Adaptive pre distortion on a 6500

  • 2
  • Praise
  • Updated 3 years ago
Yes the rumors are true Adaptive predistortion while using a 6500 was heard on 75

K2WS ALAN Congrads .

Hardware version right on. Great signal great imd Straight line signal.

And just think you didn't have to beg to get it.
Do hope you offer it to all
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes

Posted 3 years ago

  • 2
Photo of Peter K1PGV

Peter K1PGV, Elmer

  • 541 Posts
  • 314 Reply Likes
OK... I'll say it: I don't understand a single thing in this post.

Well, I know what ADP is... But beyond that.

Could the OP perhaps post something a bit more clear, to avoid our having to guess what he means?

Peter
K1PGV
Photo of Reg

Reg

  • 120 Posts
  • 73 Reply Likes
Peter:

It is actually a very clever process.  If you have a transmitter driving an amplifier, then the amplifier is going to introduce some distortion into the transmitted signal because it is not perfectly linear.  Now if the transmitter samples the amplifier output and distorts (yes, distorts) the transmit signal in a manner such that the amp distortion results in a perfect signal then you have adaptive pre-distortion.  The process is sometime called "adaptive pre-distortion" and sometimes referred to as "pure signal."

Quite some time ago Steve posted that it was not a top priority at the moment but they would eventually get to implementing it in SmartSDR.  No timeline.

Reg
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
I believe many of us understand what ADP purports to do (I posted how Warren described how he made it work at Friedrichshafen 2014) but I am with Peter in being unable to understand what exactly Ed is talking about or how it might have been accomplished.
Photo of Jim K4JAF

Jim K4JAF

  • 288 Posts
  • 71 Reply Likes

Yes, the post is somewhat cryptic and lacks details..  Please supply more details please.


Photo of Jay / NO5J

Jay / NO5J

  • 1364 Posts
  • 196 Reply Likes
Why the secrecy?
73, Jay - NO5J 
Photo of Martin Ewing AA6E

Martin Ewing AA6E

  • 252 Posts
  • 52 Reply Likes
At Dayton, FRS said that they had polled us users about APD/"Pure Signal", and only a small minority said it was important to them.  Many people may not know what it is, of course.  But the main beneficiary of super linear transmitters is everybody else on the band.  I.e., you're being kind to your fellow ham.  Good for you, but don't be surprised if not everyone is keen to spend money or development time on it.
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
No secret

I have written ,requested ,talked about pure signal sooooo!!! many times i just took for granted that someone actually was reading .


Reg,s description is pretty good .


Now this event was done with the 6500 Flex and hardware , the anan uses software . There were also 3000 and 5000 flexes using hardware to accomplish this pure signal.


Results

IMD 3rd order -50 , -60 , -70DB , down.

With no excuses just clean transmitted energy .


If you sample and include the amplifier both the radio and amplifier will exhibit this similar trait.

When you look at the signal on a panadapter you will see the straight line going from the noise floor right up to the max signal. unlike any any radio out there without this adaptation .


So there is now a 6500 flex out there running a hardware version of pure signal FANTASTIC .


There was no begging to implement , just a judicious dose of pure ingenuity . Hardware or software this is a beautiful way of making a transmitted signal CLEAN .


For me I use Pure signal every day (Software ,external sampler, . and now a 6500 user can also .


ALAN again what a accomplishment WOW a hardware preview of Version 2,000 or whatever .

For all who don't want to produce a clean footprint
We know who you are
(Edited)
Photo of Peter K1PGV

Peter K1PGV, Elmer

  • 539 Posts
  • 310 Reply Likes
So, I think you're saying:

"Alan K2WS, a clever guy who's been involved with Flex SDRs since the Flex-1000 (remember his OXCO mod?), has managed to perform hardware modification to a Flex-6500 to implement ADP... and you like it.  A lot."

Did I get that right?

Maybe Alan will be making a mod kit available?

Peter
K1PGV
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
No secret
I have written ,requested ,talked about pure signal sooooo!!! many times i just took for granted that someone actually was reading .

Reg,s description is pretty good .

Now this event was done with the 6500 Flex and hardware , the anan uses software . There were also 3000 and 5000 flexes using hardware to accomplish this pure signal.

Results
IMD 3rd order -50 , -60 , -70DB , down.
With no excuses just clean transmitted energy .

If you sample and include the amplifier both the radio and amplifier will exhibit this similar trait.
When you look at the signal on a panadapter you will see the straight line going from the noise floor right up to the max signal. unlike any any radio out there without this adaptation .

