For those interested in what's probably going to be the best auto tuner
ever - or just the challenges of designing one, here is the latest Alpha
email. Subscribing to Alpha's email has been a good experience for me.
I only get two or three emails a year, and they are always interesting...
73
Hal N4GG/4
-------- Original Message --------
Subject: DreamTuner Update including Youtube videos!
Date: Thu, 13 Mar 2014 11:29:20 -0400 (EDT)
From: Alpha Amplifiers <newsletter at rfconcepts.com>
Reply-To: newsletter at rfconcepts.com
To: n4gg at arrl.net
Hi, just a reminder that you're receiving this email because you have
expressed an interest in RF Concepts/Alpha Amplifiers. Don't forget to
add newsletter at rfconcepts.com to your address book so we'll be sure to
land in your inbox!
You may unsubscribe
<http://visitor.constantcontact.com/do?p=un&m=001n82xbzOcRnOgRgBtKXQJFA%3D%3D&ch=d1c9c710-461b-11e3-8fe6-d4ae5275b3f6&ca=221a0781-bbbc-4792-aace-b12cbeac6e81>
if you no longer wish to receive our emails.
600 w
*All things Alpha Amplifier*
*March, 2014
*
HOME PAGE GRAPHIC
DreamTuner 4040 Update - March 2014
The DreamTuner 4040 internal pictures
We keep testing and fixing the weakest link - and will until we find all
the weak links!
After Gordon and Ken took over the project, they've tested the heck out
of the core tuner.
They came back after a full review and had a list of changes they felt
needed to be made.
We're more interested in building the perfect Tuner than getting
something on the market before its time.
You can click on each pictures' link and see a bigger version with notes
on the photo.
Here's a picture of the original design. Although a great first try, we
found a few issues with the physical implementation. Some of the
elements couldn't handle the power, we weren't happy with the dynamic
range of the coupler, and the layout had too much stray inductance and
capacitance. Bad for 10 Meter operation.
Larger Picture of the Original Design
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpzwt0gGd6mBYcG24f603MKNem_sEqTTutugIF7HZKbFibz6FXx2uhsJm_d9NId5RILM7SxhckTBrNpI-1xYDUW6jMd4ggnYsF-w6UF_dR9g4me4rJWifu1PE0urnkoBA4S4vtid8ptm6URE52BR3XLmM4G0XRoEZAG_lPysZTO3NbhocFE4zJCWG4VOs57n5C88BjaFSAggMfEUlMmwtrNA==&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
This is the new layout. Notice we're now using a FPGA based coupler, and
have moved the capacitors back in the enclosure to facilitate a straight
run from inductor to capacitors.
Larger Picture of the new physical layout
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpSZRYTllS2D9SOtYctrUNaBc7Tl4fVj8JFHHGxFhmcNs-1hrLtU5Q1qc5zBv36jswv1Uz3iSMScIGL4virO3Vo1FZ88UmhEbzBuspHsHcP0X6UR-MLEkel_K80Qg-Y8iTvHO_pGq-qw0PrYi16gQZcTTgemFUcTuT_3KmkEsb77Vn1C0r8RkWk1htLMT8UY6ODscdQzL9aFU=&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
We've added a Vacuum fixed capacitor for the pad cap and a shorting
relay to electrically eliminate half of the inductor when not needed.
Larger Picture of the new Vacuum pad capacitor
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQplTJcwfhjCeslk5HRScCwPsSaO28ACBaHMlpumqVXrRdp9gqVqCy29pPYS_cPB20bTkyxLaDfvAzUk-ABxVNW34CAsrklNTtn61F8cK-gLEpCCR5g1H_Tl31Ouv33IRP54qeHwraica3k8HL0ghLidNNDUs6KVLW7-5TXJKsyrouBxjcbcXjy_QaQezqSv4eOUDPEL6xJaen148d82Z5nQe1noSYylsr1Jd6nqTuNVD8=&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
Here is a picture of the new coupler. Its FPGA based and does a
fantastic job of deriving R+jX in a consistent, accurate way.
