I was overdue for a blog entry and thought as I am laid up after some surgery that I could share some thoughts about the new Flex 6300 and the state of the SDR market right now. These thoughts were triggered with some chatter on eHam's SDR Forum.
The bias for and against Flex Radio Systems seems a little more tempered these days. I know in my own mind from conversation with N8MSA (Mike Alexander) I learned a lot of things I really didn't know about the 6K signature. They are real reflections that only a seasoned SDR owner can know and Mike has the critical foundation of owning the next best thing to a Flex where he can compare in real time. An Apache-labs Anan by the way if your wondering.
This post won't delve into that too much, however, the gist is that the 6K series moved the computer back into the box, so you have Linux running inside. While their ADCs and FPGA are ground breaking technologies as is the Flex 6K series is holistically, we are in all reality still tied to a commercial/consumer Operating System under the hood.
Being as objective as possible, really the HPSDR team could easily do the same if they wanted. There are lots of viable Linux distributions and many of the HPSDR developers have been working on Linux versions of the software.
The problem now though from a competitive prospective is the the commercial product (Flex 6300) offers a highly compelling alternative! Objectively there would seemingly be no compelling reason now to buy a straight Anan 100 over a Flex 6300. We need to see if and when there are ever valid performance comparisons if that become a driver or not. My view is merely an opinion right now and I have no bias really for one or the other apart from unoffical conversations I have had offline about Flex and Anan performance comparisons. Just look at the two, compare price, compare specs (you can get more RX slice receivers in CUSDR or Studio 1with the Anan - no TX though) or the same number in SmartSDR vs PowerSDR mRX.
The fact is that if the HPSDR team wants to offer a high-end alternative at a competitive cost they need to finish CUSDR so they can realize the increased number of slice receivers. We likely will get news of where all that stands at the German Ham Fest (International Exhibition for Radio Amateurs) in June. See Calendar at SDRZones wall for dates.
Apache-labs might consider embedding a computer in future Anan's. They could still move fast if they want to catch-up and even leapfrog Flex. Rumors might have this happening in 2015.
So what now? Why am I bringing all this up?
The reasons is that for a would be SDR purchaser you really want to survey the landscape well before making a purchase. Here is my view of what you might consider.
Technology
Who is leading and advancing faster?
Apache-labs has slowed in my opinion. The Anan 200 only really offered remote jumper configuration and an RFTap for Puresignal. Flex has those built in already as well, so soon you'll see their PureSignal equivalent. On the flip side, Apache-labs came quickly out of nowhere, the developers behind hermes have been nothing short of inventive and there is more to what you can find out if you talk to and or develop relationships with the right people.
HPSDR has not delivered Gigbit Ethernet yet, so you can get 5-7 of the potential 14 Receivers on the dual ADC models.
You won't get a full 100 watts in all the operating modes on a Anan without risking burning up your Anan. On the flip side, if you have an good amp as I do, you dont care because I for one never need more than 30 watts from my Anan anyways.
The roadmap right now for multiple devices and platforms is greatly impacted by the Ethernet protocol in use. I have heard there may be changes that will solve some of the limitations there.
The software development for a brand new SDR control package looks limited with no real advertised server functions. Maybe GNU Radio, if and when it gets there.
Flex has commercial support (at a cost for software updates) and a dedicated team that really is picking up speed now and is clearly innovating.
They are not hampered by their network architecture.
Flex still needs to build a secure Internet protocol and Apache wont have anything until and or if Networking is changed and a server engine is developed.
Bottom line though is that the Anan's outperform their Flex counter parts at typically lower cost.
The Companies
If you want to get the best product and support you need to look at what have they done lately and how likely are they to do more for me during ones ownership of a radio.
Flex is rapidly developing a strong network of third party developers. I predict their 6300 offering will increase their installed base over 10 fold in the next two years further growing their installed user base and funding pool for continued software development.
Apache-labs has no real software support, or even software for that matter. It's a big problem for inexperienced new SDR users. Just go look at the forums and many of the hams that do offer help love to treat you like an idiot or force their egos into the help you receive. The leading source of Documentation was pulled from the yahoo site due to infighting. Fortunately you can still get some of it on the Apache-labs site.