So there is now a 6500 flex out there running a hardware version of pure signal FANTASTIC .

There was no begging to implement , just a judicious dose of pure ingenuity . Hardware or software this is a beautiful way of making a transmitted signal CLEAN .

For me I use Pure signal every day (Software ,external sampler, . and now a 6500 user can also .

ALAN again what a accomplishment WOW a hardware preview of Version 2,000 or whatever .

For those who dont want it
We know who you are
Photo of Jim K4JAF

Jim K4JAF

  • 205 Posts
  • 57 Reply Likes

Still no details on how "pure signal" on the 6500 was accomplished.. 

Details please, your so cryptic in your answers..


Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
please read
It was done with a hardware package attached to the 6500 . It emulated the compare and re config in hardware .

If you wish the exact details ASK ALAN I listed the call.
He may share it with you .

II guess you can also ask Flex
FYI there are many Hams who have contributed to this Pure signal event.
It is just really great to see a 6500 using this function
(Edited)
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 1638 Posts
  • 426 Reply Likes
Gerald already said the radio is set up for pure signal a long time ago, this shows he was right. Something to look to when flex finally does it.
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Actually I really don't like to post here. And if it wasn't a 6500 I never would have .

See It is pure innovation .

And BILL I'll be as kind as I can " Its questionable on FLEX ready ." I had a good laugh on that one .

So please excuse me I,m on Island time here .
Photo of Richard McClelland, AA5S

Richard McClelland, AA5S

  • 252 Posts
  • 45 Reply Likes
Periscope up.  Another pre-distortion controversy.  Periscope down.
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 1638 Posts
  • 426 Reply Likes
Since Flex has been truthful in the past, I will accept what Steve said, it is already set up for pure signal, that's why it's working. And it won't need a crummy add on box to do it.


Photo of Steve - N5AC

Steve - N5AC, VP Engineering

  • 657 Posts
  • 534 Reply Likes

Official Response

When we designed the FLEX-6000 we had adaptive predistortion in mind so we included a tap from the transmitter that feeds back into the receiver. This gives us simultaneous access to the desired and actual transmit spectrum in the FPGA so that we can implement linearizarion. We also have sensors for temperature, voltage and load match of the PA as these can be used to improve linearizarion. The work that needs to be done for the FLEX-6000 is to experiment with algorithms and input sources to determine what gets us the most bang for the buck and then finalize the implementation.

The FLEX-6000 already has the highest precision DAC of any amateur transmitter and excellent IMD numbers so we decided to focus first on features that the community is waiting for and return later for implementation of predistortion. Predistortion in the cellular business is critical because they are putting many carriers in a single PA and spectral purity when you're doing this is critical. But for ham radio, it is more of a "nice to have" and "nice neighbor" sort of thing -- it's not going to prevent someone from copying YOUR signal and the FLEX-6000 spectral purity is well above the legal limits. Barring significant hype that makes this a marketing issue, we'll look into this after finishing the items on the road map that you, our customers, have told us are important for you.

Since we thought ahead, you will not need to use an external device to tap the actual transmit signal and feed it back into the receiver when the time comes. Incidentally, our VP/GM of our government business, Graham KE9H, has considerable experience with predistortion having run a company that focused on predistortion in cellular PAs. He also developed PAs for HPSDR.
(Edited)
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 1638 Posts
  • 426 Reply Likes
All that said,,Nice work Allen, I wonder if someone will build this box and sell them to people who want one?
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
Flex has an Open API...

Without getting into the details of how to implement ADP, I would suggest that if there really is so much pent up demand for ADP, then perhaps some third party could write the software to implement it using the existing hardware taps that Steve identified.


OTOH.. if as I suspect that it is a "Nice to Have" feature and since no third party seems to be rushing in to write the ADP Software then perhaps we will have to wait for Flex to get around to it sometimes after V2.0 WAN and other necessary features which have a huge "Need to Have" demand
Photo of WA6FXT Mike

WA6FXT Mike

  • 207 Posts
  • 37 Reply Likes

AMEN, to that!

73's

Photo of K1UO - Larry

K1UO - Larry

  • 705 Posts
  • 117 Reply Likes

It will be so funny to hear all the chimes of "wow", "this ADP is awesome"  and "why didn't they do this sooner"" come out after a few more version releases of SSDR.... It appears that everything is relative after a period of time and research.


Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Keep you fingers crossed. Spin around a few times. And drink lots of water.
Yes it will all be better
Photo of k3Tim

k3Tim

  • 801 Posts
  • 149 Reply Likes
I tried this, it didn't help.
Please advise
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Did you click your heels. 3 times????
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 640 Reply Likes
I was not aware there even was a rumored sighting of a 6500 running apd on 80 mtr the other night. If it doesn't even exist, my guess is the 'rumors' were false.

Why is there such a 'clamor' for 2.0 WAN when it appears those wanting to work remote already are?
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
Because running rWAN via VPN is a kludge. A sit does not solve the major issue of running WAN without a PC attached to the 6000 when the 6000 is located remotely.
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Warning warning. Cud a is out there. In test.
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Actually I would hope everyone copied my signal

There was growing pains involved .A large amount of fellow hams helped .
And no one should stop were hams are now Call it Progress. Driven by innovation , Experimentation and really great gifted people not sitting on their hands and minds

Finally someone actually recognized and thanked ALAN ,
Thats what the post was about .
Not please ,Pure signal . Not even a discussion about it .

So hats of to Dave Alan Doug Gerald and about 1000 others for actually building implementing a system that produces a cleaner signal.

By the way Software for this function being implemented into a LOCKED system

It Wont Happen by a 3rd party vendor What a laugh

If it's pure signal in software has to come from Flex

Oh yes !!!! A complete closed loop control is prefered . Not just in the box (Including any AMP )
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
The API is OPEN

Third Parties are developing all sorts of useful software. Already.
Photo of Walt - KZ1F

Walt - KZ1F

  • 3040 Posts
  • 640 Reply Likes
Not for the radio they aren't. My guess is adp needs to be implemented in the radio, not control surface.
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
FreeDV is a Third Party Product inside the Radio
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2081 Posts
  • 495 Reply Likes
I would not be surprised Steve has already been testing ADP in the lab to be sure things are set to go before they release it later.
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
ADP provides some benefits when transmitting barefoot.  But if you do not have some external box to sample an external AMP then the AMP will just reintroduce IMD.

I wonder if Flex will include an ADP sample tap in the new 4O3A amp?
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2081 Posts
  • 495 Reply Likes
Howard, I bet they have that all thought out, and I bet Steve has an external box as you said before to connect an amp to the Flex for sampling. But I expect this new amp will have it all
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Gee I think I'll make the buttons red.

If you want innovation package the software and lease development packages.
Then stand back
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2095 Posts
  • 497 Reply Likes
As flex is already set up for pure signal as Steve said, and they already know how they will implement it.  Why don't the pure signal community do the same with an Icom, now thats doing something....since the flex is already pretty clean.
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2095 Posts
  • 497 Reply Likes
I was listening on 40M this morning to a couple guys talking about the pure signal on the flex. They said that getting it to work on the flex has been very hard work and nothing simple about it. It is working but not even close to being ready for prime time. They said that if it were simple to implement flex would have done it a long time ago. No truth to that as we all know.
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
NAH!!!!! The KX3 wants to go remote . but they are building theirs so it will fit on a wristwatch

FYI :::::: The flex solution that Steve expounds is a in the box solution . Let me be clear that means there is no provisions for closing the loop to include the amplifier.

Here is the difference. In the cell phone world (Where Pre or pure signal came from) there were no AMPLIFIERS to even consider .JUST many many users

It's not -60 plus DB down in the box its into the noise level even with a external amplifier.

OR Solid state amps that you can live with , less than clean amps that you can live with .
That we all YES We all can live with .
Photo of Peter K1PGV

Peter K1PGV, Elmer

  • 541 Posts
  • 314 Reply Likes

Anybody know K2WS who could contact him and actual get a clear, factual, account of what's up... in proper sentences that make sense?

Like I said before, I know he's a very clever guy, but that's all.  If nobody here is acquainted with him, I'll be happy to blindly send him an email and ask him "wazzup."

ETA: Not that I care that much about APD, as a feature it ranks just below the user-customizable clock, but I'm always curious to know what's going on.

(Edited)
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2095 Posts
  • 497 Reply Likes
I am not an expert in this, I'm sure Steve has it all worked out, it was in the plans when they designed it. It will be interesting to see how it all unfolds don't you think?
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Yes Write ALAN A very positive move. and a great guy .

OR AC2IQ OR Warren . Or John K2YN He has a auto version under covers. (ALL Fantastic individuals) And There are more out there PLAYING with their toys .

Bill !!!!The plans did not include any external use . This really changed the dynamics of the application . AMPS were just not in the equation.

