What Are The Practical Limits On How Good Can An HF Receiver Get ?

Paul - WB5AGF

I've been going through some of the ARRL Product Reviews, and Rob Sherwood's latest receiver numbers, and what seems quite remarkable is how much better the numbers are for some of the more recent radios (thinking of the the FLEX-6000 and Elecraft's K3S) compared to what was considered quite reasonable a few years ago.

I think what really surprised me personally is the (comparatively speaking) unimpressive numbers yielded up by the Drake '4-line' receiver design; there was a time during which my R4B figured prominently in my daily life when I was stationed overseas (the day would not have been complete without getting the news from the VOA and the BBC).

I'm left wondering how much more close-in dynamic range can be squeezed out of chips that can be put into designs that we hams can afford ?

(Didn't someone look into using active devices that would normally have been considered part of a transmitter design in an attempt to cobble together a receiver with an almost uncrunchable front-end ? I believe that it was the Squires-Sanders receiver of the 1960s that put a 7360 mixer as the first active device.)

- Paul, WB5AGF

Bill -VA3WTB

This goes for me as far as all technology, I wonder if and when we hit a wall.

Stan VA7NF

The "wall" that was the super-het has been breached, and a new wall is appearing in the distance.  It may take some time before we stand at the base of this new wall.

Ken - NM9P

Duane_AC5AA

As I understand it, the dynamic range numbers are very real if you intend to operate in a multitransmitter environment, such as a big contest station. Or, if you happen to live next door to a ham who likes to frequent the same bands and modes you do. I have a neighbor two streets over who would clobber my (excellent) receiver with his TS-430 if we were on the same band. I rejoiced when he got a TS-570 (I think it was) with a much lower phase noise figure. Yes, the transmitted phase noise was the culprit. Likewise, I can now be on 20m while he's running around 800 W on digital modes because my 6500 does a good job of not responding to his strong signal. So, it all depends on what numbers you're improving and whether or not it affects "my use" on the air. For me, with poor antennas on a city lot, I need a good, sensitive receiver that is as impervious to my neighbor as possible, and the Flex certainly fills the bill. Better even than the Ten-Tec Orion that it replaced.

WA2SQQ

To some extent aren’t many of the improvements technology is providing bringing us closer to the point where we can measure the improvement, but it’s extremely difficult to easily realize the improvement?

 My first PC was an IBM running at 4.77 mhz. I upgraded to a 20 mhz “286” CPU which was four times faster. That was a very noticeable speed increase. Today, the technology has advanced to the point where CPU’s are running at GHZ speeds. While we can benchmark incremental increases, the actual user's perception is barely noticeable. Given the amount of RF pollution that is present, isn’t all this added sensitivity / selectivity less significant (for most locations)?

Ken - NM9P

Agreed.  I meant to mention RX & TX Phase noise in my comments.  These have indeed been some of the greatest improvements in practical close-in dynamic range issues in the last decade or so.  My 6500 is so much less affected by three neighboring hams less than a mile away, two of them running over 1 KW,  than even my old Kenwood TS-850SAT.  In fact one of those nearby stations is running my old rig and I can see his pattern.  It is very nice, but I couldn't get as close to his frequency as I could if He were running one of the newer rigs, or especially a 6000 series.

But they are much more affected by my 100 Watt signal when I am on the band than I am affected by THEIR signals....So I give them a little more room.  The neat thing is...I can SEE where they all are and give them space, or at least wait until they are done with a DX contact before I chime in.  The panadapter helps me be a better "neighbor!"

Bill -VA3WTB

I agree, It seems to me some things just have to be understood before technology takes off.
I read some were that technology has advanced faster in the last 30 years then it did in 200 years before that. Because things were understood for the first time it led to new directions. I wonder were we will be next time we understand something we don't get yet?

Walt - KZ1F

At some point though, the practical limit, as we can already 'hear' through the noise floor, is when the atmosphere just does not reflect a signal our way. I think that is the hard limit, for HF at least. For UHF, not bound by the ionosphere, we can bounce signals off of satellites, man made or otherwise. Of course there, UHF has an issue with passing through organic matter, like foliage and water. Of course, as someone here recently stated, FRS surely is working on telepathic modulation. I, actually, am not holding my breath for that, nor am I holding it for subspace communications, due sometime, as I understand it, 2350.