Larger Picture of the new FPGA Coupler
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpkXmQ2cebwdm4UGueAmfp5VEtrIvKVvrsLveFFI8SvQV2lE14BMFiCLyYRckjrPrreZoHtTidalxfkmZp1WdixR0DTzrArCMqzHP8m3ppq97qtzgu_Zbj9bqr2EoN89Doy4aLr9ue5vhkgdnV92-8MhJanuoRBW2O0am3k9t94F6kLFHoMGdWb0pDagCzzi_dgSpqIMyGs6Qu-A6AY-GTywBaIbKeUkHY&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
No more internal Balun! We've removed the Balun and are planning on
offering an external Balun for balanced applications. This means as
many balanced vs unbalanced loads as required.
Larger Picture of the "now missing" Balun
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpwiQMYKCnNDywTRqhrVBWFs5-awcicLVIfi0h-zCjLqf0yRxWtdegeZIGt6QPhV_qRjM1wuK8umc9yYKehdGoCiGpwEd3GWdwh_usczcL3R4V5OqN1xqrphugNVad2brjXcIeaftZdM8dFEhrelV6_AFsOE5piaO90KyFKyJGW55fzGK1ejf33aShXQDD8315WGxmSpnUwng=&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
Here's a picture of the shorting mechanism for the edge wound inductor.
With it, we can remove half the inductance from the circuit.
Larger Picture of the straight run from in to out
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpCl-y34MdRaRhqsY2336ABDBRBFQPVOx-J3J0u7O7ULRqUQwhnmFfMhiPF9SHNvFCVGm1z2zcGxJWrb4fztYL4gpG-_6ATOXhEBAmOcLGPiCZfGumm1JwzZGnE3uE9ahDspBYTOJk1WRnaNTJ9kFQfhRcC2ZRsIAcyW_nIdwMel6qSqJ3FiXH-G_diyMS1QA-uZnegEoCnxv2lpiH432QqTUJgTXlk2Gk&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
A's sent while operating CW: .........
Standby. That's been an all too familiar reply to the forums. We've
been hard at work on our favorite project, but it's time to take a break
and share. This time we've got status, pictures, and a couple of
Youtube videos.
To say we underestimated the time it would take us to complete this
project would be a gross understatement. But, we've had our best and
brightest focused on getting the DreamTuner completed. Gordon, Ken,
Brad and Steve working full time on the project, and we've got four
programmers that work when we need them to step up. The tuner has
actually taken an interesting turn in the last few months. Originally
it was a good size electromechanical project with the design of the
inductor and switches. It then turned to a huge software and firmware
project where subsystems were built and tested, along with the
communications systems to be able to survive the high RF environment.
But the big surprise recently was, when we put the coals to it, we
found that it turned back to an electromechanical problem again.
4,000 watts is a lot of power, and just because some of these components
looked like they would handle the job, when we really tested them at
high SWR, they called uncle.
So, read about our exciting progress - and (some of) the sausage that
went into getting there - take a look at our videos and pictures - and
come and see the tuner in person at Dayton in just 2+ months time....
73,
Michael, AA6DY
Chairman, RF Concepts/Alpha Amplifiers
Our new internal layout. Notice two of the stepper controlled switches
and the teflon pad cap have been replaced by more conventional
components and the vacuum variable caps are moved back to facilitate a
straight run from input to output, through the 'T' network, of course.
Automatic Tuning on the 4040
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQpHqRFaZneAouAUiCVd0TAVjqr8fZARrtvASVXaIw7oMSRQvCVPiQ1zTotc_KG6ynChc4KgjzV1F6XBnEz8O0wW0v_6wNjXbO4wh_24kstFMV9mc_wIZATMA==&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
Automatic Tuning on the 4040
Our first of two youtube videos showing the DreamTuner in action. The
10 Meter band was the most difficult. It really pushed the elements to
the limit and pointed out the switching problems we had. And,
typically, 10 Meters has a sharp dip so data has to flow quickly,
consistently, and accurately so we don't miss the dip.
Tuner updates from the Alpha team
Just in case you thought the engineering staff at Alpha was off honing
their contesting skills or participating in a great DXpedition somewhere
- we'd like to update you - yes REALLY update you on our Alpha 4040
tuner project. Our tuner is getting close to shipping - sure you've
heard that before, but take a look at the details and the pictures below
and form you own opinion. We'll show you some of the guts that are
going into making this the best tuner ever offered to the amateur market.
