Roadmap inclusion proposal

  • 5
  • Idea
  • Updated 4 years ago
Photo of Steve - N5AC Steve posted this in the general forum.  However, I moved it here to get a wider response to my proposal.  See my proposal below his comments.

Steve - N5AC, VP Engineering

  • 331 Posts
  • 185 Likes



The roadmap shows major features we are working on that we believe are core functionality our customers are looking for.  As we talked with customers about the scope feature, we found that even though it was a feature of PowerSDR, most customers did not use it and stated that it was not a feature they were interested in having quickly.  For this reason alone, we dropped it from the roadmap.  This is not to say that we will not do a scope function, just that it is no longer a key feature that we are promising in this time-frame.

The same thing can be said of predistortion -- we've been asked about it a lot, but when we ask which is more important to you as a customer: predistortion or remote capability, remote almost always get the nod.  We asked the same question at the FlexRadio Banquet at the Dayton Hamvention and out of ~180 customers, only 2-3 raised their hands for predistortion over remote.  For us the message is clear: focus on network and bring the other capabilities out after.

I hope that we've demonstrated that while we are working on the big-ticket items, we are also innovating and creating smaller features all along the way.  Band edge markers is a personal favorite of mine that's I've been discussing with staff members at FlexRadio for over two years.  I want to get to this and make it work well, but networking and the other items on the roadmap will be our focus.  We'll fit in other items as we have time.


Steve,

I like to suggest that in the future the poll questions should be place here rather than at the conventions. 

I believe this menu has more active participants and a wider audience with diverse interest than any other place.

For example, pose the question and give the forum members a week or two to vote on the proposal. 

Only one vote (remember we are not all in Chicago) per radio purchased.

On the ballot the purchaser name and call (if they have one) is verify by Flex records.

This would be the fairest way to get the members feeling on what want or like to see in their radios.

Let the majority rule!




Photo of Ernest - W4EG

Ernest - W4EG

  • 583 Posts
  • 113 Reply Likes

Posted 4 years ago

  • 5
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1678 Posts
  • 573 Reply Likes
Ernest,
Good idea -- Here is something that is different but related.  I put in a suggestion on the "GetSatisfaction" page about the "like" button on ideas.  

Since most of the communities like FRS use the number of likes to rank ideas, it is important that the method to support an idea is clear and easy to use.  They used to label the button "ME TOO" but that was confusing so they changed it to "LIKE."   However that can also also be confusing, especially since you can "like" a comment in the post but that doesn't add any support to the idea.  And it is not clear by looking at the label whether you have added your support previously or not.  

See the snapshot below that summarizes the proposed change.  I suggested it about a month ago and the GetSatisfaction team agreed.  Here is the status:

Stephanie, Community Support Manager said: We agree that this makes sense to change it to "Vote" for ideas. I'm going to mark this as planned so we can add it to our roadmap. Thanks for feedback! :) 

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com


********** CLICK to see FULL size ********





 
Photo of Steve - N5AC

Steve - N5AC, VP Engineering

  • 1007 Posts
  • 969 Reply Likes
FYI we are always talking with customers and asking what they are looking for in our products.  

The Ideas section of the community is monitored to see the suggestions and ideas posted by individuals that are up-voted by others in the community.  As of today, 9 of the top 20 ideas suggested have already been implemented and another 3 are planned.  Some of the suggestions made simply are not practical or possible so not much we can do about that.  My general recommendation is to post an idea and encourage others to me-too your idea.  We look very seriously at these and implement the ones we can.

The roadmap came out of all of this information and there are smaller items we are doing that just didn't warrant a mention in the roadmap.  The roadmap is a "big picture direction" about what we're doing, generally.  The banquet was where we chose to announce the roadmap -- we did not use the banquet as a forum for soliciting ideas.  One customer asked in the questions period about predistortion.  I commented that it is something that we want to do, but that I didn't feel that it had the support over remote/networking and that we have to do things based on priority.  To prove the point I asked for a show of hands.  But we already know, based on daily interactions with customers, that remote is the killer app and predistortion is a nice-to-have for most (but not all).

I have long felt that a grass-roots user group would be welcomed -- one that existed to carefully organize and poll for the important features and then organize and present findings.  The user community (here) is a big step in this direction and it really helps us do what you want most.  But an organized group would be neat.  We had these when I worked for Digital Equipment (the group was called DECUS) and it was run by customers.  Of course all the customers worked for companies that could send them to an annual users' convention.  As a DEC employee, we went to hear what customers wanted, to announce what we were doing and to help our customers be successful.

Our customers are awesome and I love talking to them and hearing about how they use our products and what we can do to make them better.  Designing, administering, running, presenting the results of and acting on a poll is a LOT of work.  I ran a poll a few years back about VHF contesting because the state of affairs so bothered me.  The poll results and the paper I wrote in the wake of that are posted on my website (it's 67 pages long).  It was a lot of work and after it was published many of the VHF contesting rules were changed.  I do not know to what extent my poll and results helped, but I like to think that it was part of the equation.  It was hours and hours of work to collate, consider and explain the results.  

