The AM Forum

Specs and technical details look impressive.

(http://cart.flexradio.com/thumbnail.asp?file=assets/images/radioimages/flex-6700%20angle%20view_thumbnail.jpg&maxx=200&maxy=100)

Flex 6500 Transceiver - intro pricing stated as   $3999
Flex 6700 Transceiver -    "       "          "      "   $6999
Flex 6700R Receiver   -    "       "          "      "   $5999

Target date 4QT 2012

Additional details here: http://cart.flexradio.com/Signature-Series-Radios_c_17.html
Product brochure is here: http://www.flex-radio.com/FLEX-6000.pdf

Oh, y'all who want to work ham radio by operating a computer must be about to die and go to heaven  ;D

I'm thinking about all the wonderful boat anchors I could get for $7000.   I could put up a damn fine antenna for that money.

Yet another "Last radio you'll ever need!" ::)

Tragically, unlike the German transceiver we have seen, this one does not come in a variety of colours.

Can't get it past the front door, therefore.

Whew!  what with all the negative replies, do I detect fear of the future?  I wonder what y'all're typing on to send these messages around?

I wonder how many computers you've replaced since you religiously placed your treasured Commodore 64k on the display shelf, never to be thrown away and spoken of in hushed tones to the annointed only?

Well DDS is here to stay, whether the display is separate with keyboard or buried deep within the rig.  It will only get more bandwidth, resolution and programming features.

The only curve not followed yet is price, most likely because the ham equipment market is so small compared to mass consumer electronic market.

Those of us who hoard ancient parts, are still able to build, repair and find such for repairable rigs are fastly becoming even more of a minority.

Soon, in quasi-geologic terms, our cherished anchors will be found only on museum shelves, admired for quaintness and useless in a digital world.

How many analog TV stations are still transmitting?  How many of you are still glued to your CRT TV's?  Have you tried the internet on many TV's and virtually all DVD players now sold?

Soon we'll all be speaking in "Dige",that is if we can get by Dec 12.  ;D

Very interesting. They copied HPSDR and raised the bar with faster A/D and FPGAs.
But then I don't have anywhere that kind of money invested.
The only downside I see is clock phase noise. I've worked on SAW oscillators in the 1GHz range. They are about as quiet as it gets.
When you take a low frequency clock and PLL it up in an FPGA the close in phase noise suffers. Notice there is no phase noise spec inside
10kHz. DDC needs very low clock jitter to work properly so we will see when they drag it out to the quarter mile and bleach the slicks.
So we will see but it clearly raises the bar if they can make it take advantage of the faster clock.
Flex 5000s should be real cheap soon

There's always the The DZ Sienna HF rig, it is available with different face plate
color schemes  ;D

These new SDR rigs should be dirt cheap in another several years as the latest
and greatest models emerge.

Pete

I feel a serious SMUG ATTACK coming on!!   ;D    ;D

It may make a nice retirement present Al!!  With Yaesu and Kenwood selling rigs for greater than 10K the Flex seems like a real bargain plus they sound excellent on AM. 

Joe, W3GMS

Um, the website says, "Your On-Line Guide to AM and Vintage Radio"  so what did you expect? 

Someone has to convince me there's something vintage about the Flex Radio.  Good luck.

Why does A.M. HAFTA be Vintage, Rob?

It doesn't have to be vintage Budley but up at the top it says

...AM and Vintage Radio.

not

AM or Vintage Radio.

I extend your point logically to also mean that

if it is one OR the other, than any vintage radio topic is okay

vintage slopbucket...vintage CW....and maybe that is true.

My point was not about approval of topics since that's not my job but rather that a little derision is to be expected in an on-line location frequented by vintage radio enthusiasts.

I'm into my dual RX HPSDR for under $1500
16 bits 122 MHz sample

The new Hermes single RX XCVR board will run about $900 and all you need is a final.
QSR1 with the exciter plug in will run you about $1100.
I want to run CW that sounds like W2ZM's spark rig.

Notice the $199 charge for software upgrades and support for 1 year. Not for me
Flex must think they are zuckerburg.

Yay, it has rack handles!!!! I'm gonna buy one for my wife for her next birthday! I'll let you know how that plan works out!

Rob W1AEX

Interesting Frank. Another shoe drops, a yearly rental fee on a closed system.
You were right in your notes about the uocoming Flex before announcement.

