The AM Forum
October 10, 2024, 01:39:33 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: The Polar Explorer - A High-Efficiency Transmitter  (Read 16896 times)
0 Members and 1 Guest are viewing this topic.
k1kp
Member

Offline Offline

Posts: 9


« on: April 17, 2019, 01:02:13 PM »

The Polar Explorer is a high-efficiency HF transmitter that produces 500W output. By increasing the power efficiency to 80-90%, The Polar Explorer seeks to reduce the cost, size, and weight of mid-power HF transmitters.

Originally conceived in partnership with Brian, K1LI, and with the ongoing help of Jeff, N1RD and Jeff, N1TK, the project  has reached critical mass. It's now time to test the waters and see what sort of interest there might be in others using and evaluating the design.

Our website at www.polex-tech.com  is now live to share the adventure!

Please take a minute to visit the site, read about the history and progress of the project, take a brief survey, and learn about opportunities to become a Beta tester.

We will be making a presentation at the upcoming NEAR-Fest XXV hamfest in Deerfield, New Hampshire on May 3-4. Visit the Near-Fest site near-fest.com for the exact time and location. We will also have a table where you can stop by and ask questions about this exciting new technology.

Then it's on to the  Dayton Hamvention on May 17-19. Be sure to stop by Booth 6810 and say hello, learn about The Polar Explorer, and find out about our Beta testing program.


73,

Tony, K1KP
Logged
KA0HCP
Contributing
Member
*
Offline Offline

Posts: 1185



« Reply #1 on: April 17, 2019, 02:00:22 PM »

Nice looking design. Good luck Tony!

p.s. Pssst, Reading ahead, it's a Class D transmitter.
Logged

New callsign KA0HCP, ex-KB4QAA.  Relocated to Kansas in April 2019.
VE3ELQ
Contributing
Member
*
Offline Offline

Posts: 258


« Reply #2 on: April 17, 2019, 03:27:17 PM »

" The class D power amplifier is driven by a phase-modulated signal, which is amplitude modulated by the drain supply in the amplifier."

Now thats very interesting.  Would love to see more details.  A block diagram would be helpful.  Big fan of class D here having built many for AM.  Good luck with it.

73s Nigel

Edit:
I found this article which explains much of the theory.  Very interesting.  Got my attention for sure.
https://www.arrl.org/files/file/QEX_Next_Issue/Mar-Apr2017/MBF.pdf
Logged

A "Tube Free" zone.
WD5JKO
Member

Offline Offline

Posts: 2000


WD5JKO


« Reply #3 on: April 17, 2019, 05:21:51 PM »

"Now thats very interesting.  Would love to see more details.  A block diagram would be helpful.  Big fan of class D here having built many for AM.  Good luck with it."




I attach one such amplifier, class AB for low level, and class F on peaks...

Jim
Wd5JKO

* Class_AB_F_Linear.pdf (2629.51 KB - downloaded 713 times.)
Logged
k1kp
Member

Offline Offline

Posts: 9


« Reply #4 on: April 21, 2019, 05:40:50 PM »

I added a few more slides to the slideshow.

-Tony, K1KP
Logged
W1ITT
Member

Offline Offline

Posts: 580


« Reply #5 on: April 21, 2019, 05:52:21 PM »

I've looked at the Polar web page and I don't see anything mentioned about transmitter IMD figures.  With the Pure Signal HPSDR in the Anan series, the new standard is minus 50 db or so for third order.  All the rest of the modern wonder-boxes still seem to hang around minus 30-ish db, as they have for decades.  The world doesn't need any more dirty 30 db stuff on the bands.  I'm hoping that the Polar Explorer can meet the new state of the art.
Logged
k1kp
Member

Offline Offline

Posts: 9


« Reply #6 on: April 22, 2019, 07:11:13 AM »

David-

I'm sure you will be happy to hear that The Polar Explorer has had clean output a goal since the very beginning. As described in the 2017 QEX article, output phase and gain are measured as a function of signal level and used to precorrect the signals to the class D PA. Given that the measured signals are in the same format as the drive signals (amplitude and phase), it is a simple matter to compute and apply these corrections. While this aspect of the Polar Explorer is still very much a 'works in progress' (as it is with many less efficient linear amplifiers currently being sold), it continues to be a design goal to achieve the best possible IMD performance. Lab measurements already indicate that it will be significantly cleaner than currently available amplifiers.