I would really hope a Available hardware solution would be offered for a 6500 . It would be definitely 2 or 3 light years ahead of the curve.

.AND I really think it could be built inside .
Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3403 Posts
  • 1292 Reply Likes
I have had several discussions with Greg and Steve at Flex about Adaptive Predistortion [ADP] and how it might be implemented.   FYI ("Pure Signal" is a ANAN Product Name)

Flex obviously plans to get around to doing it but clearly ADP is still a low priority item as there has been little visible demand for it except from a few ADP Fanatics.

ADP helps with SSB.. it does not do that much, if anything for CW and likely can make things worse for some digital modes. So, not withstanding the experience of the digital cell phone industry, there is a lot of experimentation still to be done for the ham radio environment.

With the hardware taps inside the box, it is already possible to write software to do ADP in a 6000.. As much as Ed dismissed the possibility to have a third party write ADP for the 6000, the API is open and waiting for someone to do it... 

Using an external amplifier is a totally different issue as one would need a hardware box to sample the external amplifier's output and feed that back to the internal tap...  That is definitely going to require some hardware engineering to get it right without cause unforeseen consequences. In addition, different brands of amplifiers have very different non linear response characteristics, so I suspect that there will need to be much tweaking of the software parameters to customize any ADP hardware/software solution to work with a specific amp.

All in all ADP is a major and potentially very expensive Science Project
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Howard
It is truly beyond the scope of API . There is a necessity to have access to a processor and that is just the start.

That is Also correct Some very intelligent guys worked on this for the past 10 years . The result I use every day on every band at every power level with exactly the same quality expected .
Oh yes on a litany of amplifiers.

Saying that only came true when Warren gave API auto function .

Can an outside source make it for FLEX Absolutely . Right now I would say Flex has a garage full of headaches . Or 2.0 would have been there years ago.

So to be fair the hardware approach a nice interim great accomplishment .
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2095 Posts
  • 497 Reply Likes
Here is were I stand on this Howard. If ADP does not happen for the flex, I will be a little disappointed. But the Flex will always be a radio of choice with or without ADP simply because it is a wonderful radio. And I love many of the plans for it, sets it apart. And anything else not using SSDR, I'm not interested. SSDR rocks.

The nice thing is, the Flex just as it is right now has a small foot print and very clean, would ADP make it better? sure, but never a deal breaker for me, I'm happy with the performance.  On the other hand I have seen an Anan without ADP and they don't fair so well, so they really do need it.

That's how I stand on this. And we all know flex will not drop what they are doing to implement ADP. They will stick to the plan as they have set out. There just has not been a very large group asking for ADP. It seems strange to me some who do want it are not even Flex owners,,,funny.

I think ADP will be part of a package after V2 for purchase if anyone wants it. Till then I will gladly wait, no big deal.
Photo of Peter K1PGV

Peter K1PGV, Elmer

  • 541 Posts
  • 314 Reply Likes
APD is primarily being used by Apache Labs (the makers of the Anan radios) as a marketing tool. How can they make their open source Hermes-based radio that runs open source PSDR stand out (and get Americans to pay them a lot of money for it)?  Answer: They point to, and actively promote, their only differentiating factor: What they call Pure Signal.

There are tons of nice features in all sorts of radios. Users will evaluate which features they care about, and vote with their wallets. I voted when I bought my Flex.

That doesn't mean I'm not curious about what K2WS has done. Which is how I got stuck in this thread in the first place. Since nobody's yet volunteered, I'm going to drop K2WS a line and ask him what's going on.

Peter
K1PGV
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Don't procrastinate drop ALAN a e mail

And Actually There are more differences than just pure signal. Both have there good points .
SDR radios are fantastic .

I did not start the thread to sell ANAN or Pre distortion or Bash Flex

It was , and still is to praise ALAN for the 6500 Flex Pure signal or whatever you wish to call it operational.

And if you, Anyone, really understood how difficult this task was or is you too would be praising his attempts.

There would be no GEE WiZZ flex has it planned or ANAN does it better or the best reason Mine is already great . Yea sure.

Possibly you would understand it is only as good as delivered . And only better as in HAVE .