As you know, but for others, For the record I don't have a Flex but do have a QS1R and enjoy it very much, particularly the ability to see activity across whole bands, what bands are open when viewing the entire 60 MHz sweep, who's loud, who's ragged, who's swishing their vfo's, etc. Lots of fun.

As to vintage, other Frank, whole topics are devoted to class d and e on this board.  Welcome to the 21st century.

It's also fun to fire up my 32v,  hb 813's, and confab with the "greats" of AM.

There are many members here that no longer use vintage anything to operate AM. Discussions about "vintage" SSB, CW, technical(as it relates to our hobby), etc. are not uncommon here.

From: READ BEFORE POSTING ON THIS SITE
This is a radio site dedicated to AM Radio. The reason most come here is to meet and exchange information about our mode and hobby.
In an effort to provide our community a meeting place on the internet for AMers we had conceived the QSO section for general topics.

If the topic doesn't interest you, don't read it.

Entirely new software (SmartSDR) and not open-source. PowerSDR for the Flex 1500, 3000, and 5000 will remain without any yearly fee.

Actually the color of the box is unimportant. There's nothing to look at or diddle with. The viewing color/style is on the screen. See some of Rob's PowerSDR screen skins: http://www.w1aex.com/psdr/psdr.html

I wonder what ADC they are using...
All it says is the sample rate is 245mhz

Rats, just missed 2 meters natural. 
Well, we'll just have to over sample.

...only a matter of time.

Pete, those skins were pretty nice.

Heck I would love to build a big spark rig but as you say Rick, the spectrum display is great.
This stuff is magic to me. Just like a 6V6 and a crystal back in '63.
It's all cool

I just looked up the A/D used in the new Flex.
Check out effective number of bits. (Hint 12 bits above 5MHz input)
Now look up the 16 bit converter used in QSR1 and HPSDR yea only clocking 130 MHz max.

Does the Flex 6000 have the same audio bandwidth ?(Transmit Bandwidth Default 300-2700 Hz (Variable 50-5000 Hz) for both 6500 & 6700

Yes, I finally found it in the documentation...

The 6700 using AD9467 at 16-bit 245.76 Msps

Read right over it twice, and somehow missed it.. :-)

It's in the product brochure: Transmit audio bandwidth is continuously adjustable up to 10 kHz for AM and Enhanced Single Sideband Modes.

Yep. That did it for me right there Frank. They produce and market a rig that relies on softwa.......wait, lemmee re-phrase that. They sell a $7000 rig that is worthless without software support and system upgrades, and they have the unmitigated nerve to charge $200 a year for providing support for their own product??? Good luck with that Flex.

I don't care if it has ears like the Arecibo telescope. Maybe they should rename the company to Flex radio Service.

  

Remember. With Facebook you are the product not the customer. They are selling you.

I like paying for products (not too much) instead of being the product. I also don't mind paying for software upgrades. Like everyone else, I draw the line at paying for bug fixes. That's just wrong.

There are many hardware and software companies that charge for support either through service/maintenance contracts or "what to talk to a real person", pay up first scenario. Service support, system upgrades, feature enhancements, or new features should have a monetary value associated with them. They all require time and effort and generally are not associated with zero cost to the manufacturer. However, as a consumer, no one is twisting the customer's arm to buy into this additional cost. At $200 a year, that breaks down to $.55 a day. If you just spent $7000 for rig, plus the investment of $1K to $2K for a super wizzy computer and monitor, I don't think $.55 a day is going to roll back a lot of eyes. A $7000 product investment that has the feature of continual improvement, features, etc. for $200 a year seems like a good investment. Many of the current amateur rigs depreciate rapidly when purchased and never have any option for enhancements or new features. Want new enhancements or features, go buy or wait for a newer rig to hit the market.

That's pretty horrible. An over priced radio to begin with.
The price clearly shows that all the R&D and software development costs
will already be paid for with the "introductory price".
( which will likely increase by the end of next year).


Software updates used to be a selling point for SDR.

Now they are just another sales item.


10 years with the radio, and assuming there are updates available each year, that's an
extra $2000 .

At least they don't make you pay for the software the first year.   ::)



Hopefully, they will print an explanation describing what "upgrades" ( bug fixes etc...) come at no extra cost.
That would make some folks a little more comfortable signing onto an ongoing expense.

