Why is SDR so expensive?

[jwlehman]

Please excuse my question if it seems ignorant to those well-versed in SDR, but I've been reading up on SDR and researching what's available and the prices seem high to me. Is it because it's a relatively new technology? You would think that cost would on average be less than what a stand-alone radion with the same capabilities would cost since the computer/software provides a lot of the radio's capabilities (processing power). Perhaps it the cost that developers expect from the front-end software packages that are on the market?

Regards,
JW

 

[smason]

SDR can be done on the cheap. With the really cheap RTL USB sticks selling for as low as $10.00
And the very good Funcube dongle for a little over $200.00 you can do a lot. Sure the "big boys" will outperform, but it doesn't have to be expensive.

 

[WA4HHG]

For the capabilities, SDR is fairly cheap. The $995 Microtelecom Perseus compares favorably with the $5,000 + Yaesu FTdx 5000 in the receiver dept. That's quite a value !

 

[prcguy]

Or the $999 Elecraft KX3 beats both the Perseus and Yaesu FTdx-5000 and its handheld size and runs on AA batteries.
prcguy

For the capabilities, SDR is fairly cheap. The $995 Microtelecom Perseus compares favorably with the $5,000 + Yaesu FTdx 5000 in the receiver dept. That's quite a value !

 

[kilokat7]

Or the $999 Elecraft KX3 beats both the Perseus and Yaesu FTdx-5000 and its handheld size and runs on AA batteries.
prcguy

Please explain how the Elecraft KX3 "beats the Perseus". Inquiring minds want to know.

 

[prcguy]

Rob Sherwood is a recognized test lab that runs various HF receivers and transceivers through tests that determine their ultimate performance. He then rates them based on 2KHz narrow spaced dynamic range, which is probably the most difficult to achieve and determines which receiver would best pull out a weak signal buried between close interfering stations.

Sherwood rates the Elecraft KX3 #2 on the list and above just about all the megabuck radios on the planet except a $20k Hilberling unit made in Germany. I've used many many of the top radios on the list and can say the Elecraft KX3 deserves to be on top of the list, its receiver performs well above most radios ever made anywhere at any time.

I agree that the Perseus is no slouch and performs in the upper five or so that Rob has tester, however I prefer a radio with actual knobs that is not tied to a computer.

See: Receiver Test Data

Please explain how the Elecraft KX3 "beats the Perseus". Inquiring minds want to know.

 

[SCPD]

Or the $999 Elecraft KX3

It's hardly $999.

Add up the required options and it's more like $1500 or thereabouts.

 

[SCPD]

Rob Sherwood is a recognized test lab that runs various HF receivers and transceivers through tests that determine their ultimate performance. He then rates them based on 2KHz narrow spaced dynamic range, which is probably the most difficult to achieve and determines which receiver would best pull out a weak signal buried between close interfering stations.

Sherwood rates the Elecraft KX3 #2 on the list and above just about all the megabuck radios on the planet except a $20k Hilberling unit made in Germany. I've used many many of the top radios on the list and can say the Elecraft KX3 deserves to be on top of the list, its receiver performs well above most radios ever made anywhere at any time.

I agree that the Perseus is no slouch and performs in the upper five or so that Rob has tester, however I prefer a radio with actual knobs that is not tied to a computer.

See: Receiver Test Data

The test data notes that the optional roofing filter is required. So this isn't the default $999 package.

 

[SCPD]

Please excuse my question if it seems ignorant to those well-versed in SDR, but I've been reading up on SDR and researching what's available and the prices seem high to me. Perhaps it the cost that developers expect from the front-end software packages that are on the market?

JW,

First off, it's not an ignorant question. It's a fair question and this is what kept me out of the game for those first few years of the SDR evolution.

I've owned a Perseus SDR for over a year now. I (did) own an Icom 7600 which was my HAM transceiver but also my primary SWL receiver as well. I didn't think there was ANY way that this little box on my desk could replace my Icom 7600 -- but it has -- and for good -- as I just sold my Icom 7600.

I have HF in the SUV and I found myself using the Icom less and less. As others have heard me spout on the #wunclub channel, the SDR experience can really change how you view radio and the RF spectrum. It's definitely worth it. The speed at which I can discover signals has jumped 10-fold. With more experience, you can learn to classify signals in about 1-2 seconds. I remember my early days of SWL where it would take me (a novice) at least 5 minutes before realizing the difference between SITOR and RTTY.