73,

-Tony, K1KP
Logged
k1kp
Member

Offline Offline

Posts: 9


« Reply #7 on: May 14, 2019, 07:32:58 AM »

Nearfest was a great success! There was lot of interest and excitement in the Polar Explorer and several prospective Beta Testers signed up.

Now it's 'On to Dayton' - if you're making the trip, be sure to visit the Polar Explorer Booth, #6810!

-Tony, K1KP
Logged
k1kp
Member

Offline Offline

Posts: 9


« Reply #8 on: May 20, 2019, 01:48:53 PM »

The Polar Explorer was very well received at the Dayton Hamvention! The booth was packed on Friday and Saturday with folks learning about the new technology transmitter. It's gratifying to see how quickly the average ham 'get's it', recognizing the real practical advantages that a high efficiency transmitter can offer in size, weight, and cost. Over 500 business cards were handed out, and several more hams signed up for the Beta testing program.

Two videos about the Polar Explorer have been posted to Youtube. The first one is from a live Dayton Hamvention 2019 video chat with W5KUP.COM:

https://youtu.be/kLuAApJMIZo (fast forward to 20:50 for the beginning)

The second one is from the Nearfest, NH flea market:

https://youtu.be/R3SQq-UrzUE

Tony, K1KP



* Booth cropped.jpg (716.83 KB, 2861x1858 - viewed 895 times.)
Logged
KA0HCP
Contributing
Member
*
Offline Offline

Posts: 1185



« Reply #9 on: May 20, 2019, 03:47:44 PM »

Tony,
Can you give us a ballpark cost for materials?  Thanks, Bill.

p.s. Nice interview with Burt.
Logged

New callsign KA0HCP, ex-KB4QAA.  Relocated to Kansas in April 2019.
W1ITT
Member

Offline Offline

Posts: 580


« Reply #10 on: May 20, 2019, 04:57:22 PM »

Do we have 3rd and 5th order transmitter IMD numbers yet?  Years ago I worked for a boss who told me that "really good" wasn't a valid measurement.  He said that if I couldn't show measured data it wasn't engineering, but just an opinion.  If this implementation can show high efficiency as well as IMD that meets that of HPSDR  (Anan, etc ) running Pure Signal, it'll be a real game changer.
Logged
k1kp
Member

Offline Offline

Posts: 9


« Reply #11 on: May 20, 2019, 05:15:32 PM »

Materials cost for a Beta unit would be around $1000 in single quantities. If there are 10 or more beta testers, then obviously the parts cost can be reduced.

As for TX IMD, it is still a works in progress. The Polar Explorer achieves better than 40db on IMD3/5, as per standard ARRL testing methodology, on 160 and 80 meters, without using any precorrection at all. It naturally degrades with frequency, but by applying a static (one-time per specific load) measurement of output amplitude and phase, the numbers improve and approach 40 db on all bands. In the future, more development may lead to improvement of these figures using the static correction, or a dynamic correction approach may also lead to further improvement. 40db may not beat what Anan can achieve, but it certainly beats what any non-corrected 500W (amateur) amplifier currently can achieve.

-Tony, K1KP
Logged
W1ITT
Member

Offline Offline

Posts: 580


« Reply #12 on: May 20, 2019, 08:49:49 PM »

We could live with 40 db third order.  The old sweep tube stuff did well to make it into the low 20s and most of the modern twelve volt finals do well to make 30 or 33 db third order.  And that's when the DXpedition excitement doesn't push the knobs a little further to the right.  Original S-Lines could be made to get 36 db or so, but it was mostly downhill after those days.  So if we could get more people at 40 db IMD numbers the world would be a better place.  But I'll have to take my vitamins to live long enough to see people phase out the current crop of radios.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8886


"Let's go kayaking, Tommy!" - Yaz


« Reply #13 on: May 20, 2019, 10:29:24 PM »

Hi Tony,

Good show!

Yes, -40 DB 3rd would be a huge improvement over the typical -28 DB or worse we often hear on the bands.  However, a ricebox driver if often the weak link of the RF chain. Even a super clean amplifier is ruined by a poor driver. IE, amplifier cleanliness can never be any better than the driver itself.

It is good that you are doing these above-average IMD figures with a built-in driver to eliminate this problem.

I spent a lot of time building an old school legal limit class A and AB1 tube RF chain that eventually did about -55 DB 3rd IMD, but it took a huge effort and lots of heat, so I can appreciate your accomplishment.

Over the years I have often asked RF people smarter than I, why haven't they come up with a ham linear amplifier that runs class D that used an SDR digital modulation (power supply) system to achieve 90% efficiency?  I was told it was coming, but that was 20 years ago.