I have mixed emotions about this, vis. N8VB's much more open system with user input into,( gasp,) the actual programming, user bug catches and fixes, ideas carried out for improvement. 

On the other hand, Apple didn't become valuable and for that matter, viable, by one man operations, soldering own boards, getting a critical tool (20 power microscope) from wife as a present and possibly getting bogged down from dividing time and abilities between hardware and software.

So, yes, $200/yr. support seems to be the trend.  Closed systems like IBM's sooner or later will be reverse engineered.  Just as Smex (Smugex?) has "pirated" or "added value to" HPDSR and Phil's  DDC, so did Apple with Zerox Parc's mouse and GUI. 

My bigger problem is what kind of outfit takes orders a half year in advance and expects you to pay for the privilege of waiting with baited breath for the news letters about progress?  Further  what kind of people (yeah I was one once, OTI Quantum 6 in. Maksutov) puts down the money and then waits long after the promised delivery date and can't wait to shower the 'benefactor' with kudos even before the product is delivered.

How will that company now take care of it's stranded sound card, instantly obsolete, user base?  The two systems are imcompatable for upgrades. 

In Apple's defense, it tried, somewhat unsuccesfully, to maintain backwards compatability with The "II forever" 6502 design, what with a 16 bit IIGS, etc.  -But finally had to radically change over to the Mac (Jobs), yet still in our era maintains compatibility with microsoft/OS op. systems.

 A lot of decisions make or break(brake) a company.  Many 'fancy' ham companies leave the market after dipping into the aero-space/military market.  Quantum/OTI was one in optics.  I now have a very nice orphan with one moving part, thank God.

"Wrex" may have finally misread the amateur community.  A community not cheap, just frugal and savy  enough about the business to know the design, know the cost of parts in single and bulk lots, know enough about business to estimate overhead, design talent cost, hype and character.  What closed systems miss is "family."

This will be interesting.  I'm really impressed with the 6000 specs. About time a faster than 125 MhZ DDC came along. - Interesting ancillary chips going with it too.  Don't think thought that I'll "get in on the ground floor."   As previously mentioned others will step in with the same or better in the not too distant future.

On a bright note; "GO, USA"

Here you go,
Build your own.  Analog devices engineer puts it all together for you.  Note the discussion on clock jitter for increasing analog input of the AD9467.

well, a good start...   ;D

http://www.analog.com/en/analog-to-digital-converters/ad-converters/ad9467/products/product.html (http://www.analog.com/en/analog-to-digital-converters/ad-converters/ad9467/products/product.html)

It should be very interesting to see how it all ends up. The funniest thing I read was the whining from a predictable group of Flex owners who were upset that they were being left behind with the addition of the new 6000 series. As if everything should be frozen in time so their investment wouldn't become "obsolete". Many of those folks are the same ones clamoring for better performance and improvements. Very shortly I expect to hear the arguments that the 1500/3000/5000 platforms are better than the 6000 series because (insert stupid reasons here that sound like SDR-1000 owners explaining why the 1500/3000/5000 platforms suck).

There is an obvious subset of people obsessed with juvenile "one-ups-man-ship" within our ham radio community who have to make it known that their rig (that they bought) is the best of the best. Seems like many people lose perspective and forget that it's only a hobby. I'm no Flex fanboy but I do like the rig I purchased from them. I don't mind spending a few bucks for something that I enjoy, and I also don't mind building or restoring something that I enjoy, but the key word is that since it's a hobby I want to enjoy myself.

At any rate, the reality is that for most hams a vastly improved receive capability (if this proves to be true) simply means that they will hear their neighbor's appliances and the utility company's grid faults even louder. I applaud Flex for adding new stuff, and as far as the path they've chosen for service/software/support is concerned, it will either succeed or fail based upon how receptive their client base is. That's the way it should be.

Rob W1AEX

Oh yes...

I am indeed happy that Flex finally took the step into RF sampling devices ( ADC/DAC) rather than using audio devices.
That WILL change the appliance operator's experience in ways they won't even realize.

Do I think Flex should make the radio available at cost? or with very small profit?  NO, that would be silly.

Do I think taking deposits ( refundable deposits only..) and waiting until delivery? no, that's ok too.
Though I wouldn't think that Flex NEEDS to do that... I bet they have plenty of revenue from other sales
to float manufacturing costs.



