. But,
>> we can discuss that in the future.
>>
>> 73,
>> Barry
>> K3NDM
>>
>> On 8/21/2012 2:57 PM, Joe Subich, W4TV wrote:
>>> Barry,
>>>
>>>> The issue is really when will all those bits go to a logic 1. With 16
>>>> bits, it will be ~96 db where db=20log V2/V1. I know this is not
>>>> what most perceive, but it's true the A/D convertor is a voltage
>>>> sensitive and not a power sensitive device.
>>> The issue with dynamic range is just as much the ability of all those
>>> bits to go to logic 0. If the receiver has so much noise - whether
>>> atmospheric noise at the antenna, thermal noise in the IF, masking
>>> due to A/D/A conversions in the demodulator, or thermal noise/hum in
>>> the audio output stages, the sound card will never achieve even the
>>> theoretical 96 dB dynamic range for 16 bit A/D.
>>>
>>> Add to that, switching noise on the typical USB power line, the lack of
>>> secondary regulation in inexpensive USB sound "dongles", the failure
>>> to provide a separate *regulated* reference for the ADC, failures to
>>> provide *regulated* references for op-amp preamps, lack of attention
>>> to grounding and bypassing, etc. and the noise floor in commodity USB
>>> sound devices (and "inexpensive" amateur products) further limits the
>>> achievable dynamic range.
>>>
>>> With external noise sources, it doesn't matter if one has a 16 or 24
>>> bit ADC, the achievable dynamic range will be limited to the area
>>> between *all* of the noise contributors and the maximum voltage
>>> handling capability (typically < 5.0V) of the ADC ... or between
>>> 65 and 70 dB in many of the commodity and inexpensive amateur
>>> devices.
>>>
>>> 73,
>>>
>>> ... Joe, W4TV
>>>
>>>
>>> On 8/21/2012 10:48 AM, Barry LaZar wrote:
>>>> Joe,
>>>> I agree with your comment on internal noise. However, 24 bits of
>>>> quantization will get you a better dynamic range. But, It becomes an
>>>> interesting point when the effects of receiver AGC gets into the
>>>> discussion on this point. Someone in a much earlier post said that the
>>>> receiver's AGC would keep things within bounds for a 16 bit card; he may
>>>> be correct. But, if you run without AGC on the receiver and use AGC
>>>> developed by the SDR software, having those extra bits will make a
>>>> difference, particularly during noisy periods or contests.
>>>>
>>>> The issue is really when will all those bits go to a logic 1. With
>>>> 16 bits, it will be ~96 db where db=20log V2/V1. I know this is not what
>>>> most perceive, but it's true the A/D convertor is a voltage sensitive
>>>> and not a power sensitive device. There are some other interesting
>>>> subtleties like how much dynamic range do you really need, etc. Since I
>>>> haven't had my first cup of coffee yet, I think I would not like to get
>>>> in to those at this time.
>>>>
>>>> 73,
>>>> Barry
>>>> K3NDM
>>>>
>>>> On 8/21/2012 10:13 AM, Joe Subich, W4TV wrote:
>>>>>> However, this card is not current; I think the 0404 replaces it.
>>>>>> Someone can address this for you.
>>>>> The current version is eMU 0204 - approximately $125 at places
>>>>> like B&H Photo Video.
>>>>>
>>>>>> You really do want a 24 bit card and not 16 bit one.
>>>>> Unless you need the high sample frequency for wide panadapter service,
>>>>> any well designed 16 bit/48 KHz USB sound device will provide all the
>>>>> performance you need at less than 1/4 the price of something like the
>>>>> eMU 0204.
>>>>>
>>>>> > The other thing is that that card should have low internal noise;
>>>>>
>>>>> This is *the most important* determinant of performance. Internal
>>>>> noise dramatically reduces the effective dynamic range of the sound
>>>>> card.