Given that, my guess is we're already talking about dancing angels on the pin's head.

Bob-N4HY

Here is one person's opinion.  I have been at this since I first bought the Elecraft K2.  I feel that Wayne Burdick showed that we could make major improvements and Burdick made it reproducible in a kit!!  Since then,  I have always sought to better a receiver.  I did it for Uncle Sam for years and the learning that came from those experiences were incorporated by Flex into the 6000 series when Gerald, Steve, and I began talking about the future.

We are close to the fundamental physical limits on receivers.  Our transmitters are about to follow with predistortion.  I have not yet calculated the impact on phase noise but there is a reason the most expensive single component in a Flex 6000 series without a GPSDO is the oscillator.'

We are close to the end and now it is time to work on software, features, future,  decreasing cost of components, etc.

Remember, this is one person's opinion, I am so on the edge of Flex decision making these days that the first I knew of Maestro was the day I could put a deposit on one.

I am speaking for myself and not Flex.

Bob
N4HY

Walt - KZ1F

Actually Bob, given your stellar credentials, you are an excellent person to weigh in on this topic. I actually got thinking about this some months back vis-a-vis at what point is propagation the final brick wall. Yet another reason I miss the days of the HEO birds, which, btw, perhaps the last remaining one is or shortly will be en-route to Blacksburg.


Ken - NM9P

Don't remind me, Walt!  I was severely bummed out (a highly technical medical/psychological term...) when the Phase III Oscar 40 bird had a major malfunction when they tried for the second stage of orbital expansion.  I had been waiting many years and purchased some new equipment in anticipation of launch.  I had actually copied the satellite when it was in low orbit at an S-9 on my IC275A with only a 19 inch wire in the back of the rig as an antenna as it sat on the ledge next to the window in my ham shack.  I had just moved and hadn't even unpacked my shack yet.  Then not long after that, BOOM!  No more hopes for a high altitude Phase 3D bird.....It would have been glorious!

Walt - KZ1F

Yet there is hope. AMSAT-DL has the last remaining 3D bird and is shipping, or has, to UVa in Blacksburg to be launched at some unknown point in the future. Personally, I am not going to spend a lot of money to work flying repeaters. When we lived in RI a guy came up to have a, I guess, grid square, dxpedition from our back yard. That was the primary reason I went from Advanced to Extra as Extra can be control ops. I was hooked.

KY6LA_Howard

Yes the laws of physics have practical limits...

For some factor we are already beyond the practical limits

One of my favorite specifications is MINIMUM DISCERABLE SIGNAL.. MDS

Radio manufactures play the specification bragging game claiming they have the best MDS with ranges below -140dBm or ever better.. yet for most practical purposes such as receiving in a city on say 20M, your background noise level will be at least -120dBM or greater.. so with -140dBm you are just amplifying the noise... hence.. we are already past the practical limit for that specification.

Another is Dynamic Range.. with Legacy Technology achieving dynamic range greater than 85 dB is actually pretty hard to do.  you need all sorts of roofing filters and other techniques to push the range higher ... With SDR technology achieving dynamic ranges higher than 85 dB is much easier... the dynamic range is done is math and it is a direct function of the number of bits in the A/D converter.  So a 16 Bit A/D gives you 96 dB of dynamic range...  So how does Flex with their 16 Bit A/D produce Dynamic Ranges over 100dB.   they use a very fast A/D converter and over sample each bit so effectively they are simulating a 20 bit A/D....  Already in the audio world you can find inexpensive 24Bit A/D giving 144dB dynamic ranges... it should happen in the RF world soon too...

However.. what it the practical reality...  Sherwood states that most hams in normal use likely only need about 85 dB of dynamic range...   Realistically if you are in a contest or have a strong nearby signal every dB of dynamic range will help but for the rest of the time... we are likely already there...

Walt - KZ1F

So that makes sense in the context of what an FPGA can do. How do non-FPGA radios achieve the > 100db dynamic range, same A/D converters?

That was a good description Howard...thank you for that!