New products are available all the time, those that feel left out because suddenly there's something better should just calm down and think a little!   ;D

All this P&Ming about closed software and upgrade costs mostly by people using the Window OS. Pot to kettle and duh!

Is that sort of like the pot calling the kettle black? 8)

..and are you using a water clock driven, peg and wheel computer, rachet driven stylus?

We have far more in common than p&m differences.

No. I'm using a closed software OS and I pay for updates. BUT, I'm not P&M about the Flex equivalent. Others are and that's my point.   ;)

Bottom line: If you don't like the Flex 6000, don't buy it. This mine is better than yours BS gets old.

They have a LiveStream web cam at the Hamvention FlexRadio booth but the position of the camera doesn't really show the 6000 series.
The LiveStream video URL is here: http://www.livestream.com/flexradiosystems

Marketing scam the A/D they are using is only 12 effective bits above 5MHz
But they will pump the sampling rate to the fools.
It will be interesting to see some real specs because I only see fluff.
glad i don't have to pay $.55 a day for my Drake TR7A or V2 CDC.
Get the sucker hams to pay for beta testing and do the testing for free then market to the mil guys for big bucks

We all know that the amateur radio world is full of conspiracies, marketing hype, products that resemble a gaggle of boards hooked together with an octopus back-plane, bad vibes, imaginary products, and back-room gangsters. $.55 a day for a Drake TR7A or V2 CDC is silly; they're both dead products.

I have an SDR 1000. It's way better than the later models because it took lots of fiddling and fussing to get it to work and that is the essence of fun.

I am crazy for the spectrum display but alas I use the Flex very rarely. Boat anchors are even more temperamental. In fact sometimes they don't work at all when yesterday they did. Fabulous.

At first glance from gagle land I thought the radio quite cool until I looked closer.
Then I just stated fact.
Heck I bet the 5 year old 14 bit Perseus is a better radio.
Sorry it burst the bubble and set off the BS detector

Hmmmm, Frank.  Velly Intellesting.  We need to look at the rest of the chips too. I'll have to look closer at the modified AD9467-250EBZ circuit eval. board along with the companion HSC-ADC-EVAlCZ FPGA data capture board shown in the AD engineer's video to see if such evil lurks there too.   Boy did you get a load of his wall of test apparatus? 

Oh just plug into your Rhody here and grab a handful of filters there...
Just cool down your helium dewar there, insert your TXCO there...
Calib. vs. GPS here and Vector two stars 'till morning.

Flex should've hired that guy.

OK, found it.  Page 4 of AD9467 specs PDF file. "AC specifications"
5Mhz and up 12.1 to 12.4 effective bits, +/- a tenth of a bit or so.

Maybe thats why the FFT spikes are so flat topped and cartoonish in their demo GUI spectra..   ;D   (well, not really, but.... )

Decimation and log calculations aren't in my vast list of strengths, but there was a very interesting question regarding the 6000 series 16 bit A/D architecture, sampling rate, and the 25 bit DDS in the Flex reflector. I won't pretend to understand what the Flex programmer said (Steve N5AC) but you two guys can probably make sense of it:

http://www.mail-archive.com/flexradio%40flex-radio.biz/msg47919.html

Get your calculators out and enjoy!

73,

Rob W1AEX

Mine is better. All others are junk. Anyone who has one of "those radios" is a fool.

How sad.

One of the great minds told me Flex tried to hire him. The pay was the honor of being part of the team.
Rick, glad you understand the spec. I don't want anybody to think I'm making this stuff up. The first thinng I noticed was it wasn't an ITAR part like the 2208. Now check out the offset slop. I think it was 150 counts.
Some smart guy please explain how extra oversample makes up for 4 bits of resolution???

That line of thought always reminds me of Tom K1JJ's reference to the much beloved "TS-XXX Pro Field Disco Duck" rig that will always be at the top of the heap.

Incidentally, many eyebrows were raised at the brochure's TX bandwidth limitation of 10 kc for double sideband modes. In the FlexEdge forum fellow AM'er Chris Sanger (KB3KRJ) had the following dialog with Tim Ellison from Flex that indicates that Flex has decided to lift this restriction:

Chris:  The specs call for up to 10k wide transmit bandwidth ... on AM will this be 10k on either side of carrier?
Tim:  5K max one sideband.  10K max for double side band modes.
Tim - 24 hours later:  We have decided to expand this beyond the 5kHz listed on the brochure. It will probably meet or exceed the FLEX-5000 bandwidth.