Hams actively pursuing innovation and demanding the best bounce for the buck drive the future
Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2087 Posts
  • 496 Reply Likes
Yes Congrats to Allan and all who worked on that project, while it is interesting I fail to see how flex owners can benefit from it, unless it is soon to be ready for all to use.
Peter I am with you, since this is with a flex 6500 maybe Allen will post an description of the work done and how everything connect, and how SSDR works with it. As this is about the ham community he should not have any problem sharing.
Photo of Stan - VA7NF

Stan - VA7NF

  • 392 Posts
  • 83 Reply Likes

Some notes:

Congratulations Ed:  ADP had disappeared from this forum for a while, you brought it back for a brief peak;

Be careful what you wish for:  Remember that a wide signal is a benefit to contesting.  The same feedback circuits that reduce splatter can be used to increase it to objectionable levels.  Just imagine how far I can keep other stations away from my running frequency with custom tailored key clicks and IMD. 

Oh yes:  FRS can you introduce a series of enhancements so I can add a tailored chirp or buzz to my CW so it is much more recognizable in a pile-up or to keep a run frequency.  (tongue only slightly in cheek)

Photo of Rick Hadley - W0FG

Rick Hadley - W0FG

  • 529 Posts
  • 116 Reply Likes
Yeah, that's the ticket...a tailored chirp so we can sound just like a Polish station from the 1960s!! ;-)
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Actually ALAN reminded me with his 6500 and pre distortion.

Yes I try to stay away from this forum . This occasion was a KUDO to ALAN .

It brought out the usual snide comments and ignorance of some . Hopefully it did not diminish the True success of ALAN and his 6500 .

I do care if Flex planned Pure signal. Promised , Lie or will offer it tomorrow. It is a fantastic system . And very few radios out there can duplicate this feat .

I do know of one however that produces -60DB plus all day long without pure signal
A story for another day.
And to coin off one of the comments Designed with this in mind .

As we go into the down cycle of dxing more and more hams will gravitate to local bands for entertainment.
Good operating practices , Pure signal makes a good neighborhood
Photo of Peter K1PGV

Peter K1PGV, Elmer

  • 541 Posts
  • 314 Reply Likes
OK, folks...  from the man himself.  Thus spake Alan K2WS:
I have built a hardware APD for use with the 6500.  It is strictly a personal project to learn how APD operates.  I was very impressed with AC2IQ’s demo of his hardware APD in the winter of 2013.  Dave was not willing to share his design schematics, so I did a lot of research and cooked up my own!   I have discussed the project with others on 75, but have not demonstrated it on the air... yet!  

My system does work, but there are a few loose ends I’m cleaning up.  Perhaps in a week or so my APD will be on the air.

It works by comparing the signal envelopes via separate log detectors, with a difference amp ;  the internal  0dBm Exciter signal in the 6500 with the external directional coupler’s signal ( barefoot or Linear Amplifier).  That  difference amp generates the error signal.  It drives the gain control port of a VCA which pre distorts the input signal and then drives the 6500’s 100w PA. 

It's always best to go directly to the source and no believe the noise in the system.

Many thanks to K2WS for taking the time to clarify what he's working on.  I admit that it does sound *very* interesting.

Peter
K1PGV

Photo of Bill -VA3WTB

Bill -VA3WTB

  • 2095 Posts
  • 497 Reply Likes
Thanks  Peter, again great job Allen.
Ed reported it being on the air on 75M
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
Glad Alan let you know what is happening. The absolute important thing is ALAN discribed what he is doing with the 6500 not me .

The really neat part is that it is a variation of what is out there . Not AC2IQ not Warrens but ALAN's
How cool is that . Still Pure Signal or whatever
.

It's semantics my friend Lets see Demo , Testing

a difference???

I see it as a polite I'll show you later.

Lets see it was -58 down and yes on 75 TESTING Or was it a Demo Gee I dont know.

Look guys whatever you want to call it it's pretty good.

The results will be getting Better Long before that 2.0 BS WOW.
OH did I mention a recording GEE WiZ I forgot that .Sorry BILL And any one else stop defending FLEX and wake up and smell the roses .
(Edited)
Photo of Ed baker

Ed baker

  • 114 Posts
  • 8 Reply Likes
On another note :
There is a transceiver out there Yes on the AIR Gee I have a good idea on that one too ,which not running Pure signal Blows the FLEX or anything else away on 3rd order IMD .

With all implemented how does -80DBM sound boys . It's creative engineering my friends by HAMS . And Yes I Have been in QSO with him. ON 40 Demo or testing WHATEVER
Photo of Jim K4JAF

Jim K4JAF

  • 288 Posts
  • 71 Reply Likes
I think it may be a good time to put this thread to rest.. 

This conversation is no longer open for comments or replies.