KY6LA_Howard

FPGA is the computational engine that has little to do with the A/D converter other than being a fast part that does the math.

Flex for example uses a 16 bit part - by definition 96 dB Dynamic Range.. But the A/D part works at around 245 MHz so it effectively oversamples a 65MHz signal 4 times for each bit of resolution.. add a little math and you get 108 dB dynamic range.

Here is a link to a good discussion which does not oversimplify things

http://www.eham.net/ehamforum/smf/index.php?topic=84676.30 

It will help you become "Sophisticated"  Hi Hi

Walt - KZ1F

That was a legacy inclusion from prior conversations we've had re: SDR will revolutionize HamRadio...the only thing I could fathom to support that was cost of manufacture. I think this topic dovetails very nicely into that prior conversation. If we are already at the margins, super duper electronics still can't overcome the laws of atmospheric physics. So, what would that leave but cost??

km9r.mike

@ Bob

I absolutely loved the receive performance of my K2. Perhaps the first time I came across a rig that I could contest all weekend with w/o fatigue inducing receiver noise. What a pleasure to use. Too bad it's filtering was not stronger. The Flex SS is of course like the K2 on steroids (but in a good legal way). Hopefully much features to be added in the areas you noted. 

For next gen, what about the possibility of utilizing dual adcs in a bose noise canceling configuration for noise mitigation or would it just be more pratical to tell the om on the other end to qro?

Bob-N4HY

I fear that it is hardly possible to get analog parts to go that far. Direct sampling is the ultimate and with a bigger FPGA we could improve on the DR dramatically and other fancy things but it likely makes no sense to do so as Howard has pointed out

KY6LA_Howard

To further amply Bob's point.. look at the Hilberling €18,000 to achieve a marginal improvement over the best legacy radio and yet still not as good as many much less expensive Direct Sampling SDR

Costs.. SDR Parts will ultimately get cheaper - so hardware costs will drop

HOWEVER- Walt - you are the expert - Software costs keep on climbing... so savings in hardware seem to be quickly eaten up by software.

Can we push things a bit farther.. yes.. lots of brilliant software ways will emerge.. the EE journals are full of new brilliant ideas on how to squeeze out a little bit more...

But to what end...

Not sure any human except some superhuman at WRTC can actually make use of it...  

OTOH.. if we had robots doing comm (WSJT-X  or Bob's latest satellite for example) those extra few dB might still mean something....

Ken - NM9P

As many of you have said, MDS and Dynamic range have achieved amazing performance at or beyond practicality.  Phase noise is getting there and already is present on many rigs.
Ultimately the next practical frontier is going to be major refinements in noise reduction, signal enhancement, diversity reception, beam steering, multiple receiver geo-diversity, and increasing the speeds of these processes and reducing latency.  As processors and A/D converters continue to increase in speed and power, we will see amazing things in this frontier.  I wonder what is waiting in the wings with what Bob, Gerald, Steve, and all the others have learned from their "government" work?  (assuming some of the techniques are not restricted to government contracts....)

KY6LA_Howard

@Ken... a brilliant thought provoking post

Walt - KZ1F

Howard, that is completely a false equivalence. Now in the 6000-series or take your pick the 6700 6500 6300 there's one fpga I think and so each unit that ships has stuff in it that's distinctly different than software where the software is written once and then copied so it's kind of like the first color tv was millions of dollars and each successive one was way less than that's a bad analogy because those are hardware well but the software yes for the very first one is very expensive when you put it in a couple of thousand or tens of thousands then it's that cost divided by a couple of thousand or more. That is what is the basis of the suspicion of cost being the primary benefit of SDR

Paul - WB5AGF

"... the suspicion of cost being the primary benefit of SDR"

Walt, I'm puzzled by that comment (many things puzzle me ... this is just one more).

If ease-of-manufacturability was the driving force behind the use of SDR design concepts, and not improved performance, then why are the close-in dynamic range numbers so much better for recent designs ?

