Apache Labs ANAN-200D a competitor?

Greg Johnson

I think that flex has a nice concept going with the thin client. That is how they manage. The **** client is like the new electric car that has a motor over each wheel that no one understands how it works. Psdr is just not the same. If it was, it would work for existing flex radios. psdr will be replaced most likely with a transmit version of cusdr.

The **** client focuses on a few things thin clients cannot compete with, and thin clients focus on a few things that **** clients cannot do.

Enter the gigabit protocol for **** client, now you have 3 x the information moving in that pipeline. A set of processors that evolve with each new computer purpose. With the addition of the add on computer to take the info from the whole spectrum and put it into a server based network solution protocol will allow the most basic Anan to have 7 open receivers with one dac .

I don't consider these radios to be in competition with each other. They are both different. Enjoy the strengths of each. The software used to write for the **** client is open source and moves much faster. This is why it evolves so fast.

I have never had a unsolicited good audio report from the flex. I have from the cw side.

Greg kc8iir

Joe, KQ1Q

The debate between embedded vs client-side approaches and implications for thick vs thin clients are broader than ham radio. There is a continuing tension in the digital design community between special-purpose embedded processing (DSP, FPGA, etc) vs general-purpose processing via CPU, which in this case means a **** pipe or client. 

The old view is DSP/FPGA hardware is much faster at signal processing than a general-purpose client-side CPU. However recent Intel CPUs can be faster than some DSP or FPGA hardware. This is a result of continual progress in instructions per clock efficiency, higher core counts, and better floating point and vector instructions. E.g, the 3.5Ghz i7-4770K is benchmarked at 182 Linpack GFLOPS if using AVX2 vector instructions. That's nearly 2000x faster than the original Cray-1, and modestly faster than the Virtex-6 FPGA used in the Flex 6000 series. This would seem to drive the design decision toward client-side implementation.

However several offsetting factors justify a server-side, thin(er)-client approach. While new CPUs are fast, older ones are less so. It is very hard to make real-time software run well across a broad range of x86 CPUs with widely varying performance and instruction features. Also general-purpose operating systems are often not well suited to real time applications. Even if the CPU has the computational horsepower it bogs down the entire computer doing that, leaving little left over for logging, email, utilities, or anything else. Finally FPGAs are getting faster and more dense, probably at a quicker rate than CPUs.  Intel is even moving toward integrating an FPGA on some Xeon CPUs: http://www.extremetech.com/extreme/184828-intel-unveils-new-xeon-chip-with-integrated-fpga-touts-20x...

In an embedded application the FPGA is a lot less expensive and burns far less power than a general-purpose CPU with similar performance. Admittedly on the client the CPU has already been paid for, but on the server the relative FPGA efficiency enables also using dedicated DSP chips and other hardware to assist the FPGA. It provides a stable, consistent platform to support reliable real-time software.

So despite how fast client-side CPUs have become, a server-side FPGA approach for an real-time embedded solution like SDR seems the best (although not only) path forward. I think FRS made the right decision.

Steve - KD8QWT

Justifying a purchase after the fact is a loosing battle, especially in the electronics world.  Buy the best you can afford and enjoy it.  There are many more important things to get emotional over than a radio.

Andrew VK5CV

Cognitive dissonance 101.

Any discretionary purchase like this must have an emotional component the "I want it". I have sometimes found it takes a while to find a car I want enough to pay for over that old one I have.

But when I saw a colleagues' 27inch high resolution monitor I just had to have one too. It looks great with 4 pans.

The Flex products are a better fit for my limited discretionary time. Andrew VK5CV

Walt - KZ1F

Totally agree Bill. Frankly, I think I see some of that on here, "me thinks the lady doth protest too much". I think those that vociferously try to justify what is clearly a subjective opinion are rationalizing... Steve, totally agree with that too.
Certainly most people on here don't care about their underlying OS, the computer allows them to surf, email, social media, etc; it's an appliance.  Up until two months ago, it was a livelihood. I realize my chief complaint only resonates with a handful of people on here. As I said, I can mitigate that to a large extent. Would a person content running Microsoft Windows toss SSDR to run what I release? I am not sure I would but, those that aren't happy with a forced uSoft environment or those that want more functionality likely will. The proof of the pudding....