For the cost question, here's what it boils down to and like everything in life, you get what you pay for... but that axiom is sort of starting to change:

1. Quality of electronics/hardware
2. Quality of software
3. Customization (optional features, add-on boards, etc.)

For #1, this is vitally important with respect to how fast the SDR can process digital information. Basically, the top-end SDRs have a FPGA unit -- which is a fancy name for a microprocessor that is very advanced and is essentially a "computing device." The quality of the FPGA itself is important. XILINX makes a good percentage of these FPGA units and a single chip can cost anywhere from $50 to $15,000. The latest units are extremely expensive. In very simplified terms, think of the difference between the old 8086 computer processors and the latest Intel Core i7 processors.

These higher end units are used in multiple applications including military, defense and of course, communications.

If you're a HAM, then take a look at the latest FLEX SDR transceivers. Their latest model uses a XILINX version 6 chip which is quite expensive. (The cost of the SDR transceiver reflects this as well.)

Not all SDRs are created equal. Some require the SDR to be interfaced into a soundcard. The higher-end models do not. Some require external power and some get power from the USB port.

In general, the higher the cost means better performance and better software of the SDR.

For #2, this is what pushed me to the Perseus. The Perseus application is very simple and by (2012) standards, isn't even all that sophisticated. That being said, there's something very nice about simple -- speed. If you're about hunting for signals (like me) then you can understand how quick acquisition and analysis (which takes all of 2 seconds) is a huge benefit. The design of the software interface plays a huge part here. Some prefer complex designs and others prefer a clean design. Complex designs may offer other benefits such as signal decoding, analysis, etc.

There are free versions of SDR software available and some are quite good. Again, you get what you pay for...

For #3, this is very subjective and only important to those who use a SDR for something other than RF monitoring. The SDR can be used in many ways in the field of RF electronics. These options though can increase cost dramatically.

Top SDR Models (for receive only)

Perseus
WinRadio Excalibur
QS1R
SDR-IQ
Funcube Dongle Pro+ (HF, VHF and UHF)

(There are other brands/models but I just listed the most popular ranked by cost.)


Hope this helps,
-Nick

 

[prcguy]

For general SWL use the stock configuration will still beat just about any radio out there for $999. If you want it to perform the same in the Sherwood tests yea you need the filter and that totals about $1030 but you don't need the other options for SWL.

I was using my KX3 today on vacation (outside in 30deg weather) and it never ceases to amaze me how well it performs.
prcguy

The test data notes that the optional roofing filter is required. So this isn't the default $999 package.

 

[Token]

Please excuse my question if it seems ignorant to those well-versed in SDR, but I've been reading up on SDR and researching what's available and the prices seem high to me. Is it because it's a relatively new technology? You would think that cost would on average be less than what a stand-alone radion with the same capabilities would cost since the computer/software provides a lot of the radio's capabilities (processing power). Perhaps it the cost that developers expect from the front-end software packages that are on the market?

Regards,
JW

SDR is not expensive, in fact for what it brings to the table it is pretty inexpensive. Lets take for example the $45 Softrock Ensemble II, maybe $100 after adding a power supply and a nice case.

While not a fan of sound card based SDRs you have to admit that no past HF receiver ever sold offered the capabilities found in the Ensemble II for so low a cost, performance per dollar if you will. For the price of an entry level portable you get performance similar to a $600 table top RX, plus the flexibility of the SDR filters, waterfall/spectrum displays, and ability to display and record up to 192 kHz of bandwidth. And there are several examples in this performance range from $45 to $300. Combine this with a $500 PC (and you can get PCs that will work for this application cheaper than this) and you have performance exceeding the same cost in a tabletop radio.

Stepping up in performance you go to the DDC SDRs, not sound card based. These start in the $250 range and go up from there. To be sure, top end units seem at first glance to be expensive, peaking at around $1600 for HF only coverage, but that is pretty much because there is little to compare them to. Today you simply do not find mid to upper end conventional hobby receivers.

The JRC NRD-515 was about $1300 in 1981, call that $3200 in buying power today. The McKay Dymek DR44 was selling for $1750 in 1983, or about $4000 purchasing power today. And neither of those were “pro” gear, like the expensive Rockwell-Collins stuff or the upper end Racal. Even the Kenwood R5000, at $1000 in 1992, would be about the same as $1600 buying power today.

But today’s $800 to $1600 HF only SDR is simply better in almost every measurable way from 30 years ago $1000+ radio. Sure, you can talk about wanting to have knobs, not be tied to a computer, etc, but that is preference, not actual performance.