The community gives us a very good look at what customers are thinking.  Would more be better?  Sure, but I think we're doing better in this regard than many companies.  Speaking for the engineering team at FlexRadio, we have lots of work to do just to get the things done that we've already committed to you.  Today, we're focusing on that work.  As we move forward, we'll always be looking at the next great ideas -- I think we've shown that we're not at all averse to doing and trying new things.  So a poll is something we might do in the future (but when we have a bit more time on our hands) or it might be something that you guys do.  If it is executed by the community and everyone gets a shot at expressing their opinions (it is perceived as fair), we'll definitely use it to plan what we're going to do and provide feedback on the things that don't make sense/we can't do, etc.
Photo of K1UO - Larry

K1UO - Larry

  • 731 Posts
  • 119 Reply Likes

Steve... 

While I can wait for predistortion I believe its something that ALL will benefit from on the Bands :-)  (some just don't realize it)   Ill bet a lot of the people at the banquet didn't respond positively to predistortion because they really didn't understand what it was or does.  I didn't, until I took the time to research it a bit.  I think enhancements to existing algorithms you are working on are what many are concerned with but when they not mentioned as an item in the roadmap. i.e. NB,NR, APF, split enhancements, docking of various panels etc.  Im thinking that's where the concern is coming from overall.  OTOH.. I've polished up the "Vote it up" button and ready to go........

Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1678 Posts
  • 573 Reply Likes
I meant to also comment that I agree --- FRS does a much better job than most of the other vendors regarding customer input and upgrade plans via the road map. It really shows your roots as a software company at heart.

Regards, Al
Photo of Dale KB5VE

Dale KB5VE

  • 410 Posts
  • 56 Reply Likes
Steve being from a industry that is required to do what it puts in print, I would like to see the items in the original sales pitch completed. I have been a happy loyal Flex user for years but all the fluff is nice but I feel what wast advertised should be the main goal of flex. As you know remote is high on my list but as long as we are headed in the right direction I can wait. If I was a project head I would have the advertisement on the wall with red x's going through what we had done and my goal would be to have that flyer all red ASAP. Then the fluff. Under promise and over deliver I have lived by that for 45 years of business success .
Photo of Robbie - KI4TTZ

Robbie - KI4TTZ

  • 479 Posts
  • 77 Reply Likes
Here's a really easy way for anyone to make a quick poll, of course it wouldn't be limited to those who have purchased a 6000 series:

http://strawpoll.me/

Create a new poll on that site and post the link here as a new topic. :)

-Robbie


Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1678 Posts
  • 573 Reply Likes
Hi Steve,
I didn't realize you worked for DEC.  I used to belong the user group here in Atlanta.  The company I worked for is still using some VAXes today.  (hard to believe we developed the software to run on 11-750s initially in the 1980s and migrated it to a number of the later VAXes).  We developed a lot of software in Macro32 and DEC C to run our manufacturing floor.   I had a full book shelf full of orange VMS manuals and microfiche.  Digital built great hardware and software, it was sad to see them go.  

I've done customer surveys before and agree they are a tremendous amount of work and only good for a point in time.  I think voting for the ideas on the community could be a good tool with much less effort.  But in my opinion there are a few drawbacks in the way the GetSatisfaction handles ideas:

- The voting process can be confusing (as I noted in the post above).   I think if everyone realized how it works and is used the numbers might be higher.  Ideas with over 20 votes are rare.  This seems low given the number community members. 

- There is not a good reporting tool.  Instead of paging through all of the ideas, it would be nice to have a simple spreadsheet type report with 1 line for each idea, the status, and a link to the idea for the details. 

Devoting time to an independent surveying process would be close to a full time job so using the tools you have makes sense.  Maybe developing some reports from the community data would be a compromise to consider.  It might encourage a higher level of user participation in the voting process. 

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com
Photo of Dave - WB5NHL

Dave - WB5NHL

  • 266 Posts
  • 60 Reply Likes
Steve, I remember that there were a large number of comments about improving the NB & NR functions. While obviously not a new function deserving an entry in the roadmap, I assume that improving theses functions is VERY high on your list of to do's.
Photo of Mike K5UX

Mike K5UX

  • 34 Posts
  • 13 Reply Likes
Dave, I'm really glad you mentioned this.  I think FRS needs to improve the DSP-related functions, particularly.....NB, NR, ANF, APF.  This stuff is considered basic essentials in any DSP rig today.  If I had to guess, I would think FRS is already working on this, but it never hurts to bring it up, particularly when there is a long silence and it's not on the current road map.