>>>>>
>>>>> 73,
>>>>>
>>>>> ... Joe, W4TV
>>>>>
>>>>>
>>>>> On 8/20/2012 4:40 PM, Barry LaZar wrote:
>>>>>> Howie,
>>>>>> The dongles I have seen are not spectacular. They are about
>>>>>> on par
>>>>>> with your average stereo laptop soundcard, OK but not great. I use a
>>>>>> E-MU0202 from Creative Labs. This card is a 24 bit card that can sample
>>>>>> beyond what Elecraft recommends. However, this card is not current; I
>>>>>> think the 0404 replaces it. Someone can address this for you. You really
>>>>>> do want a 24 bit card and not 16 bit one. The other thing is that that
>>>>>> card should have low internal noise; the high end Creative Labs card
>>>>>> meet this as does several others.
>>>>>>
>>>>>> There are a number of software packages around that you can
>>>>>> use. I
>>>>>> use NaP3 because it can control the KX3 and allows me to use VAC so that
>>>>>> I can pipe the output to FLDIGI for reading the digial modes. I have
>>>>>> essentially a full up computer controlled SDR that I can take out for
>>>>>> Field Day and operate portable with knobs and a battery. It's the
>>>>>> craziest!.
>>>>>>
>>>>>> I think, once you have it all lashed up, you will really have
>>>>>> some
>>>>>> fun. Good luck and do use a good card with ASIO drivers for best
>>>>>> performance.
>>>>>>
>>>>>> 73,
>>>>>> Barry
>>>>>> K3NDM
>>>>>>
>>>>>>
>>>>>> On 8/20/2012 12:37 AM, Howard Hoyt wrote:
>>>>>>> Hi,
>>>>>>>
>>>>>>> I have been following the fun on this reflector while awaiting my KX3
>>>>>>> (next Friday!). I am very interested in using the I&Q outs to interface
>>>>>>> with my laptop, and as many have noted, you need a good, compatible
>>>>>>> sound device to have a usable panadapter display. Does any one have
>>>>>>> experience with a 192k sampling USB sound dongle for their laptop?
>>>>>>>
>>>>>>> TIA!
>>>>>>>
>>>>>>> Howie
>>>>>>>
>>>>>>> Howard Hoyt - WA4PSC
>>>>>>> CE - WXYC-FM 89.3
>>>>>>> UNC Chapel Hill, NC
>>>>>>> www.wxyc.org
>>>>>>> 1st on the internet
>>>>>>> ______________________________________________________________
>>>>>>> Elecraft mailing list
>>>>>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>>>>>> Help: http://mailman.qth.net/mmfaq.htm
>>>>>>> Post: mailto:Elecraft@mailman.qth.net
>>>>>>>
>>>>>>> This list hosted by: http://www.qsl.net
>>>>>>> Please help support this email list: http://www.qsl.net/donate.html
>>>>>>>
>>>>>> ______________________________________________________________
>>>>>> Elecraft mailing list
>>>>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>>>>> Help: http://mailman.qth.net/mmfaq.htm
>>>>>> Post: mailto:Elecraft@mailman.qth.net
>>>>>>
>>>>>> This list hosted by: http://www.qsl.net
>>>>>> Please help support this email list: http://www.qsl.net/donate.html
>>>>>>
>>>>> ______________________________________________________________
>>>>> Elecraft mailing list
>>>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>>>> Help: http://mailman.qth.net/mmfaq.htm
>>>>> Post: mailto:Elecraft@mailman.qth.net
>>>>>
>>>>> This list hosted by: http://www.qsl.net
>>>>> Please help support this email list: http://www.qsl.net/donate.html
>>>>>
>>>> ______________________________________________________________
>>>> Elecraft mailing list
>>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>>> Help: http://mailman.qth.net/mmfaq.htm
>>>> Post: mailto:Elecraft@mailman.qth.net
>>>>
>>>> This list hosted by: http://www.qsl.net
>>>> Please help support this email list: http://www.qsl.net/donate.html
>>>>
>>> ______________________________________________________________
>>> Elecraft mailing list
>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>> Help: http://mailman.qth.net/mmfaq.htm
>>> Post: mailto:Elecraft@mailman.qth.net
>>>
>>> This list hosted by: http://www.qsl.net
>>> Please help support this email list: http://www.qsl.net/donate.html
>>>
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:Elecraft@mailman.qth.net
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft@mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:Elecraft@mailman.qth.net
This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