Walt - KZ1F

Really Greg?? What model do you have. On phone I generally get very glowing reports, unsolicited. I use the Heil HM-Pro, use the equalizer settings Ken gave months ago for the PR-40, as I recall. It works great with HM-PRO. I also found that the default gain setting is too low. I generally leave it set at high 70's and I am maybe 8"-10" away from the mic. Ah, I also use processor and hi processor for dx. Except for Navassa and Malawi I find I do exceptionally well with 100w. However, I do exceptionally better with 600W. ;-). The bulk of the audio reports were barefoot. Right now my rotor is frozen at 278 degrees so working Caribbean and Indian Ocean off the side required some extra umpf.

Barry N1EU

"It is my understanding that the way they are going to achieve this is adding NVIDA CUDA cpu boards and running multiple copies of Powersdr on those boards. Again I am not a hardware engineer but this seems like a hack approach. This is why I keep referring to the Anan as a science project."

Sorry, maybe I'm wrong, but this description of their software implementation sounds like utter nonsense.  I will check with the developer(s) and report back.

Barry N1EU

WA2SQQ

I’ve owned Toyota’s exclusively since 1976 and I’ve decided that I’d never buy another American vehicle. They’ve been reliable and the customer service has always been 110%. To me GM means “Great Mistake”. Yet, if you ask a GM  owner, they would likely have negative things to say about Japanese vehicles. There is no argument that we will all support and defend the brand we’ve invested in. Ask a Flex owner what they think about the Anan and you know what the answer will be. Also, there is no argument that some of Anan’s features (like NR) work better than Flex’s (for now), and the same can be said about some of Flex’s features. No product will ever satisfy every ham.

Each of us did our homework and decided which works best for what we expect from the radio.  For me the choice was crystal clear. Flex is a US company with an extensive background in state of the art communications products. Their design people are members of our hobby, so they know what we want. The end result is a quality amateur product designed and built by hams – in the US. I can pick up the phone and speak with a real person in the US, or I can post a question here and get responses in almost real time, often 7 days a week. Isn't that what it's all about in the end?

Walt - KZ1F

That's an interesting prospective Joe. It might even be true, I am the wrong person to ask on a technical side. However, on a business model side that would not be a deciding factor, beyond using it in marketing collateral. Popularized post WWII, don't make things that last long. There is more money to be made selling a slightly better (better to be defined separately) product to the same people over and over and over again. I forgot who it was that first said this, but I do agree, a ham operator will be very stingy on many things, however they will open their checkbooks wide for what they perceive to be a better radio. I would place good money on when FRS announces the 7000, 40% +/- of the 6000 customer base will be ordering it, if for no other reason than the perception their 6700 is now obsolete. It's called planned obsolescence. You see that in cars, household appliances, stereo gear, TV's, commodity computers, boats, golf clubs and, yes, hobby radios. It's called being in business to make (and continue to make) a profit. Companies without a consistent cash flow in excess of burn rate don't survive. Everything else is perception.

Bill -VA3WTB

The platform for the 6000 radios will last many more years yet, they have long range plans for the 6000's. The hardware is able to do much more then it is right now, it's just getting started.

Joe, KQ1Q

Walt, I agree companies need to sell new products, and from a business standpoint "planned obsolescence" is often used. However the definition of planned obsolescence is an intentional design effort to *artificially* limit a product's useful useful life in order to produce more revenue from new sales. I don't see that as being an intentional design priority on the Flex 6000 series.

If in fact the 6000 series cannot be upgraded as fully as the 100D, causing a shorter product life, that will become obvious in due time. As it stands now, the 6000 series shipped before the 100D, has been on the market longer, yet it appears the 100D might require major hardware upgrades sooner to remain competitive -- despite the theoretical upgradability advantage of a software-oriented "thick client" approach.

I don't remember the numbers, but I think the FPGAs in both Anan and Flex 6000 have a high % of  unused resources, which indicates they both have a lot of server-side upgradability. I think the Anan series are cool radios, and I'm glad there's more than one company doing leading edge work in this area.

Walt - KZ1F

Not brand, model. I am certain Kenwood has a loyal following, not as rabid as here but loyal. So they start with a 530, go to the 830, go to the first solid state successor to the 830, etc etc all the way into the 2000 or 990. Ditto with Yaesu. Actually the 2000 is getting kind of long in the tooth now isn't it? When did the 3000 and 5000 come out? They are discontinued now. I hope you're right Bill, but last I heard FRS has plans out to 2.0. I am sure we'll know a little more in a month. Honest Bill, I believe you are incorrect in your prognostications.