So taking a $1000 SDR and using it with a $500 PC (real world example, my $500 low end Gateway PC runs both a WinRadio Excalibur and my Perseus) results in a better receiver than you could buy for the same number of dollars in 1980, let alone the value of those dollars today.

The issue I have with SDR (and I have firmly embraced the technology, so obviously this issue is not to much of a biggy in my mind) is going to be the life span. Sure, I get high performance at a reasonable price. But my 1940 Hallicrafters SX-28 still works just fine today, will my G31DDC be working in 2082? And if it is will I (or whoever owns it, I will be long dead) be able to find a PC and OS to operate it? I would not bet on it at all. I look at the SDR of today as a 15 year lifespan item, I expect to be able to find hardware and software to support it for the next 15 years, anything after that is gravy.

T!

 

[Token]

Top SDR Models (for receive only)

Perseus
WinRadio Excalibur
QS1R
SDR-IQ
Funcube Dongle Pro+ (HF, VHF and UHF)

(There are other brands/models but I just listed the most popular ranked by cost.)

Nick, don't forget the Afedri SDR. I just this past week got the chance to play with one for a few days and I think I have to rate it above the QS1R or the SDR-IQ. I am thinking about grabing one to replace my spare SDR-IQ on my SDR-Radio server. Also, the NetSDR and the SDR-IP need to go up there in the Perseus / Excalibur area.

Of course, you know I am not going to agree with the order you put the SDRs in , I am convinced the Excalibur exceeds the Perseus. I know it does in my own bench testing.

T!

 

[prcguy]

To say the Excalibur exceeds the Perseus would mean its in the top running for the best performing receivers ever made. Is that what you mean to say?
prcguy

Nick, don't forget the Afedri SDR. I just this past week got the chance to play with one for a few days and I think I have to rate it above the QS1R or the SDR-IQ. I am thinking about grabing one to replace my spare SDR-IQ on my SDR-Radio server. Also, the NetSDR and the SDR-IP need to go up there in the Perseus / Excalibur area.

Of course, you know I am not going to agree with the order you put the SDRs in , I am convinced the Excalibur exceeds the Perseus. I know it does in my own bench testing.

T!

 

[SCPD]

Nick, don't forget the Afedri SDR. I just this past week got the chance to play with one for a few days and I think I have to rate it above the QS1R or the SDR-IQ. I am thinking about grabing one to replace my spare SDR-IQ on my SDR-Radio server. Also, the NetSDR and the SDR-IP need to go up there in the Perseus / Excalibur area.

Of course, you know I am not going to agree with the order you put the SDRs in , I am convinced the Excalibur exceeds the Perseus. I know it does in my own bench testing.

T!

Yes I left out a bunch that I couldn't think of off the top of my head. There are several others out there as well that just don't have the following yet. In time, we'll have a lot to choose from and like with the desktop receivers, everyone will have their subjective viewpoints as to which is better. We humans love to argue.

With SDRs however, I think it's very important to note that the software can really affect its usability. Great SDR hardware is useless without quality software. You may have great software but it's useless without decent hardware. Granted, there's a middle layer developing where you have mediocre hardware with mediocre software and for the price (< $30) some people believe THAT is what makes a great SDR.

So yes, it does come down to money and how much you're willing to spend. SDR can be "expensive" but it is affordable -- and that is directly relative to your own pocketbook.

This reminds me of the old rec.radio.shortwave newsgroup days where we could spend days arguing over which SW receiver had the best VFO design.

:roll:

 

[SCPD]

I was using my KX3 today on vacation (outside in 30deg weather) and it never ceases to amaze me how well it performs.

FWIW, I'm not bashing the KX3. I think it's pretty cool and wouldn't mind one myself.

I hope Elecraft joins the receiver-only market. They could probably come up with a very portable, high-performing SW receiver.

 

[Token]

To say the Excalibur exceeds the Perseus would mean its in the top running for the best performing receivers ever made. Is that what you mean to say?
prcguy

If you are rating receivers based on the Sherwood Engineering list remember that list is sorted on “Dynamic Range, Narrow Spaced” and no other factors. As far as I know it is not a claim of best to worst, other than narrow spaced dynamic range. In my opinion narrow spaced dynamic range, by itself, does NOT define “best”, there is no one factor that determines how good an RX is. It is a combination of specifications that make up how good the RX is, so I do not agree the Sherwood list, in the order it is presented, is necessarily the “best to worst” receivers. To be sure, narrow spaced dynamic range is very, very, important, and better receiver designs will generally have better narrow dynamic range, so at the top of the list the order is about right, but the further down you go the less correct the order is.