It hasn't all been a walk in the park getting to this point - we have
stumbled a few times, and had to go back to engineering first principles
and test, test and test again. Right up front however, we want to admit
to our biggest and worst mistake in the whole design process - and it
was our first mistake too.
It wasn't the switches that you will see needed to be redesigned, or the
inductor that has had some changes or the boards that needed to be
turned again, or even the software.....
*/Our biggest mistake in the Alpha 4040 tuner process so far was to
underestimate the time and energy it would take to build our DreamTuner.
Our 4,000 watt spec proved much more difficult to deliver than we
originally thought. /*
*//*
But, even though we've taken a bit of a beating on the forums, we stand
by the Alpha way. We won't ship this project until we feel it's Alpha
quality and you'll be delighted. Once the DreamTuner leaves the factory,
we don't want to see it again! I'm sure everyone knows, but we didn't
take a dime in deposits - all we have is a list that has grown to
hundreds of future owners of the DreamTuner.
We are in the midst of an intensive testing phase. High power testing.
Here are some pictures of parts of our test setup.
Lab test setup
*Lab Bench Test Setup*
*Component Test Setup *
All of the switching components were tested previously using a mixture
of techniques at DC or 60Hz (hi-potting and over-current with DC) in an
attempt to make sure there was plenty of margin. Of course, as the
phrase on cars goes "your mileage may vary". It turns out (after
building some fairly elaborate test setups) that there were some
shortcomings when the components were tested at RF. These manifested
themselves either as outright failures or as problems that would have
resulted in reduced lifetime.
At the beginning of the process, when we found the enabling element -
the Vacuum Variables, we built a tester that would cycle the capacitor
to the minimum capacitance, HiPot test it at 5,000 volts, record the
leakage, run the capacitor to the maximum capacitance, and do the same
test. We tested multiple capacitors to failure. When we saw that we
could cycle these capacitors a minimum of 50,000 times without cracking
the vacuum bellows, we approved the capacitors. They're rugged and we've
been very happy with them for a long time.
Our edge wound silver plated copper inductor did well until we really
cranked the power. We found two problems with it - 1. The opposing
wipers were causing half a loop in the circuit, and we needed to find a
way to take much of the inductor out of the circuit when it wasn't
needed. Two problems found late in the game, but fixed.
Switches were a whole different animal. Here is an example of a voltage
breakdown failure in one of the beautiful switches we designed from
scratch. Although there appeared to be a lot of margin in the design,
something happened at 28 MHz when we applied the highest voltage that
could be expected. RF at 28Mhz and very high power looking into a big
mismatch.
We tested numerous commercially available relays, to see if any of them
might be useable. We "cooked" a
/lot/ of relays! Here is one that looked promising for a while, but
eventually had "corona" occur on the open contacts, that lead to heating
and ultimate failure. This relay is used in other company's tuner
products, and so we thought it would be worth testing.
We tested many more. One that is touted as a "40kW" antenna switch
emitted a strong odor of fish when the current through it was pushed
close to what we expected at 4kW, into our specified SWR. Parts of it
were getting so hot they were causing the dielectric parts to outgas.
40kW seemed fishy indeed...
Clearly there is more to this than meets the eye initially. So in
parallel to the hardware testing we developed a model of the tuner in
software so we could probe all the various things that were going on. It
turns out that, in a "Tee" tuner like this, that there are an infinite
number of ways to set the components to achieve 50 Ohms on the input.
Some of these result in low loss through the tuner- but very high
voltages or currents under some power/SWR/frequency combinations. This
can result in overheating, or in severe cases, arcing which can destroy
components. So sometimes it is better to run with slightly /higher /loss
in the tuner in order to keep the components within the limits they
would like to see for a nice long life. The difference here might be
something of the order of one-twentieth of an S-unit! But it could
double the life of the tuner.
