Dayton 2015 Report - Unofficial

KY6LA_Howard

I will drink almost anything.. but the XYL.. is far too sophisticated a wine afficiando for her own good... or my pocket book...

BTW... really looking forward to drinking Draft Guinness in Ireland again

Steve K9ZW

Howard - do you go to give the presentations or for the ****?  Very interested in how you work it all into your schedule?

BTW for my vote I'll agree with the Pilsen for lager, but put me down for "Old Hooky" from the Hook Norton Brewery for an exceptional bitter.  Greatly prefer the "live" as served at the brewery.  For a dark the "Baaad Boy Dark Wheat" by Three Sheeps Brewery, Sheboygan Wisconsin gets my vote. 

Whatever the tipple, always love visiting the cellar/brewery/caves...

73

Steve

K9ZW

km9r.mike

I have tried Guinness in the states and in the temple bar district in Dublin and learned that I have no taste for it in either location. I would however enjoy a Bushmills at either but preferably stateside. Jamersons not so much but that is another whole can of worms especially there.

Richard McClelland, AA5S

Regarding the New Belgium's Ranger, their brewery is open daily in downtown Fort Collins, we could move the hamfest there.  I'm sure a good time would be had by all.

William Hein

I'll second the brewery option!

Joe, KQ1Q

KY6LA: "...Elecraft tried to misrepresent the K3S as an SDR but clearly it is a Legacy Superhetrodyne Radio with a DSP audio stage...."

It is  interesting that the definition of a software defined radio (or "software radio") has been debated since the invention of the concept: https://en.wikipedia.org/wiki/Software-defined_radio

However the general consensus is that SDR refers to a signal path where detection and demodulation happens in the digital domain. This could be at either RF or IF stages. My Denon consumer stereo receiver does EQ in the digital domain but nobody calls it a "software defined receiver". It has an analog signal chain then digitizes the signal at the AF level and does filtering, EQ, speaker time delay and 5.1 bass management in the digital domain. But the bulk of the signal path is analog.

The Elecraft KX3 is often spoken of as a "software defined radio" with knobs. I'd be interested if a detailed functional block diagram of that radio has ever been published which corroborates that description. I'm not knocking the KX3 -- it is a great radio. However I wonder about the exact technical basis for describing it as an SDR.

KY6LA_Howard

I try to stay sober at least until I have answered all the questions at my presentation.

Friedrichshafen has such a great variety of FREE alcohol that staying sober is very difficult..

At least I am not talking on Friday.

On Sunday after, I am driving with a bunch of my OL3A Contest Group friends to Pilsen for their **** festival...  (Yes Burt Contesters make friends too) followed with several presentations about SDR's at ham clubs in Czech Republic and Austria...

KY6LA_Howard

The KX3 is most definitely a Direct Conversion SDR by my definition a First Generation SDR.

Here is a preview of one of the pages from my upcoming June 27th Friedrichshafen Presentation

"Four Generations of SDR" 

What Is A Software Defined Radio?

•Modulation and Demodulation modes in software, so they are changeable and new ones can be added.
•
•Signal Processing in Software, so it is changeable and expandable.
•
••Control Surface, whether a computer display or physical encoders, is Reconfigurable in software.
•
•Can Add New Features and Capabilities with new controls to operate them completely in software.
•
•Radio is Controlled by Software and is very likely networked. An application program interface (API) may be available.


By this definition a Superhet with DSP ( such as a K3S or an IC-7851) is NOT an SDR.

N7AIG

Thanks for the URL to the Wikipedia article. Interesting that the history actually appears to begin with the old Gold Room and E-Systems. I remember those days very well. I was a constant Midas user.

This tribalism demonstrated here with regard to SDR orthodoxy is also repeated endlessly in discussions of the "best programming language", "best music synthesizer", best... whatever.

Seems to be a human trait, but it really does grow tiresome. It would be nice to live in a world of Spocks sometimes, where we just observe and report, without value judgments. But that doesn't seem to be within the grasp of humans.