Mike
K5UX
Photo of Robert -- N5IKD

Robert -- N5IKD, Elmer

  • 488 Posts
  • 152 Reply Likes
I don't want a radio that is designed by committee. I want a radio that is designed by a visionary. 

That being said, I do appreciate the Flex crew monitoring our opinions and offering polls to gauge our interest.
Photo of Robbie - KI4TTZ

Robbie - KI4TTZ

  • 479 Posts
  • 77 Reply Likes
Very good point.  Best comment I've seen all year.
Photo of Lee

Lee, Elmer

  • 677 Posts
  • 285 Reply Likes
I'm totally in agreement with the above comment and I am totally against turning the radio's design into something political using polls.  I think its probably more profitable to have a robust discussion of pro's, con's and synthesis laid out to the engineering staff who have a much clearer idea of constraints and possible or impossible implementation.  It also better helps to control expectations.  If you have polls people will be constantly kvetching about "hey wait a minute 10 of us wanted the dohickey moved on the GUI and you haven't done it YET!!!!!!"    

Personally I want pre-distortion in the receiver as in noise abatement far more than I want the resources used to make already great audio better.  I would have been a vote of one hand.  But I bet getting rid of static crashes as we enter into summer may be a more desirable result for most operators (and a far more useful marketing tool) than making some distortion meter go from -45 dB to -50 dB.  I think 170 pro remote base to 3 pre distortion is a very clear and defensible consensus, but 15 jokers who want the whodicky on the GUI moved is very unclear and will just lead to factions and friction.

Another problem with polls is the radio is not a blank sheet of paper that allows you to noodle endless "what ifs", but three radios each with some what different capability and capacity.  These radios are not endless in their capacity and filling up the space with "political choices"  which were not really well thought through in the end will hinder you when a really good idea comes along.  Once you stuff it in there it's damn hard to get it out.  (witness: Obamacare) 

At some point you spend all of your engineering time chasing down rabbit holes and sort of loose track of where you were going in the first place.  The creation of skins in the SDR-1000 comes to mind as an example.  Lots of time to get a kind if silly cosmetic result.  Skins had a tiny bit of marketing juice but in the end were just a fad  (certainly not meant to disparage whoever spent the hours developing skins)

I think the road map is a really good idea.  It gives functional momentum to the design and I think if you can deliver more than promised then its all to the better.  I think Flex has gone a long way in creating some systems to drive both production and design as in delivering radios on day one of announcement and over delivering on expectations as far as software development.  
  
73  W9OY
(Edited)
Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 313 Posts
  • 132 Reply Likes

> "I'll bet a lot of the people at the banquet didn't respond positively to predistortion because they really didn't understand what it was or does."

That was certainly the case at my table.  At the present time, I don't believe one article on predistortion (PD) has appeared in QST.  It's time.  Although I subscribe to QEX, I just don't recall if the topic has seen any print in that publication.  The subject badly needs broader exposure that QEX does not provide. 

PD is not a marketing gimmick.  It has a 20+ year proven track record in land-mobile communications.   I have used it when owning an ADAT ADT-200A transceiver.  Many of us have seen real world experiments where 3rd-order IMD is brought down from -30 dBc to -60 dBc on +12V transceivers -- and actively kept low through continuous adaptation.  Witnessing PD on a Flex panadapter is a jaw-dropping experience.  Where PD really shines is when trying to listen to weak SSB stations next to very large signals.  The difference is truly night-and-day.   

Flex has managed to attain excellent stand-alone transmit IMD performance without PD.  But we would be better served if the Flex 6500/6700 could accept an RF tap from an external power amplifier.  Frankly, it would be a major disappointment (and waste of company resources) to put work into PD and not include an external tap - even if that tap required end-user or factory retrofitting . 

Paul, W9AC

 

 

(Edited)
Photo of Robbie - KI4TTZ

Robbie - KI4TTZ

  • 478 Posts
  • 77 Reply Likes
I gotta admit, I'm intrigued even though I've never heard of predistortion.
Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 313 Posts
  • 132 Reply Likes
Just sent it to you through e-mail.
Photo of Stan - VA7NF

Stan - VA7NF

  • 398 Posts
  • 85 Reply Likes

I question the need for a RF tap from an amp back to the flex. 

The flex predistortion would give a clean RF signal to the amp.  It is the amp's responsibility to faithfully amplify that signal.  So the amp should provide -ve feedback within itself to correct it's distortion.

Photo of Paul Christensen, W9AC

Paul Christensen, W9AC, Elmer

  • 313 Posts
  • 132 Reply Likes

Stan,

Good point.  Ideally, each manufacturer is responsible for the performance of their own product.  However, consider a real-world situation where a Flex 6500 is teamed up with a modern solid-state (SS) amp like the SPE series  or Prometheus.  While the Flex provides reasonable IMD performance, the top-end SS amp manufacturers have difficulty in obtaining numbers even into the mid -30 dBc area.  It's not nearly enough when working weak signal SSB next to big, dirty signals in the -30 dBc area on a mostly quiet band like 20m. 