The last statement indicates an equivalence to the 5000A which means at least plus and minus 20 kc to allow a maximum of 40 kc for double sideband modes.

Rob W1AEX

When you claim to be the best, you should be

Wow, there is a lot of ASS_U_ME in that flex article. 25.5 bits, yea right. assume no clock jitter assume it is really a 16 bit converter. we will see when a guy like Sherwood puts it on the bench.  FYI the gagle machine and QSR1 gain a few dB over the 96dB due to 16 real bits. Notice flex spouts in band IMD only, another worthless spec. I bet the R390 can beat that number for in band since it has an AVC.
Their processor and back end look good but the A/D is very lacking. Also the hi pass and low pass filters look good. But why use such a crappy A/D the heart of the system.

The LTC 2208 comes in two, right up front, flavors or packages.
The orig. LTC 2208 is 16 honest bits and the LTC2208-14, is the 2208 lite flavor.

So how does oversampling work for 4 missing bits?   Time.
-The added dimension if done properly.

Steve,
All humans are tribal, built into the genes; civilization is the altruisic, group effort to overcome hedonistic, self greed.  A good part of altruisim is honesty.  A team works better with transparancy.  I suppose that can be interpreted or spun in both conservative and liberal ways.

Besides, "my radio is better than yours" is fun and we like it.  We like being informed and learning when possible.   Many of us have tons of radios which we all like for various purposes, some just 'cause they're pretty or have black crackle, or...     

I admire your propensity to keep things on track, logical and based on the trivium where possible... but being human is fun too.

I hear ya Rick. Not denying anyone being human. I just don't think any of this mine is better than yours crap adds anything of value to the conversation. Until rigorous measurements are made, all the rest is just prepubescent chest thumping. The reality is that most of these ballyhooed specs make ZERO difference in 99% of amateur radio operation. In other words, they are irrelevant and it follows chest thumping about them is also irrelevant.

You can increase the dynamic range of a system by oversampling and dithering even if the A/D does not have the dynamic range, this was used with radar 10 to 20 years ago when the best high speed A/D converters were 8 bits.    Usually the signal riding on noise provides a natural dithering and the higher speed of the A/D provides for the oversampling or signal integration.   When coherently integrating I and Q channels the noise is reduced by the square of the number of samples integrated.    It really works.

Does this not require extra clock cycles in comparison with having the additional parallel processing four bits?

Not sure what Flex uses.    With the radar, if you are using coherent integration, you take many samples of the I and Q channel and integrate each by itself, the result is a single I and Q value.   Do this many times and use the values in the processing.    The result is wider dynamic range if all the
samples in the I and Q are independent.    Also the spectral bandwidth is reduced and the signal to noise ratio is increased.    This requires higher sampling rates and wider bandwidth as compared to just one-to-one sampling.    If you can't oversampled then you are limited by the number of bits in the a/d converters.  I would like to see what type of processing they are doing in the flex with all this radar signal processing being applied to ham radio.

So you take the 2208 with 16 honest bits and sample at 122 or 130 MHz then compare it to a 12 bit machine running at 245MHz, where is the gain. You only need to divide it by 2 one more time.
There is no chest thumping here Steve just wondering what is the true performance.
I follow all new A/Ds that come out, kind of a hobby. the 2209 is the only one close to being better. LTC still seeems to be the best for us guys on the street.

Lots of chest thumping. Let's review.






You've never used the radio and have no measurements to go on, yet you've claimed it's marketed at fools. By way of contrast you proudly proclaim how little you've spent on your set up. Sounds like chest thumping to me. The reality is that most of these ballyhooed specs make ZERO difference in 99% of amateur radio operation.

I've made something like 6000 contacts over the past few years. I've heard tons of Flex radios but only one or two HPSDRs. If they were so great, I would expect to hear more on the air.

Overheard at Dayton this weekend:

"This flex-girdle is killing me!"

Over sampling also allows you to move what would normally be in-band noise up in frequency out of band. It's been done in audio A/D for decades. In fact, nearly all A/Ds in audio gear since the 90's have been using a 1-bit delta sigma converter and oversampling. This arrangement can produce up to 24 effective bits.

if you claim to be the best you should be the best
An 813 isn't a 4-1000A.
And yes I have operated the Flex 5000.