Joe, KQ1Q

Re my own point about the KX3, this block diagram shows it is an SDR. As Howard says it down converts to an IF, then digitizes and does all demodulation and modulation in the digital domain: www.elecraft.com/manual/KX3SchematicDiagramDec2012.pdf

Not sure I see the point about a reconfigurable control surface being part of the SDR definition. The KX3 is obviously an SDR yet it mostly has physical knobs which in practice cannot be reconfigured since the labels are physically printed. Likewise the BaoFeng UV5R -- whether the encoders themselves can be reconfigured is immaterial since the labels are printed. Similarly many military SDRs have minimal ability to reprogram the user control interface, yet they are clearly SDRs with all the other attendant benefits:  http://1.my.cdn.eesb.net/photo/3869/1416456302_896082_z.jpg

There are "knob/button" SDRs which have numerous soft buttons which can be reconfigured like the ADAT: http://www.adat.ch/img/adt200a_front.jpg However I don't see how a reconfigurable control surface is a mandatory part of the SDR definition.

KY6LA_Howard

My definition is:
Control Surface, whether a computer display or physical encoders, is Reconfigurable in software.

The point being that the software can change the functions of the knobs.....

KY6LA_Howard

BTW.. I am still working on the paper which i need to finish before Tomorrow Nite as I leave the USA on Sunday...

Joe, KQ1Q

My point is a software *controlled* radio -- where a microcontroller can reprogram the interface -- is separate and independent from a software defined radio which traditionally refers to detection, modulation and demodulation in the signal path.

W9OY

In my understanding the K3S is simply the logical extension of the TT Omni 5 from 1980.  It is clearly hardware defined and it's nature is to take the RF and continuously whittle and hone that RF down to a few Khz or hz bandwidth.  This is both it's strength and short coming but no matter how you look at it the blocks of the radio which takes the RF to a few khz are hardware and set in stone,  You can hang a little dongle off the end (DSP) and a PIC but in its major design it's an Omni 5.  

The Omni 5 was designed to take over the high performance ham radio market.  In the 70's the ARRL published some technical techniques to test for and compare receiver performance and the TT engineers realized if they could design a radio that beat everybody else on these tests they could advertise the **** out of it and sell a lot of radios.  They primarily did this with the introduction of the "roofing filter".   The other thing that happened is a few technically astute hams at the local ham clubs mastered this technical mumbo jumbo and the race was on regarding who could beat the numbers, as if beating the numbers somehow meant you had a better radio.  Pretty soon EVERYBODY needed a roofing filter.  I call these crystal radios because they rely on crystal filters and all the distortion and expense that introduces to achieve the design goal, which is to beat the ARRL tests and gain bragging rights.  Even the KX3 which a SDR and basically has the topology of the SDR-1000 3 board stack has a **** roofing filter, which severely limits its functionality.  

Gerald came out with a radio, the SDR-1000, that did the opposite of the Omni 5.  His radio takes a wide bandwidth of a couple hundred thousand HZ (192 khz) and digitizes it and puts the digital data in a database and then works mathematical magic on the data USING SOFTWARE.  The same 192khz of data is available on the antenna of the crystal radios, but all of the information in that 192khz is thrown away by the whittling and honing of the roofing filters etc.  Gerald realized "hey man I got all this data what can I do with it?"  The first thing he did was to demodulate and filter some of the data.  If you wanted to listen to 3900khz LSB, he plucked that data massaged it and turned it into audio for your ears.  This is dimension 1, something for the ears.

He said I got all this data what can I do with it, I don't want to just throw it away, so he  displayed it in graphical format aka panadapter.  This turned a 1 dimensional radio into a 2 dimensional radio.  You could now not only hear a narrow bandwidth, but see both the bandwidth you were listening too, and you could see + and - 100khz with all the stations contained in that expanded 200khz bandwidth.  Effectively the panadapter is a second receiver, a receiver for the eyes.  Next he added a third dimension.  He stored panadapter data over time and created a waterfall.  Now we could not only see bandwidth we could see history.  The waterfall shows us a third dimension, history as well as intensity over time.  Finally a forth dimension was added.  Decoders were created such as CW skimmer or the various digital interfaces.  This added the dimension of intelligence across a given bandwidth.  As opposed to a crystal radio which whittles and hones to a single signal, a SDR grabs the data and analyzes data, presenting to the operator information otherwise lost.  So while the legacy crystal radio crowd is left scratching over who has 1 dB better dynamic range, in a land where 1 dB better dynamic range MAKES NO DIFFERENCE, the SDR crowd is left to explore vistas and panoramas through the mathematical manipulation of data.