While I don't like that I have to pay for support from Flex, the upshot is that that I will be owed support.
History of the two companies is wildly controversial. Apache labs has limited history and depending on whether your a Flex Hater or Lover you have varied feelings out there. The fact is while Flex had some delays getting the new 6K series out the door, they look to be living up to their commitments now and rapidly making good on promises. I have to confess historically I was skeptical about Flex surviving. The look like they are going to be strong now. They sold out of 6300's before Dayton even started.
Bottom line here is that you get a full service product offering with Flex and a hardware support experience with an Apache-labs offering, with Open Source support for the software. Support is out there if you ask in the Anan world, but not owed to you and it might come with bumps and bruises.
Apache-labs looks like its ascent to higher levels is steady and we will see Flex still moving fast for a while before it tapers off once Smart SDR is rounded off and they address some bugs at the OS level.
Personally, I am still on the fence about whether to get a 6K and which model I would get if I did. I will wait to see if we get fully functional cuSDR release with TX this summer and decide then. The truth is that my Anan 100D is still more promising than the prominent cost of a Flex 6700. I cant really see myself gaining anything in my hamming expereince for the extra dollars I would be laying out.
My Anan 100D experience right now is rock solid. So despite all the cons I have pointed out in the spirit of being objective, I have one hell of a radio that I ended up getting at one hell of a price, especially if cuSDR or Studio 1 software gets transmit capabilities. If they don't then I still get a really nice experience with PowerSDR mRX thanks to the HPSDR team. They do a great job giving us cool ground break technology with patience being ones greatest cost. Apache-labs needs,to be careful in my opinion that they don't get so far off on the hardware end without software to back it up.
All said and done, I never really had a beef with my Flex 5000 either or my ownership experience. I got into it later and Flex had worked out the kinks. I really havent had any problems with my Anan 100D either.
This brings me to my last points, its beneficial to wait and see sometimes. I got lucky with buying the 100D.
You need to see what each SDR company has done lately and where that leaves you for the period of ownership your planning. Consider your own ownership style, it might not be different than what you expect from an auto ownership expereince. Economy with performance, or luxery with full service.
My advice is think about a 4 year ownership period if you buy a an SDR right now. Take its full costs, all of them combined, and then decide it by either 36 or 48 to get your cost per month of ownership subtracting out its realistic value it will be sold or traded. I'm would use a resale price of 20-25 percent to be safe and realistic. So if the radio costs $5800 all in by the time you buy everything, pay support costs then assume it will sell $1000. That makes $4800 you'll spend and if we divide that by 4 years you get a $100 per month of ownership cost.
On my Anan 100D, I spent about $3400 all in and I am a year end and could likely get at least $2000 for it. I think in 2 more years it will still be worth $1000 so my ownership costs will be pretty good with a 3 year target of $2400 spent or under $70 a month.
Flex 6700 for 3 years would be $180 month
Flex 6500 or Anan 200 for 3 years would be about $90 a month
Flex 6300 or Anan 100 for 3 years would be about $55 a month
These are all approximations and attempt to consider some of the software maintenance costs.
I hope this helps some you looking to purchase SDRs. There are of course numerous options, most low power though, so you may need to factor in Amps ect depending on how you want to operate as 5 Watts probably won't be satisfactory more most, especially if your going to work pileups.
Why I ordered an ANAN-100D - Part 4
In this blog, the (perhaps) final post on my decision to order an ANAN-100D, we will look compare what we know about the two products (FlexRadio 6700 and Apache Labs ANAN-100D) and examine my decision to order the 100D.
First, a quick recap of the two radios: both are direct-down-conversion (DDC) software defined radios (SDRs). Both use two (possibly more - read on) high-performance analog-to-digital converters (ADC), feeding a data stream to a field-programmable gate array (FPGA), a fast and flexible logic device that is used, in this application, to pre-process the data from the ADCs to better manage the burden on whatever is executing the decoding and encoding (demodulation and modulation) algorithms. And this final item is where they seem to deviate the most...