- Paul, WB5AGF

Bob-N4HY

Noise mitigation is always at the forefront of what we think about. I'm personally more interested in trying to move us into the future of using lower power and digital encoding of everything. Joe Taylor was and is a good example, a good friend, and my prime example of someone who was ahead of their time and still angers people who work CW. We can go multiple ways in the future and I'm certain that given what I know but am not at liberty to say, Flex was and will be one of the leaders on the forward path. They are superb engineers, business people, and awesome at serving the users and always looking to get better.

73s Bob N4HY

Kevin Va3KGS

Pokémon Go

I started with CB Radio and Fox Hunting.  Today its Cell Phones (Public Transceivers) looking for virtual intelligence. Technology improvements will keep us inspired for years to come.  Wait for these young people to figure out what we old timers do for fun.  Happy Hunting!!

Kevin, Va3KGS

Walt - KZ1F

@Paul, how do you get from my comment, primary beneficiary of SDR being costof units to ease of manufacturability? This is my conclusion based on what likely was a 2+ year conversation with Howard to explain to me what the incentive was for sdr. It's not like B&W vs color, it's more like plasma vs OLED. And even that is likely a better case than legacy vs SDR. This, btw, I think is an outstanding technical ham radio conversation. So FRS, apologies in advance. Look at Rob Sherwood's list, specifically the top two. Cost wise close and performance wise virtually identical. The only difference is technology. Yes, I know everyone has their personal favorites but if you stipulate Rob is an unbiased researcher, the only logical difference has to be in cost. Having said that, features aren't free, the 6700 can do things the K3S can not do. But from the perspective of atmospheric physics, the two technologies appear to be neck and neck. Now, again, apologies, look at the 7300. IITs in the top 10 of in production radios on Rob's list and meets or exceeds Rob's threshold performance floor as far as atmospheric physics. Agreed, compared to the 6700 it is feature poor, but this thread is about technology practical limits. Oh, the 7300, in that YouTube vid appear to be statistically indistinguishable from the 6300, at a significant cost savings. Again, not talking feature sets here. Acknowledgement, the 7300 also wasn't built in Austin or SoCal. But, for the guy that doesn't pour hu dress of thousands of dollars into their hobby.... We're talking practical limits not which has the better features of is more politically correct. Yes, everybody bemoans we should put Americans back to work. They do this on their way to Walmart.

WA2SQQ

In digital photography there are two types of people who buy a camera. First, there is the customer who wants a camera to take good photos, a real photographer. We also have the customer who examines every photo at 300% of the normal size, looking for artifacts that 99% of the people aren’t even aware that they exist. This is what we call a “pixel-peeper”. We hams fall into two groups also – those who bought the radio to chat and enjoy the hobby (me), and those are just obsessed with the quest to have the best specifications that money can buy. Maybe it’s time to sit back on 75m and smell the coffee? Time to chill out and be happy your name wasn't listed in this month's "Silent Keys"!

Walt - KZ1F

Yet this discussion is on practical limitations. For myself, I wouldn't mind driving a multi-hundred thousand dollars car. But the analogy would be what is the technological difference between them and a Chevy Volt. And, since you mentioned it, I am thrilled my name is still not listed under silent keys.

Rick Hadley - W0FG

To chase that birdwalk a little further...the first thing I do when I get a new QST is check to make sure neither I, nor anyone I know are listed there.  At this stage of the games (ham radio AND life), I can't conceive of needing a better receiver than that of the 6500.  I'm just hoping to be able to hold on for another full sunspot cycle before the Lord makes me go QRT.

KY6LA_Howard

As Walt said, we have had this discussion for a couple years now off the board... Basically within the very narrow confines of MDS and Dynamic Range and Phase Noise,- Legacy radios have pretty well hit their practical limits in that it becomes extremely expensive to use analog parts to squeeze slightly more performance out of them.. OTOH.. SDR have the advantage of being digital and which means that the some features such as Dynamic Range are really a math function which can be improved beyond the human ability to discern them for relatively lower costs.

HOWEVER... SDR's do offer lots of future opportunities to do a lot more than currently being done by Legacy Technology...Bob hit on the concepts of noise mitigation where the ability to add new noise mitigation mechanisms is rather limited in Legacy technology..

But there are other areas where SDR's are superior.. just look at the fact that Flex added several new modes such as DStar --- and of course the available API so that users can add features and functions they dream up...