I have several radios on the list, and I can say for sure that my Heathkit SB-104 does not outperform my Yaesu FT-2000 in any real world way, in fact the FT-2000 blows the SB-104 away. And yet the Sherwood list, the way it is sorted, has the SB-104 transciever 49 positions above the FT-2000. For that matter, my Yaesu FRG-7700 does not outperform my FT-2000 in any way I can tell in actual use, but it is 6 places above the FT-2000. Of course, I like that my FTDX-5000 is near the top of the list sorted the way it is

With regards to your question, the G31DDC Excalibur, on my test bench and with my copy of the Perseus and Excalibur, come in at about the same narrow dynamic range, possibly a dB or so delta between them with the Excalibur on top. Also, the Excalibur has better MDS, better noise floor, and better sensitivity. When selecting very high values for the “demodulator filter length” (a setting in the Excalibur software) the ultimate filter rejection of the Excalibur exceeds the Perseus. The LO phase noise of the Perseus is very, very, slightly better than the G31DDC, I suspect unit to unit variations would actually make this too close to call.

So yes, the G31DDC Excalibur would be in the running as one of the best performing receivers, as defined on that list, probably just above but possibly just below the Perseus based on the one sorted value. And the G33DDC Excalibur Pro is even better, as it has better sensitivity and lower noise floor than the G31. In use I have seen the Pro outperform the non-Pro. I have not had the Pro on the test bench to measure it myself yet, but suspect its narrow dynamic range might be better than the G31, and that would also place it, on the Sherwood list, in the top few receivers.

I can say that in side-by-side use, outside the lab and in the real world at my listening desk, that the Perseus and the G31DDC Excalibur are very close, but I would have to give a very slight advantage to the Excalibur. It just seems to pull out the marginal stations a bit better.

T!

 

[jwlehman]

Wow! I really appreciate all the detailed, and well-thought out responses! Very insightful!

Thanks,
JW

 

[FreqMeister]

It's kind of sad when a person asks a reasonable question and over half the responses in this thread are essentially "Who says your Ford is better than my Chevy?" The O.P. wasn't asking if a Perseus outperforms an Excalibur or a KX3.

As a few have pointed out, you generally get better performance, and certainly a better feature set, for less money with SDR than with conventional receivers. That said, if you want a true wide bandwidth direct conversion SDR (vs one using only soundcard bandwidth) with high dynamic range the A/D converter required, and supporting circuitry, are not cheap. While the signal path is a lot shorter than in a conventional receiver, it requires some serious digital hardware and a high bandwidth connection to the PC and/or a relatively costly high bandwidth DSP engine.

The Afedri SDR-Net is one of the least expensive direct conversion HF SDRs with a respectable bandwidth I know of at around $250 but it doesn't have very good dynamic range and is susceptible to overload problems depending on where you live. As others have pointed out, prices start at about $25 and go up from there.

SDR, in my opinion, is still in its infancy. Winradio has been around perhaps the longest, but their most affordable products are not true SDR (as defined above). A lot of the newer options are from relatively new small companies and are made in low volumes which makes them more expensive--especially where a lot of R&D was involved. The manufactures have to recover their development labor over a relatively small number of units sold. I'm not sure how big the overall market is, but if a company like Uniden starts mass producing a decent wideband SDR in China it would probably be very reasonably priced.

The open source products have the advantage of "cloud sourcing" the otherwise expensive software R&D required for SDR and/or they use existing (usually soundcard-based) software. But the hardware is often lacking compared to a good dedicated conventional receiver. Good analog RF design is a relatively rare skill and requires expensive test equipment. A lot of the low-end SDR hardware I've seen seems to suffer from substandard RF front end performance (like the Afedri and cheaper options).

One has to wonder how long the big players, i.e. ICOM, Yaesu, Kenwood, etc., are going to stick to the old paradigm? They have the RF expertise, and economies of scale, but so far haven't offered a true direct conversion black-box SDR. Apparently they would rather try to sell you a $13,000+ radio like the IC-R9500 but those days have to be numbered.

 

[SCPD]

It's kind of sad when a person asks a reasonable question and over half the responses in this thread are essentially "Who says your Ford is better than my Chevy?" The O.P. wasn't asking if a Perseus outperforms an Excalibur or a KX3.

Funny, I think it's pathetic when a new, anonymous member, with a whole five posts to their credit, tries to lecture other members. :roll:

 

[prcguy]

He may only have 5 posts but I agree with what post # 5 says.
prcguy

Funny, I think it's pathetic when a new, anonymous member, with a whole five posts to their credit, tries to lecture other members. :roll: