On Sun, Jan 10, 2010 at 4:34 AM, lazysenior <[email protected]> wrote:
> Somehow I just aint't getting it? > The Facts as I know them: > > 1. The spur Problem was brought to Flex's attention a year ago. > > 2. In Oct 09 Gerald puts out a statement that in my opinion trys to make > light of the problem but promising a fix. > > 3. After ordering a Flex 5000 I am warned by 2 respected 160 mtr Gods (W4ZV > and W8JI - check out their credentials on the net) that the Flex will cause > me grief is I use it on 160 mtr CW with an amplifier. > > 4. Jan 2010, Problem still exists. > > Somehow, I can not help but think "something" is very wrong if the > President of the ARRL stops using a Flex 5000 because of spurs on MANY bands > including 160, 80, and 30. And remember you can ONLY run 200 watts on 30 > mtrs. He isn't the only well known ham who is not using a Flex any more. > Several other contesters have put the Flex 5000 away. > We all love a conspiracy. Actually, I am glad this came up. It prompted me to actually test my Flex 5000. I learned something and I even got to give myself a dope-slap up-side the head. I have been using the classical superhet radio for so long that I keep forgetting that the Flex is not a superhet, especially on transmit where it makes a much bigger difference. When I was trying to understand the results of my testing yesterday using RX2 to monitor the spectrum of the F5K's transmitter (what other radio can do that!) I kept scratching my head as to why the spurs were so prominent outside the transmitter's passband. Why the filter should have removed them! Then it hit me this morning: NO IT SHOULD NOT! Why? Because there IS NO FILTER after the D:A conversion. This is an I/Q transceiver, basically two direct conversion radios that use cancellation rather than filters to eliminate images and artifacts! Any artifacts left in the I/Q digital stream will appear at the transmitter's output. What this means is that, while supremely FLEXible (sorry, I couldn't resist), the Flex transmitter is unlikely to have the out-of-passband spectral purity of a classical Collins transceiver. But it also isn't limited to what can be accomplished with that fixed-passband crystal or mechanical filter either. (BTW, the K3 is a superb modern execution of the KWM2, perhaps the best ever. It impresses the hell out of me.) So where does that leave us? It leaves us comparing apples to oranges. If your goal is to run high-powered CW DX on 160m, I am willing to bet that the K3 may be the ultimate evolution for that type of operation. OTOH, if your goal is to have the radio that will do more interesting and different things, especially in the future, the Flex 5000 is currently the only game in town. (But I think that, even today, the Flex 5000's receiver is superior to the K3 since it achieves its performance without having to rely on crystal filters. But I digress.) I have been thinking about this for a bit and now I want to lay out a scenario that is probably going to scare the s... heck out of you contesters, a scenario that the SDR radios from Flex and those that follow in their footsteps are going to make possible, a future that is unreachable using the traditional Collins KWM2-like transceivers that most hams have today. The real hint is the CW skimmer and, to a lesser extent, the PSK monitor functions. Virtually everything we do in ham radio has been based on the, "one radio, one signal," concept. That is going to change. CW skimmer makes it possible to decode all the CW signals in a passband thus giving the CW contest op a huge advantage. But that op is still limited by his own ability to respond to one signal at a time. Also, this capability is limited to only one transmission on one frequency at a time. But here is the endpoint of all this, the thing that is going to change (and probably kill) ham radio contesting. First off, it requires the logical endpoint of CW skimmer. Why stop with CW and PSK? Why not just decode every single digital mode on all the bands at the same time? The only thing preventing us from doing that is processing power. No readily-available computer has the power to do that. But many readily-available computers do. Let's take our Flex 5000 and forward the I/Q stream over a local digital bus (gigabit ethernet would be my choice right now) to a bank of cooperating machines. Each machine uses its power to decode as many signals as it can. With enough machines you can decode everything. We can even make it smart enough to recognize the various modes and autoselect/autodecode. No operator intervention required. Now you have decoded all the signals that you can hear. Next thing is to pipe all the decoded streams to the QSO processing machine. This machine now decodes the QSO and generates the proper response, i.e. the standard exchange. Given how simple most exchanges are, this doesn't require a lot of processing power. Pattern match the call, do the dupe checking, and if you need that station, generate a standard exchange. Since we have lots of computing power, we generate exchanges for ALL the stations we can hear. We now generate a new I/Q stream containing all the transmissions (in real time) so we literally carrying on multiple QSOs concurrently. Of course this requires isolation between the transmitter and receiver(s) so we just data-link all this across town to our transmitter site that is far enough away so that, even at full power, its signal is weak enough at the receiver so as not to cause compression/blocking. The end result? One guy, with one station, at the flick of one switch, now works every station automatically, some (many?) of them simultaneously. With the ability to scan and monitor other bands the software can determine changing propagation conditions and switch bands, continuing the process. I know, this sounds like science fiction but it is possible to do this today with the Flex radios. No, the existing PowerSDR will not allow this but I suspect that the next gen software, if it is sufficiently modular, will permit someone to build just such a fully-automatic contest station. All it will take is a bit of time, money, and willingness to get the job done. Me? I couldn't care less about contesting but such an application intrigues the heck out of me. I would love to have that for just one contest and to blow everyone else away. The only thing is, once you build it, there is no more challenge. "Ho hum, another contest to win <click>." OTOH, I like the idea of a radio that would automatically determine propagation and allow me to connect to and exchange information with any other station I can possible hear. It would autoselect the best CODEC for the propagation conditions. It would transfer any digitized information I want, including voice, text, and images. It would tell me what is available in real time and what has to be deferred until conditions improve to support real-time communications. Imagine the software telling you, "If you wish to establish voice communications, increase power to X, otherwise text at Y characters/sec is available." So, I am going to stick with Flex. Even though the Flex 5000 may not be as good as the K3 doing 160m CW today (at least on TX, the F5K is probably superior to the K3 on RX but I won't debate that because the K3 is very good), the future possibilities with the Flex so completely outstrip the K3 (or Yaesu or Icom or Kenwood or Ten Tec or ...) that it isn't even a horse race. (That phrase has me thinking about the analogy of all the horse riders laughing at the first automobiles. Yup, the horse was far superior to Ford's first tiller-steered, gasoline-powered buggy and remained so for several years. But the automobile and horse, like the steam drill and John Henry before them, changed places and we moved on.) -- 73 de Brian, WB6RQN/J79BPL _______________________________________________ Flexedge mailing list [email protected] http://mail.flex-radio.biz/mailman/listinfo/flexedge_flex-radio.biz This is the FlexRadio Systems e-mail Reflector called FlexEdge. It is used for posting topics related to SDR software development and experimentalist who are using alpha and beta versions of the software.