That is why the crystal radio crowd is doomed.  Owning a 100dB dynamic range radio when 85dB is adequate is a waste of time and money.  You don't have a better radio, all you have are more and more expensive bragging rights.  Who even says the "test" by which we measure radios is even current?  It was best thinking from 40 years ago for the modulation schemes and transistor components of that day.  Let's say ham radio moves to a digital based modulation scheme that works down to 2 dB above the noise.  Would a radio system that introduces all kinds of artifacts because of crystal filters even be relevant?  The only way the crystal radio crowd can get dimensions 2, 3, and 4 is to add a SDR to their system.  Larry Phipps N8LP was the first to realize this and he created the LP-PAN which is a sound card based (first generation in Howard's parlance) SDR that grabs the data and creates at least some of those other dimensions.  Even the vaunted KX3 is only 1 dimensional.  To get the other dimensions you have to add SDR to the "SDR".  The reason is a DSP sound card chip and a PIC computer which make up the business end of the KX3 doesn't have the horsepower to get the dimension 2, 3, and 4 jobs done.

I'm not knocking anybodies radio, just discussing how I think how a bright red line is drawn between 2 approaches HDR vs SDR and how it's disingenuous for a HDR to pretend it's a SDR because it's got a **** DSP dongle hanging off its audio chip

73  W9OY

Duane N9DG

I would say the evolutionary time line is more like this:

From mid 1950's -ish through the early 1980's most everyone had settled on a superhet design with an HF 1st or only IF around 5-6 or 9 MHz.

Then Drake introduced the TR-7 with its low VHF 1st IF. It dramatically fixed the problem of 1st IF image rejection. And allowed for much easier AM BCB or below to 30 MHz coverage with no gaps. Having SW BCB coverage in radios was a big “have to have” feature in those days.

The JA brands wanted this wide non-amateur band RX coverage in their transceivers as a selling point as well, so they too jumped on the low VHF 1st IF bandwagon. By the mid 1980's they were all fully up conversion 1st IF designs. The last of the JA brands with HF IFs roughly coincided with the last of the tube finals, and also roughly with the end of analog VFOs in those brands. And also - the begining of a decade-ish of really bad phase noise performance.

What Ten Tec did, or more accurately DIDN'T do was jump on the low VHF up-conversion design band wagon. At least not for the models they were targeting for top end HF performance, particularly closed-spaced IMD performance. They stayed with the 9 MHz first IF and implemented a 9 – 6 – 9 MHz IF scheme (Corsairs through Omni VI) to get passband tuning functionality.
 
And Ten Tec also DIDN'T jump on the fully synthesized LO bandwagon either for phase noise reasons. The Omni V and VI still used crystals for the band oscillators that was then mixed with a 5-5.5 MHz synthesized VFO that was designed for best possible phase noise performance. The fully synthesized LOs on the JA brands were really pretty lousy phase noise wise until the mid to late 1990's.

Then the Orion still stayed with a 9 MHz 1st IF, and mixed from there to get the 12-14 kHz IF range that allowed audio targeted A to D's to function at an IF frequency radio use. The LO in the Orions was fully synthesized, but by that point getting fairly decent phase noise performance wasn't terribly difficult anymore. Along about this time the term “roofing filter” became a marketing buzzword even though the actual architecture itself isn't that fundamentally different than what was commonly used from the 1950's onward.

What Elecraft did was recognize that Tec Tec was basically right for achieving the best performance for a  narrow (or as I call it “sliver-band”) architecture and then set out to refine the idea further. The K2 and then K3 essentially followed that plan.