Are you driving the same car you had in 76? Wait, 76? Were you even driving then? hihi

Walt - KZ1F

Well, first off, it's software, I am not sure a 1GB nic isn't faster than a USB2. The other more beneficial aspect is in the chassis the actual radio is Linux. If they went there (PSDR requiring Linux), I would have bought 2.

np2g

Since very few have had both in front of you .There is very little difference except the HPsdr is more mature.

There is a absolute great noise blanker . The latest version is SIFI WOW
There is a absolute great Pure signal function and there is a working Diversity rec far less than the 6700 /

Now I liked the GUI on the flex and when it i s fully functional I will return.
This said it would need a vast improvement on the transmit imd . It would need that 1.5 noise campaign
AND it will have to network better than the Cuda project . Did I say cost less .

To those who say the software is Flex re=run Better look and see what little is left .
I look for CU SDR Another visual event . For me it is about ME making determinations of what I like and How I set it up not what flex thinks I need Sum it up They both are fantastic . Which is better if determined tomorrow who knows

Walt - KZ1F

Joe, that is an 'intentional design priority' in likely every company. That's why things have plastic parts, not metal. That's why car engines dropped timing chains to timing belts. Your inference makes it sound evil, nefarious, robber barons, antitrust, no, it is business survival. It's easy enough to check, when did the 3000 intro and when did the 5000 intro. I am not casting aspersions on FRS or any employee or mgmt. They are a company, not a family member, they aren't doing this for good will or peace in the family, they are doing this to earn a living and build an ongoing business. That involves business plans, product life cycles, non-techie stuff. As soon as I made the decision to order a 6000 (since they were still a year prior to production) I bought a 1500 figuring it was likely better than my TS-530SP. If I still had the KT34XA at 20m yeah, the 1500 might have been better but at 5watts no it wasn't. When it comes time to buy a new radio model, it is all perception and then, given your radiator, skill etc, it is likely the new 'better' model is no better. But it doesn't matter because it is all about bragging rights.  It's mind boggling how many people were totally relieved when Rob Sherwood said, technically, the 6000 (5 and 7) were at the top of the heap based on 3rd order numbers). Unless you have those stacked monobanders at 120', the top 10 on the Sherwood list would probably work equally well.

Barry N1EU

"Sorry, maybe I'm wrong, but this description of their software implementation sounds like utter nonsense.  I will check with the developer(s) and report back."

Nonsense confirmed.  The SDR server platform isn't even Windows.  No PSDR inside.

David Decoons, wo2x

I've got both the Flex and 200D. As others have stated, I feel the two radios serve different demographics.

The Flex 6000 series advantages

• Better construction
• Thin client requirement
• Easier to set up and get on the air (much less configuration settings than PowerSDR mRX)
• Factory aligned. Power out is linear from1 to 100 watts
• Native LAN remote with no third party software
• Full QSK break-in

Anan 200D Advantages

• Open source software for those who want to tinker
• Pure Signal (see comments below)
• Two ADC's allow multiple rx'ers on two antennas simultaneously
• Excellent Noise Blanker and NR2

I've found with Pure Signal that you adjust attenuation for the power out you are currently running. If your power out on that band changes, you need to readjust the attenuation. Also, the Anan 200D requires aligning the transmit power into a dummy load and also calibrating the S meter. Even after calibrating the power out, some bands at 50 watts drive are off by as much as 12 watts! Not as linear as the Flex.

I feel the Anan is geared towards the experimenters while the Flex is for people who want a more "plug n play" radio.

SDR Bridge has added a new dimension of operating with the Flex 6000 series with integration to both CW Skimmer and RTTY Skimmer.

It is a good time for SDR enthusiasts :-)

Dave, wo2x

Bill -VA3WTB

I think Flex is thinking well beyond V2.0 in the brainstorming meetings, and no we don't hear about them. After v2.0 is when it gets very interesting. Flex will need to tempt us with features that is now only concepts. Steve said that himself. They have a lot up their sleev's, wait and see.

Walt - KZ1F

what software implementation nonsense are you referring to Barry? Are you talking about the 6000 or the Anon? The SDR server? You mean the 6000 black box? No, no PSDR, it's linux, the actual radio is driven by linux. It communicates to the outside world via TCP and UDP.

W9OY

I have a 6300 and a 100D.  I like the 6300.  I ran them shoot off style for close to a year and the only thing I like better about the ANAN is diversity.  Both are strong radios but I find the implementation of PSDR to be very outdated in its approach.  The 6xxx is FAR superior in regard to its ancillary suite of software DDUTIL SDR-Bridge FRStack Writelog etc.  The software designers in the 6xxx series are far more responsive to community suggested improvements in the radio.  Pure signal is certainly at present an ANAN advantage but the 6xxx eventually will implement that as well.  The Flex software's feature set grows at a MUCH more rapid clip than the ANAN stuff and in my opinion given the publication of the API's will in fact become the premier hobbyist radio as well.