You haven't operated or made measurements on a Flex 6000, the topic of discussion here.

You've provided nothing that shows Flex isn't the best, just some meaningless A/D specs.

I'll still take my R-390 A.

Bill

22 receivers ahead of the 390A on the Sherwood list, including the Atlas XL350 from the 70's. WTF?

Flex has not provided any specs either and they are hoping to sell the thing
for large money.
I have only an interest in performance. So the first thing that I checked was one of the most important components and it doesn't add up.
I do think it has some cool processing and a better front end than the 5000 and HPSDR for that matter.
But IMHO the A/D looks pretty sloppy. ENOB now that is an interesting term.
150 counts of offset another.  

There are only about 400 HPSDR rigs world wide and I bet 1/2 are not on the air in TX. Most run low power. Heck, I just hit 1200 Watts PEP last week.

Drop some high level balanced mixers in a R390 and I bet it would climb up a bunch of notches. Still one of the cleanest LOs on the planet.

Definition of ENOB about half way down ...

http://www.analog.com/library/analogDialogue/archives/40-02/adc_noise.html (http://www.analog.com/library/analogDialogue/archives/40-02/adc_noise.html)

Frank ... your TX sounded excellent last night ... about the best you've ever sounded.  No carrier glitches noted.

Good reading Jay.
Notice the 9467 part spec is 95 dBFS NFDR at 250 MSPS and does not increase to 100 dB until you drop to 160 MSPS.
It would be interesting to see a FFT of the output to see what it does with all these extra samples.
Time will tell when real specs are released.

Maybe some chest thumping here:


http://www.youtube.com/watch?v=DMoWt0wnlUw

No surprise considering the online mentality for such arguments vs. real life. Brag, belittle, or otherwise thump your chest at a hamfest or on the air and a handful of people might hear you. Do it online and hundreds, potentially thousands could be exposed to it. As we've seen, that's what is truly important to some, not the reality of actual use and utility. The critical discussion is fascinating to read, but as to the P&Ming, you're right - it brings nothing useful to the table and gets old pretty fast.


Rob, you should read this as AM (whatever you choose to operate the mode with) and Vintage Radio. Rather than an exclusive site for vintage AM, it's all-inclusive for AM specifically.

I'm an old gear buff myself and, for all its faults, never see that interest waning. I've never been particularly interested in plastic radios or especially the newer computer-based SDR stuff. I enjoy working both on the air, though. As great as they sound, I could be listening to a plate modulated broadcast transmitter. I can't see 'em, sooo..... ;)

In the last few years a number of things have happened to pique my interest in SDR. First was a thread that G started on here about the Softrock application for his R-390A. Then I got to see a new-ish receiver with panadaptor and other excellent features at Steve/K4HX's station. Then earlier this year or late last, Rob/yAEX posted some samples of signals he'd received during the HMR, including what my signal looked like. Finally, after NEAR-Fest on our way south, Steve and I were treated to a demo of Chuck/K1KW's new station which absolutely amazed me. An excellent visual of what we were hearing.

So, while I'm not in the market at this point for anything 'new', I have developed a distinct appreciation for its utility. Hell, I just got a decent scope hooked up as a mod monitor in recent years. But I can certainly envision a time in the future where some of this SDR-type stuff could be very handy in conjunction with the current big iron here. Beyond the more typical features-for-the-sake-of-more like endless memories, this is stuff that is really handy for AM.

My own preference is to use simple rigs with analog filters, but the Flex radios, as a group, have the best sounding AM modulation heard on the ham bands.

I'm gaining a real appreciation for radios that don't hurt my back when I lift them

Man was borne to suffer.


klc

http://en.wikipedia.org/wiki/ENOB

as I lernt it. A trick the sound card guys played with early 24 bit audio A/Ds that actually had a hard time making 18 bits.

8) 8) 8) 8) 8) 8) 8) 8) 8)

Maybe best to Zip it!

Why?

Good stuff on ENOB.  Such "posturing" is what drove the space race, put USA on the moon and spelled the end of the USSR.  Say did you all see the first private launch from Cape Kennedy this morning?