So here we are and I'm glad you are making such great progress. I'll keep an eye on how it works out.
In addition to the very successful Pure Signal technology, this is a great step to on-air cleanliness.

Why did it take so long? I would think one of the big guns like Yaesu, Anan, etc would have done something by now. Was it software or hardware or was it a cleanliness design issue?  Hopefully you can stay ahead of the competition curve.

BTW, I visited your website with the question, "How do they generate 80-90% eff ? Using what basic technique?"  The text below is taken from your site at Polex-Tech.com and will answer this question for others:

Good luck -

Tom, K1JJ

--------------------

"How does it achieve High Efficiency?

The Power Amplifier stage of traditional RF amplifiers must be biased in the linear region to avoid creating distortion and harmonics. This constrains the efficiency to be around 55-60%, which means that for every watt of RF output, an additional .8 watts of power is wasted. This wasted power contributes to power supply costs, cooling requirements, and ultimately, cost, weight, and size.

By using a class D power amplifier, much higher efficiency can be achieved, between 80 to 90%. A class D amplifier is not linear, so instead of attempting to linearly amplify to SSB signal, it is actually created in the RF Power Amplifier stage. The class D power amplifier is driven by a phase-modulated signal, which is amplitude modulated by the drain supply in the amplifier."


Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
k1kp
Member

Offline Offline

Posts: 9


« Reply #14 on: May 21, 2019, 07:48:28 AM »

Tom-

Thanks for the kind words. The answers to 'why did it take so long' are found in the technical challenges faced in the 6 years of development of the project. Specifically:

1. Getting decent efficiency at 10 meters - this required Gallium Nitride FETs. I spent one entire summer trying every FET out there, and it wasn't until I tried GaN that the efficiency came up at the high frequency end. The critical parameter is switching time, which simulators do not accurately model. The really cool thing about GaN (for an OT like me) is that the carriers are majority carriers (electrons). They call it a 2DEG (2D electron gas) and it's acutally a sheet of electrons - like a good old vacuum tube!

2. Bandswitching a Class D amplifier - Choosing the right tank impedance and Q while managing parasitics was key. To cover 9 HF bands, a brute force approach would have 9 separate tank circuits (LC) and 9 relays. This obviously was a non-starter, so a different approach was needed. A design using only 4 relays was found and made it possible to bandswitch the amplifier for 160-10m. Six meters is still out of sight in a bandswitched arrangement, although I think 6 could be done in a single-band implementation, at somewhat lower efficiency.

3. DSP - DSP with sufficient performance has existed for quite a while, but it is the combination of DSP and the above two factors that really makes Leonard Kahn's 1952 idea become feasible.

-Tony, K1KP
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8886


"Let's go kayaking, Tommy!" - Yaz


« Reply #15 on: May 21, 2019, 09:54:07 PM »

Thanks for the info, Tony -

Sounds like a long journey to finally get the amplifier working reasonably well.

Suggestions: I would think that an additional line of class D linear amplifiers only, say, 500 watts, 1.5 KW, 3KW and 5KW pep out (export models like Henry and Alpha does)  would be an interesting product grouping.   Since the exisiting design already requires a receiver - and excellent transceivers are already commonplace, a stand-alone amplifier could be desirable. Kit form would be one way to market them too.  I know I would buy one.

The class D linear amplifier niche is where the need currently lies.

I can imagine an Anan transceiver or HPSDR rig (I own an HPSDR) that would use its PureSignal pre-distortion to lift the amplifier into the -55 3rd IMD world. The 5KW pep version would run 1KW carrier on AM with plenty of audio headroom. It would be as efficient and have the power of a 24 FET class E rig.  Or compete with a 4-1000A plate modulated rig all in a small package. Now that would be something. 160-10M with a click, little heat at QRO, all modes... amazing.

A product line like this in the marketplace right now would put competitors out of business and you would take over the world of linear sales...    Who would want a 30-60% eff tube heat-box when they could get a 80-90% dream machine?   Cheesy

I heard a big ssb station on the other day using PureSignal. He was 50 over but had zero signal on the opposite sideband and after 2.8 KHz, he was absolutely gone. I mean this was a perfect signal. Years from now, assuming anyone cares about splatter due to all the open space on the bands, running a "dirty"  -28 DB 3rd order KW transmitter will be shunned like the end of the spark gap days....  Wink

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.057 seconds with 18 queries.