*Testing high up on 10 m*
This "software tuner" has allowed us to look at a lot of things quicker
than we could in the lab. Here is an example of the component stresses
when the tuner is facing a 10:1 load at 28.5MHz and asked to pass 1.5kW
of power. Without going into too much detail, the x-axis at the bottom
goes from 0-360 degrees, representing one trip around the Smith Chart,
so all the possible impedances that lie on the 10:1 swr circle. The
important voltage stresses are shown on the left, for the two capacitors
and more interesting, the output connector- which in this case sees
close to 1kV rms- worth thinking about! The currents are shown on the
left- in this case the inductor current goes up to *17 amps!* Many of
these values can be adjusted by changing the algorithm used to derive
the component settings. This process is going on, and the difficult
areas are being subjected to further testing in the lab, as mentioned above.
The final piece is the software. I think it is safe to say that no tuner
ever made, commercial or amateur, has had such an ambitious computer
focus. The idea has always been to create a tuner that is in line with
all the modern transceivers whereby the functionality can be upgraded
over the internet. The 4040 is the first Alpha product to tackle this
head on in quite this way. Needless to say, it has provided us with
another "educational moment" in product development. Again, there is
probably too much detail to go into here. Suffice it to say, we have a
team of people working on various parts of this, with a plan to bring it
all together soon.
In case all the above sounds too negative, we have found solutions to
all the problems that we have encountered - there is no impassable
roadblock that is staring us in the face. Just the myriad of details
required to pull the tuner together into a product worthy of the ALPHA name.
Don't just take our word for it - take a look at some pictures of the
tuner in the lab being tested and the videos of it in operation. Here's
a youtube video showing the elements in action.
Elements turning!
<http://r20.rs6.net/tn.jsp?f=001RtxmBJowkMJ4-s_Ce_C8Yu7wEVSZsVEMf91H58nhbMPDlX3pAtBEzwXg22T0YxQp4iwKjF1MTSyHQzyBnVocI-BBokJz3lWHcivewneOdNDlxv7Kb2dFX-ffpmOBcQFvaCFSwO20TgiPB2z6HMwrMUELPFCodhMxuLta1DD3FDz_36b_vS-UYw==&c=3r9CRDnY4OarJf8G4UtcrwvlaV8ARAG951lm-_bGvFdC32_2yzy58A==&ch=y28-M8eZMBr4eTnjPFP7_Wo0RTxKecN55NN1mSaDauIQCAukdW7-nQ==>
Inside the 4040
When Alpha releases this product we are certain that it will be the most
thoroughly tested matching network ever released to the amateur market.
We feel this is essential based on the performance specifications we
will claim and the price of the product. This will be a tuner like no
other seen before.
We're very happy with the quality we've built in, and very unhappy about
the time it has taken to get this far. As anyone who's in the
engineering business, everything takes a long time. Especially when
you're building elements from scratch.
Well, that's it for now. Back to work.
Forward this email
<http://ui.constantcontact.com/sa/fwtf.jsp?llr=eka89lcab&m=1102056358212&ea=n4gg%40arrl.net&a=1116792808481>
<http://visitor.constantcontact.com/do?p=un&m=001n82xbzOcRnOgRgBtKXQJFA%3D%3D&ch=d1c9c710-461b-11e3-8fe6-d4ae5275b3f6&ca=221a0781-bbbc-4792-aace-b12cbeac6e81>
<http://www.constantcontact.com/index.jsp?cc=news05>
This email was sent to n4gg at arrl.net by newsletter at rfconcepts.com
<mailto:newsletter at rfconcepts.com> |
Update Profile/Email Address
<http://visitor.constantcontact.com/do?p=oo&m=001n82xbzOcRnOgRgBtKXQJFA%3D%3D&ch=d1c9c710-461b-11e3-8fe6-d4ae5275b3f6&ca=221a0781-bbbc-4792-aace-b12cbeac6e81>
| Instant removal with SafeUnsubscribe
<http://visitor.constantcontact.com/do?p=un&m=001n82xbzOcRnOgRgBtKXQJFA%3D%3D&ch=d1c9c710-461b-11e3-8fe6-d4ae5275b3f6&ca=221a0781-bbbc-4792-aace-b12cbeac6e81>^(TM)
| Privacy Policy <http://ui.constantcontact.com/roving/CCPrivacyPolicy.jsp>.
RF Concepts/Alpha Amplifiers| 634 S. Sunset| Longmont| CO| 80501
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.contesting.com/pipermail/secc/attachments/20140313/969fc9f5/attachment-0001.html>
|