Perhaps here's one of the more authorative drivers for some of Flex's turnabout.  None other than Phil, N8VB
-Related to the ENOB thread and interesting.  Also the vaunted Sherwood list is mentioned.

Following is all quoted from the QS1R yahoo board.
"
--- In qs1r@yahoogroups.com, George Works <worksg@...> wrote:
> Flex used to respond to questions on their reflector about direct
> sampling by saying that it had some benefits, but so did their analog
> filtering approach. But I suspect that their heart wasn't really in this
> response because Flex went direct sampling on their premier 6000 series
> product.
> George, PJ5/KJ6VW

Hi George,

Well, it's all more about posturing for marketing purposes these days than
dealing with reality.

I remember certain Flex Radio people arguing that Direct Sampling would never
approach the superior dynamic range of the QSD/audio ADC approach. It seems
like Flex has now changed their tune on that claim. I also remember a specific
Flex radio associate, that should have known better because of his background,
claiming that the ADC in the direct sampling receivers would be subject to
severe overload when all the signals within in the bandwidth of the ADC
magically added together to exceed the ADC clipping level. What he
forgot/missed was that in the real world signals on the bands are not
synchronous and do not add together like he was implying. Superposition does not
work that way. That scenario is only possible with synchronous signals
generated in the laboratory.

Flex Radio purchased a QS1R a few years ago to play with, so maybe their
experience with it helped to change their minds about direct sampling in
general. Also, the Perseus surpassing the Flex-5000a on Sherwood's receiver
list also opened some eyes I am betting.

73 Phil N8VB
"

Flex had nobody to design the direct sampling for them. There is more to that story.
The smart thing they did was to add additional filtering up front.
I do see my A/D saturate from heavy static. It is usually time to disconnect the antenna. I also see the noise floor jump up when there is a strong local signal. A good tunable preselector will be a big help. A Harris, Collins or Cubic prepost selector is the balls for a front end. I use a dual 13 filter Racal on my dual RX set up. When I get hermes the tunable Cubic will get pressed into service.
Direct sampling drops the parts count but you need to use good hardware to take advantage of the software. Flex put some real horsepower behind the A/D but I think they could have used a better A/D.
GIGO.
I think there needs to be a few more A/D bits before the analog guys are left in the dust. We seem to be stuck at 16. All the shuck and jive in the FPGA doesn't get you there,,,,yet. 

There have been some interesting "back and forths" between quite a few Flex owners and several of the Flex hardware/software engineers regarding the selection of the obsolete TMS320C6A8167 C6-Integra DSP rather than the TMS320DM8167 DaVinci DSP. As it turns out, this is due to confusion with a prototype design they worked with and the wrong information was put into the product brochure specs. In fact, the units now being built use the TMS320DM8167 DaVinci DSP.

Additionally, the matter of 16-bit ADC resolution rather than 24-bit as in the 5000A has been vigorously questioned in the technical forum. Several Flex engineers have gone through the math with the skeptics and are now referring them to the Amtel documentation which walks through the use of decimation and oversampling to increase dynamic range. The link they are supplying for the skeptics is here:      

http://www.atmel.com/Images/doc8003.pdf

One of the Flex engineers has said that they are enjoying the discourse and are willing to answer any questions. Anyone can join the Flex Reflector and swing away:

http://www.mail-archive.com/flexradio%40flex-radio.biz/index.html

They've also indicated repeatedly that they doubt anyone will be disappointed with any independent lab tests the 6000 series might be subjected to. Should be interesting to see what the independent tests turn up, but more important to me is whether or not the rack handles can be purchased in chrome rather than black.

Rob W1AEX

Interesting Rob, I wonder how the first release will work.
Will it be bug infested or work clean.
If it is clean then it will be fair to charge for software upgrades.
I think this will be very interesting to watch over the next couple years.

Do any of you Flexers know if they went with their plan to re-write the software in Erlang?

Kind of like la der des ders, "the last of the last", IOW the war to end all wars, 1914-1918.

Wait till some proprietary chip craps out 5 or 10 years from now, and they inform you that the part is "discontinued".

The way they list the price at $3999, $5999 and $6999 instead of $4K, $6K and $7K says it all. This tells me it is closer to the realm of "consumer" electronics than to that of amateur radio. At least they don't add that 9/10¢ at the end the way gas stations do.