Interestingly when Y and K realized that Elecraft and to a lesser degree Ten Tec was eating their lunch for HF contester use, and top notch receiver performance they reverted back to an HF 1st IF. But what they didn't realize was that the goal posts had moved in those intervening 30 years. Great sliver-band alone performance wasn't enough anymore. The PC-based SDRs had shown that it was possible to get very good performance with no traditional “narrow” IF at all. And to do equally so over as many receivers as you could spawn in the software. The sliver-band hardware designs are all fundamentally locked into forever being “narrow” frequency information recovered from the RF spectrum devices. And each additional receiver that you may want is only achievable by adding more duplicate sliver-band hardware.

And those sliver-band designs have yet to implement a truly good spectrum display or waterfall, they are all rather poor frankly. The wide-band SDR designs with their great spectrum displays and waterfalls that they have sort of came along as part of the package as a great side benefit of their wide-band architectures. The external add-on displays for those superhet designs take advantage of this as well.

FWIW I always see these raging debates about why “this radio”, or “that radio”, is or isn't a “true SDR”, kinda silly really. Because virtually all amateur transceivers being built today DO fit the strictest definition of “SDR”..... So what.

A more important and better thing to focus on, illuminate, and educate others about is whether one design is a “sliver-band” design or is a “wide-band” design. Basically I'd draw the line between “narrow” and “wide” at about 15 kHz. That 15 kHz or less of sampled bandwidth dividing line criteria basically encompasses all of the traditional superhet designs whether they are up conversion or not, DSP IF or not, or highly software configurable or not. Basically the same radios that are always being declared “not SDRs” by many of those who do use wide-band designs, and are declared “are SDRs” by those who don't.

And the SDRs that historically used sound cards, or A to Ds used in sound cards, and now increasingly the DDC designs are ALL wider than 15 kHz ish. Therefore are all “wide-band” designs. And perhaps an even more key design feature differentiation of “wide-band” SDRs from “sliver-band” designs is that they can have multitude of receivers in that sampled bandwidth. The sliver-band designs are all one RX (recovered audio) per sampled spectrum segment only designs.

Wide-band SDRs on the other hand  can accommodate “N“ receivers within their sampled RF spectrum, whether that sampled RF spectrum is only 48 Khz wide or is well over 100 MHz wide. They are all still fundamentally the same concept. Those design differences between the different wide-band SDRs are more akin to segregating superhet designs from each other based on whether they are either up conversion or down conversion designs, single or multiple conversions etc.. They are all just design, performance, and cost consideration trade-offs, nothing more, nothing less.

So that is why I also find the whole “1st, 2nd, 3rd, etc. generation” declarations of different SDR designs kinda silly too. They are certainly different design approaches, but hardly qualify as "generations". The wide-band SDRs all share a lot in common with each other, whether they use I/Q mixer designs or are DDCs.

W9OY

If you limit your vision of ham radio to a box on a table with a mic a key and a speaker there probably isn't much difference between a legacy radio with a dongle and something more software intensive.  If on the other hand you wish to bend shape and mold your ham radio experience and carry into the future having the radio in software is a tremendous advantage.  Having the software defined radio as the control nexus of an entire radio station is a tremendous advantage.  Having the radio as a server is a tremendous advantage.  Having the radio talking to the world over ethernet instead of 38k baud serial cables is a tremendous advantage.  Being able to rout data signals in software is a tremendous advantage.  Being able to control your radio remotely is a tremendous advantage.  All of this is the reason things like Maestro and SO2R will be achieved.  Better performance lower cost, more hams able to write software to effect change, more hams learning more about ham radio, and the leading edge of radio

Flex has or will have soon 2 6700R receivers flying in the ISS in a experimental program called Global AIS which is a world wide ship tracking and data acquisition project.  The 6700R works because it's I/O is Ethernet based and the "Waveforms" aspect of the software allows for the custom packet acquisition modem which addresses the Doppler issues of acquiring data while traveling at 7km/s to be placed in the radio.   The modem is incorporated in the 6700R like we  incorporate FreeDV.   Also implemented are what ever computer control is needed. (recall LEO's from the Oscar days)  I just don't see any K3's flying on the ISS for this kind of project anytime soon.  You aren't doing this with a dongle and a PIC.  This to me is what ham radio is about not just slavishly trying to acquire yet another dB of dynamic range. 