Much of the ANAN's hype is just hype.  It has a big FPGA but the software guru's are moving the radio's software/firmware guts off the radio and onto a single board computer which will have a barely adequate data pipe between the radio and this new computer.  If that's the case why do you need a big honkin FPGA?  I can easily see this architecture shift rapidly outgrowing that data pipe's capacity.  So where is the disconnect, how did this happen?  Anan designs hardware with a big expensive FPGA and the software guru's decide they don't like coding FPGA code so they "S" canned the architecture.  This is why it's called an "experimenters radio"  

The radio IMHO is at best a promise.  It's a promise on which a good deal has been delivered but not all that much.  Pretty much it's not much different than the F5K I bought half a dozen years ago.  I've seen 14 receivers "promised"    The 6700 has actually delivered on 8 receivers.  It has no internal tuner and it DEFINITELY needs a tuner.  You are totally at the whim of the hobbyist system developers since ANAN just sells hardware.  If your favorite developer flips a clot into his left main coronary artery your radio just froze in time.  The left main is called the widow maker and that clot just made your radio a widower.  This is the reason the 6xxx series costs more,  You get commercial hardware and commercial software and commercial support.  With the ANAN you get a Yahoo group with a bunch of retired military swinging peckerwoods who are full of wives tales about how to get your radio running and maybe one or two who have a clue.  I speak from experience.  (I can say that since I'm retired military.)  There are actually 2 yahoo groups.  There used to be one till politics and ego's got in the way.  Given this state of affairs you need to be very clear the 4500 you pay to ANAN buys a piece of hardware and nothing more.  I understand there is like one guy in the states who can maybe fix the thing if it blows up, otherwise its a trip to India.  

73  W9OY

np2g

Actually you are 90 % incorrect if you wish to support one piece over the other try doing it as ox. Has real feedback real functionality Real comparisons. Not state of your imagination .

Barry N1EU

Walt, you have to click on the "View Previous" to see the nonsense reference.  The software being developed for the Direct Fourier Conversion/**** Pipe OpenHPSDR architecture is brand new and is not the hacked multiple copies of Powersdr that N2LEE suggested.

Barry N1EU

Barry N1EU

Walt, you have to click on the "View Previous" to see the nonsense reference.  The software being developed for the Direct Fourier Conversion/**** Pipe OpenHPSDR architecture is brand new and is not the hacked multiple copies of Powersdr that N2LEE suggested.

Barry N1EU

Barry N1EU

(sorry, I tried to edit my comment and instead ended up with a duplicate post that it won't let me remove)

np2g

Great info Dave. I however do say similar construction. And yes running pure signal does require one to pay attention to details. Defitinely not appliance operator compatibility For all the dynamic capability of any sdr is fantastic

Walt - KZ1F

Oh yes indeed Bill. I am sure they have plans for the next 5 years. I just doubt they include the 6000 series. My guess, the 7000 series will be introduced mid 2016-early 2017 with SSDR 3.0. My point was this. While the 6000 series will still work the same as the day before the 7000 intro, 40% of the 6000 series owners will order the 7000 because they will perceive their 6000 as "yesterday's radio" vs. those that are happy with it and won't shell out another $10,000 for it's replacement. The issue is some people's buying decisions based upon perception and bragging rights. And, yes, I am sure the 7000 will be better based on technical in-the-lab merits, not necessarily how it compares to the 6000 at your site.

Steve W6SDM

I looked at the ANAN but just couldn't see myself buying a radio that was blue. 

Steve W6SDM

It's important not to get caught up in branding when buying anything.  That's what the marketing folks would like us to do, but we need to evaluate purchases based upon the merits of the product and lot the label on the box. 

On the car analogy, I went from a BMW to a Jeep Wrangler.  The reason:  My BMW didn't like dirt roads.

I think the synergy between the hardware and the software on the Flex makes it the most desirable.  We have the benefit of some really smart people making sure the whole thing works together.

Walt - KZ1F

Don't like IBM?

Lee - N2LEE

Steve in my opinion it's just, if not more important to buy a Company more than just features. Feature check boxes come and go with ever product release but if the company that is developing the product is going left and you are going right then in the long run it doesn't matter if they have the right number of check boxes.