I have to say they do sound pretty good on AM when adjusted correctly, and I like the way W9AD can operate his remotely from Mexico and sound exactly the same as he does when in ILL (I could never get away with doing that here at my QTH; lightning would wipe the thing out before the first month went by).  I have heard some on the air that, if the op had told me he was using a converted BC transmitter I would have believed him. But still, I could buy several ready-to-convert BC rigs plus spare parts and tubes, and a couple of 75A-4s or R-390s, and still have enough of my $7K left over to purchase the hardware to build an effective antenna.

One thang can be said about the SDR and Flex radios in general...

To become really proficient using one, it takes hundreds of study hours. One has to master the software, hardware and operating skills.

Sure, we could JS it and just get on the air by feel -  or we can do what Rob/W1AEX has done and spend several years diving DEEP into everything these rigs are capable of... and it's a lot.

Then there's Frank who eats and sleeps the hardware to the point of being expert in the leading edge of TX and RX ham use.

Some of us may discount the skills required to get really good at learning the deepest parts of the technology, but as one who has built up an HPSDR rig here, I was on the hairy edge of confusion most of the time -  and find working on the big AM tube rigs to be easy in comparison.  SDR stuff requires a whole new set of computer skills that are normally not used with a business/pleasure PC.

They may not be running heavy iron, but an SDR AMer who has FULLY mastered his rig sure gets my respect....   ;)

T

Tom,
You went through the learning curve so know how it feels. It is great that the gang is there to help. The smart guys who help us dumb guys really get my respect.  Guys like Phil Harman and Joe Martin etc make ham radio fun.

Is this the same "GFZ"??

Yeah, you'd never know it.  ;D

He's 009 now.
160 MHz ave.
DDCville.

Yes it was Pete, and based on the past, the Flex rigs came out with many software bugs.
So if Flex ships a clean product and charges for new features it is fair but if you have pay to clean up bugs it is a rip off.

Rick,
Gaggleville

I agree. If a feature is advertised as being in the rig, and that feature doesn't work as described, and a software upgrade (bug fix) is required to make it do was it was advertised to do, you shouldn't have to pay for it. However, if the feature is enhanced, or updated to include some additional functionally, that should fall under the payment category.

So based on their track record would you think the first year should be free?????

The first year IS 'free'.

======================

One (1) year SmartSDR™ software update and FlexAdvantage™  technical support contract for the FLEX-6000.  This provides for a year of SmartSDR software updates and FlexAdvantage technical support for the FLEX-6000 family of software defined radios.


NOTE: This is not required for new radio purchases as it is included in the purchase price of the FLEX-6000.  You can however purchase multiple years of SmartSDR updates and FlexAdvantage technical support if you so desire.

Don,
Your tax dollars have contributed to the design of this radio so your only option is to not pay for it twice.

Frank,
Margareritaville
.hey, at least we can sing it.

but not spell it.

That's what my IPad put in, seriously .  ;D

hmm, infallible machine driven to drink.

Flex Radio's are good sounding radio's...

But why some echo?

HUH??

Was a common problem with 1000's. I think maybe flaky connectors or some such. It's been a long time.

Ah yes, the infamous echo. One of the quirks is that if a station hasn't paid attention to proper grounding, particularly with the computer that is handling the DSP chores, RF feedback will be induced into the process at some point. With the newer rigs it's frequently the firewire cable that's the point of ingress. At any rate, the RF feedback and the transmitted signal will differ by the amount of latency present in the computer, so you end up with the infamous echo effect. Proper bonding throughout the station resolves this, but lots of hams seem to struggle with the concept.


Tom, that's very high praise, but my level of understanding is microscopic compared to Frank, Pete, and others around here. I'm just really good at faking it!

Rob W1AEX

Ah, now my brain is working. We jumped from the Flex 6000 allllll the way back to the SDR-1000. There were many issues back in "the good old days" with the SDR-1000 cabling, grounding, and latency. Echo, as a result of latency due to poor computer processing, can be a big problem with some of older computers. As Rob points out, even with the newer machines proper grounding, shielding, etc. are very important in keeping the RF out of the circuitry to eliminate the echo effect. I tried running my Flex on a Dell (bought in 2000). The echo effect was so bad, it sounded like I had a 60's car reverb in the audio chain. The only thing missing was the bouncing springs effect when you drove over a rough road.