Duane N9DG

All entirely true.

But the reality is that the vast majority of the 6K hardware users will use them as some form of a “box on a table with a key, mic, and speaker”. And the insistence by so many that there be knobs and buttons control for them is just an extension of that reality.

If you sit back and ponder the 6K hardware, you come to the realization that the hardware itself is capable of capturing, processing, and outputting more radio spectrum "information content” than the human mind is capable of consuming. I don't think any sliver-band radio design, be it SDR or not, can viewed as having that level of capability. The sliver-band designs are all constrained by their RF hardware architecture choices to only digitally process a 15 kHz or less chunk of radio spectrum. Basically they are forever locked down by that narrow 1st IF bandwidth.

So the limitations imposed by their traditional UI approaches are never really noticed. Their UIs match their RF signal path limitations quite well. With wide-band sampling SDR designs that is not true. The UI, and the humans ability to consume its output become the limiting factors. So the 6K, and wide-band SDRs in general, are actually being constrained by the limits of UI implementation, and of a human's ability to interact with it. And the Maestro is in reality is just a UI implementation with its own set of constraints, it will certainly not be able to expose all of the power that is inside the 6K box itself..

Burt Fisher

"the 6K hardware, you come to the realization that the hardware itself is capable of capturing, processing, and outputting more radio spectrum "information content” than the human mind is capable of consuming",

you are right so why does anyone need a 6000 series with more than 2 slices?

Walt - KZ1F

Duane, whether it (Maestro) does is one thing. Whether it can is something entirely different. Watch this space.

k3Tim

"why does anyone need a 6000 series with more than 2 slices?"

Hi Burt:
I can think of many:

A slice for CW skimmer on 40 meters looking for gray line DX
A slice to monitor ship to shore of my container ship in the South China Sea (listening to Captain Karl is always 'interesting')
A slice for WeFax out of Japan for Wx reports in the same area as the container ship.
A slice to monitor HF air traffic control out of Halifax as my corporate jet is enroute to drop Howard in Europe.
A slice to have some nice Cuban music in the background (Radio Havana Cuba)
A slice to record a local MW broadcast station

How's that for an active imagination!

Burt Fisher

Ok, you could use it but what do you think about 90% of those who bought more than 2 slices?

Bill W2PKY

I could allocate 6 slices on 6 meters alone. My addiction with JT-65 on HF consumes at lease 3 slices or more watching for new DXCC entities. Started with a 6300, soon after traded for 6500 and now looking for a good deal on a 6700 for diversity listening this fall. 

ik2egl

tnx Horward for publishg your presentation, really a must see

Duane N9DG

“Duane, whether it (Maestro) does is one thing. Whether it can is something entirely different. Watch this space.”

Will the Maestro be able to do things beyond its more traditional knobs and buttons metaphor that it is? Undoubtedly yes. But when it is used by itself, it is still ultimately constrained by the smallish screen, and knob and button count. But its use need not be mutually exclusive of also using other UI presentation elements in parallel with, and external to it. That's the power of the 6K's architecture, you can do both.

“Ok, you could use it but what do you think about 90% of those who bought more than 2 slices?”

Whether they can or want to make use of those additional slices or not is largely irrelevant considering that they bought what they needed or wanted at a cost competitive price point relative to the other options out there that are much more limited. Think of those users as having bought a generic family car priced super hot sports car to only drive to church on Sundays with.

Walt - KZ1F

@Duane I was making a shameless (yet veiled) plug for what you'll see in a couple of months.