At least two other manufacturers have implemented an approach to correct for the laziness of amp manufacturers.  ADAT provides adaptive PD from an external amp without modification.  I used it with my SPE 1K-FA amp.   Apache's transceivers do an excellent job at the task but require a bit of user modification.  Their product wasn't initially built with PD in mind.  

Whether or not Flex chooses to sample internally or externally is really not much of a software or control design issue -- it's a hardware issue.  Just a hunch, but I envision Flex implementing an externally-sampled PD solution after the completion of the roadmap, perhaps at the point of paying for continued upgrades.  I'm fine with that. 

I have seen (and heard) WA1OXT and AC2IQ demonstrate the effectiveness of PD when using an external amp.  WA1OXT uses an 8877-based amp while AC2IQ uses a SS amp.  The opposite sideband IMD suppression goes from ~ -30 dBc to -60 dBc.  That 30 dB improvement is huge when working in tight spaces, especially on 20m.  It's not a trivial "numbers" improvement.     

Paul, W9AC     

Photo of Richard Clafton W4/G7EIX

Richard Clafton W4/G7EIX, Elmer

  • 455 Posts
  • 117 Reply Likes
I also did not know what it was, and have now read a few snippets on the web.... and this video - even though it is of an ANAN seems to show the effects in a way my mind seems to be able to absorb.

https://www.youtube.com/watch?v=o8h6ParzsAs
Photo of Ned K1NJ

Ned K1NJ

  • 301 Posts
  • 75 Reply Likes
I've seen  PD in action. It is truly amazing, and a step forward for amateur radio.  Also true
that the feedback should be taken from the amplifier output.  It may also be true that many
may have made their Flex purchase with remote capability in mind.  So since this was
an advertised feature, perhaps it should come first.  Yet, we now know that Flex has PD
as something in the works, and want to provide this feature soon.  It's good to know that
eventually we will have both.

Ned,  K1NJ

Photo of KY6LA - Howard

KY6LA - Howard, Elmer

  • 3470 Posts
  • 1335 Reply Likes
Democracy in tech development tends to be a bad idea. Steve Jobs proved that time and time again. People will resist change and new ideas while consistently voting the status quo. Frankly until you show them that they need it, products like Predistortion, iPod, iPad iPhone will always be at the bottom of a democratic vote.

I seem to recall that DEC went out of business. Perhaps they listened too much to their user groups giving them more of the same rather than groundbreaking innovation.
Photo of Mickey N4MB

Mickey N4MB

  • 48 Posts
  • 10 Reply Likes
As a matter of reference, DEC did not go out of business. They were purchase by Compaq who later became HP. For better or worse, OpenVMS survives until this day and is still running at hundreds of sites around the world. 
Photo of Al / NN4ZZ

Al / NN4ZZ

  • 1678 Posts
  • 573 Reply Likes
A good approach is to have a visionary leader working with a diverse team.  It is rare that one person has all of the good ideas and approaches.  But a good leader understands that.   Over the years developing software and solutions,  I came to appreciate the benefits of team diversity.   We learned that including and listening to customers, sales, support, and marketing, as well as the IT folks provided insight we might otherwise overlook.   We didn't start out like this but found over time teams that didn't diversify didn't deliver products that reached their full potential.  

We rarely get revolutionary ideas from the users or the other non-IT contributors.  But sometimes it happens.  As others have noted, they tended to be evolutionary (tweak this, add that).   So I also understand the potential pitfalls and why it's important to have a leader that can pull it all together and innovate.

Some companies do better than others and smart ones improve over time.  I think FRS is on the right track.

Regards, Al / NN4ZZ  
al (at) nn4zz (dot) com




 
Photo of Lee

Lee, Elmer

  • 677 Posts
  • 285 Reply Likes
I think Gerald is pretty revolutionary in his thinking.

N8LP uses a rf bridge in his powder meter as opposed to a voltage bridge. This I think could be adapted to pre distorting the radio/amp as a unit. You would simply put the coupler as the last thing in the line prior to the antenna and feed DAC data into the program. It would work the same with or without the amp being turned on. If you developed a Ethernet connection to this DAC to get data into the radio it should be fast enough and you would not need to retrofit anything. http://www.telepostinc.com/lp100.html

When not predistorting you could use the coupler as a VNA for antenna analysis.

73 W9OY
Photo of Ned K1NJ

Ned K1NJ

  • 301 Posts
  • 75 Reply Likes
    Flex is revolutionary and they're good at filtering and responding to feedback from
the customer base.
    Using the N8LP sampler is a good idea.  Right on, Lee.

Ned,  K1NJ