I mentioned in the previous post that the Apache Labs philosophy is to use the OpenSDR Hermes architecture, which balances the processing load between the radio and the PC. The developers have, in recent Yahoo group posts, asserted the opinion that this strategy yields the best cost-to-performance ratio. Flex, on the other hand, is offering a "ham radio server". Their position is that there is a market for a high-performance radio that allows access from anywhere via a TCP/IP network though low-resource software, including hand-held device apps. Which approach is correct?
The answer is, or course, that both approaches are correct. Which one to buy depends on the needs of the customer. Let's look at some key capabilities:
Raw performance (selectivity, sensitivity, etc.): The Apache Labs ANAN-10 and Hermes boards are in the hands of users, and the initial reports are that the raw receiver performance is excellent. There are differences in strategy between Apache and Flex, but I would hazard a guess that both will perform very well in this arena.
Transmitter performance: Flex is renown for the sound of transmitted audio, and I think they will want to maintain that reputation. The OpenHPSDR community, on the other hand, is working on development of an adaptive distortion algorithm to "clean up" the transmitted signal. My guess is that the Flex will sound a little better at first. Both are rated at 100W out.
CW: Flex seems to be spending a great deal of effort in ensuring that CW transmission performance will be similar to, and perhaps better than, more conventional transceivers by moving CW logic into the FPGA. This addresses a major issue in the previous generation of Flex radios: marginal CW TX performance. The Hermes-based ANAN-100D, on the other hand, doesn't seem to have a CW "engine"in the FPGA and is thus expected to perform much like the Flex 5000A.
Digital modes: The Hermes architecture depends on the PC software to provide a digital audio stream to most 3rd-party software. It may also output I/Q streams, but few programs of any kind, aside from SDR software, can make use of it. The capabilities of the Flex offering are almost completely unknown at this time, but they have made statements that they will have various output capabilities.
Remote control: The OpenHPSDR is, at the time of this writing, attempting to define what a remote control protocol would look like for the Hermes platform. The solution would, in all likelyhood, be closely linked to the software. The capabilities of the Flex offering are, as with digital modes, almost completely unknown at this time, but they have made statements that they will have various remote control capabilities.
Accessories: The Flex 6700 will have an antenna tuner; the ANAN-100D will need an external tuner.
Input/Output: The primary connection to PCs for both radios will be Ethernet. Aside from that, both have multiple antenna connections, audio connections power connections. The Flex appears to have a balanced audio input connector.
Software: The Flex will initially ship with SmartSDR, which is mostly a "thin-client" human-machine interface. It will present a visual representation of the radio spectrum, and will allow the user to exert some manner of control over the radio. The Hermes-based ANAN-100D is supported by a number of software packages, including open-source variants of PowerSDR and new applications such as the previously mentioned cuSDR.
Neat tricks: The Flex has a VHF receive capability that the ANAN-100D lacks, while the 100D has the ability to accept a 3rd ADC.
That is a quick "thumbnail sketch" of the differences between the two radios. There are some additional details on their respective websites, and encourage you to examine the currently available information thoroughly.
So...how did I decide which radio to buy? The first question I asked myself is - do I need either one? Besides the joy of tinkering with a new radio, what would I be getting and do I really need it? The answer to the first question is that I would need some system to allow me to escape the limitations of what was, until recently, available. But do either of these give me what I am looking for?
I think that answer to that question is that the Apache Labs ANAN-100D gives me a little more of what I need, and has fewer unanswered questions about what the future holds for it. The ANAN-100D:
Has a large development community through the open-source OpenHPSDR organization
Has more software, with a significant range of capabilities, available today than the Flex
Has great "tinkering" potential in that it is an open platform
Strikes a comfortable balance between "all in the box" and "split between the box and the PC"
These factors, weighed together with not knowing many of the details needed to make a high-confidence decision about the Flex, led me to order an ANAN-100D.
Bottom line: I think that the Hermes-based Apache Labs ANAN-100D will do more of the things that are important to me, and many other hams, right out-of-the-box and has great development potential.
And there you have it! Will the ANAN-100D live up to my expectations? Please join me in, as I intend on sharing my experiences with you, starting with unpacking it and installing software and ending up...who knows where...
Until then - 73 de Mike - N8MSA