Ken - NM9P

I often keep a slice or two listening for band openings on six meters. Then I can have a slice listening for my buddies on our 80/40 rag chew frequencies, and use my fourth to look for other activity, either manually or via CW Skimmer. Or I can use two slices to work a DX op running split, while monitoring 6 meters on another and scanning around manually or with skimmer, or listening on WSJT-x. And I can use one pan all by itself with no slice at all, just to visually watch the activity change with changes in the MUF. Why? Because I can! And it interests me. There are lots of possibilities for more than two slices. It depends upon operating style and creativity.

W9OY

I think the comment about using only 2 slices is easily refuted.  In a pileup I routinely use 3 slices one for RX one for TX these in the same PAN.  The reason is SDR-Bridge is set up to be able to point click either on the pan or in skimmer and change TX freq.  I also have typically 2 skimmer waterfalls open because each skimmer fall analyzes the signals it receives slightly differently.  If I copy a call sign from the DX I can quickly scan the list of callsigns in both skimmer windows and instantly click to where the DX is listening or easily discern the pattern of his pileup style, so I have   It makes working DX trivial because of all the dimensions available for instant analysis. The point is my brain can easily analyze 5 PANs worth of data.  No way in **** you are doing this with a K3 and a dongle.  I also keep a PAN open (the 5th PAN) with no VFO which gives me a concise panorama of the entire pileup.  Data is a matter of presentation and analysis  I don't analyze the massive raw data delivered to the ADC, I analyze patterns and relevant data.  The 6xxx is a tool that allows this.  No legacy radio allows this.  

I can have yet another PAN open and watch a pileup grow on a different station.  I routinely watch European pileups grow on stations in the Indian Ocean as the terminator proceeds across Europe especially on 40 and 80M.  I know I am still several hours away from being in the fray but watching and listening to the DX working EU for a bit gives me a good idea of propagation and even the DX's pileup management style, well before my turn at propagation arises.  You also ain't doing this with a legacy radio and a dongle.  Recently BIN was added and W2RF started playing around with 3 dimensional spatial audio representation.  Imagine enhancing 3 dimensional audio in SO2R.  You might be able to train yourself such that you bring the audio focus from front to back as you switch between one band and another instead of right and left.  This may make you more efficient.  The result of this speculation is unknown but you sure as **** aren't going to get the chance to find out (aka push the contest envelop) with a legacy radio and a dongle.  With 2 slices you haven't even touched brain data overload.  With 5 PAN/slices you haven't touched it.  You simply have to make the PAN/slice report data to you in a way that answers a specific question like "what's the VR2XXX pileup doing" or "Is 6M open"  These questions don't require constant monitoring like the RX slice in a DX pileup, but they do require ready data for analysis once the question is asked.  

These questions can also be paired with software to make the radio more productive.  My antennas and amp band follow my radio  It means if I click 20M my radio/amp/antenna is ready to transmit on 20.  If I click 40M the whole shebang is ready on 40M.  I can memorize the settings of a pileup on 40 and a pileup on 20 and instantly switch back and forth depending on conditions, rarity etc (does this sound reminiscent of S02R?)  I have done this with 3 pileups and have worked 3 new ones in 10 minutes on 2 different bands by leveraging the software.  You ain't doing that with a legacy radio and a dongle I don't care if it does have MAN'S BEST DYNAMIC RANGE!!!!

73  W9OY 

Mike - N8MSA

I now have two 6700s; one is used for 8-band JT-mode spotting and one for general-purpose use. I will use 8 contiguous slices for pile-ups, centering each a few KHz away and varying the audio volume as needed. You have no idea how well this lets your hear pile-up calls until you have tried it.

I can think of other use-cases as well, but these are my primary uses for 8 slices.

73 de Mike - N8MSA

Bob - W7KWS -

Dayton is 1-day drive to half the U.S. population. Carrying all those boat anchors to Las Vegas on an airplane would require a very large shift in the current mind set.

VK7WH Winston

I agree, the general Dayton area does make sense, but surely there MUST be a better venue SOMEWHERE nearby. The current stadium has long since passed it's use by date. Dara may need to consider moving the Hamvention to a nearby city if a suitable alternative venue is not available in Dayton. Just my AUD 0.02 worth Winston