Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Getting the most out of the EB104 design - It can easily be done  (Read 9458 times)
K9AQ
Member

Posts: 54




Ignore
« on: June 12, 2012, 02:33:07 PM »

I started a forum in February looking for ways to improve the EB104 design.  Thanks to a lot of help from Baruch, 4z4RB, and from Ken, K9FV, I believe that we have identified the modifications that need to be made to this design to increase the output power on the higher bands and to improve the reliability.

There have been many criticisms posted on this website of the board design and the power matching of the output transformer, but I have not seen any directions on fixing these issues.

Baruch has build numerous EB104 amplifiers and later amplifiers using his own board design.  With his amplifiers, he found that replacing the the conventional output transformer, T3, with a transmission line transformer (TLT), greatly improved the performance.  We expected to see the same improvement when a TLT was used with EB104. Baruch gave me a lot of encouragement and Ken and I tested many different types of TLTs and compared the matching performance against the T3 transformer supplied by CCI.  There is just no comparison.  An easy to build TLT provides a near perfect match from 160-10 meters.

I recently overhauled my EB104 amplifer and replaced T3 with a 9:1 TLT, using a Fair-Rite286100202 #61 multi-aperature core, wound with 3 turns of 18 ohm coax (available at CCI).  At the same time I made several other needed mods, better decoupling capacitors, 12 turn bias pots, and improved RF grounding for the PCB.

I am very happy to report that I have an easy 600 watts output through 12 meters and about 550 on 10 meters! I am getting very consistent gain on all of the bands. 

I will be documenting the work on my amplifier and will make copies available to help out others that would be interested in trying out EB104 with a TLT.  If you would like a copy, drop me a note. If it works as well for you as it did for me, thank Baruch for pointing me in this direction and for the hours of support that he gave me.

Don Solberg
K9AQ
Logged
K9FV
Member

Posts: 478




Ignore
« Reply #1 on: June 12, 2012, 08:08:27 PM »

Don and Baruch did most of the "smart" work - I just slogged along and followed their lead on so many things.  I did learn a LOTS and Don has a very good write up on the mods required for the EB104 board.  BUT, as Don mentions, Baruch has a much better design board he is selling on ebay and a very decent price lacking only the MOSFETs - with the SD2933 being the device of choice.

Thank you Don and Baruch for all the work ya'll have done on this project.

73 de Ken H>
Logged
N4EDI
Member

Posts: 12




Ignore
« Reply #2 on: June 20, 2012, 09:31:33 AM »

I would like a copy of your modifications. How can I get them?

Thanks
Ed
KJ4OAS
Logged
K9AQ
Member

Posts: 54




Ignore
« Reply #3 on: June 22, 2012, 08:45:20 AM »

Ed,

I am putting all of my mods, along with photos in to a document.  I should have this done in a week or so.  Send me an email at dsolberg@wi.rr.com and I will email it to you.

Don
K9AQ
Logged
4Z4RB
Member

Posts: 5




Ignore
« Reply #4 on: July 10, 2012, 06:57:22 AM »

Great to see al that activity.
Indeed the TLT is the answer and best performance
In my new board I have also used a dual feedback system not only from the common mode trafo but from the drains too
BTW that original choke is way too small and does not have enough turns to provide good feedback that is why the feedback resistors get too warm and burn up.
thanks
Baruch 4z4rb
Logged
K9AQ
Member

Posts: 54




Ignore
« Reply #5 on: July 10, 2012, 07:10:19 PM »

Baruch,

Thanks for your feedback. I am using 10 watt caddock resistors, mounted to the copper heatsink for the feedback resistors.  I have not experienced any overheating.  I will measure the temperature of these resistors after I have been on the air for awhile.  I only work SSB and I listen a lot more than I talk, so I probably aren't pushing it very hard.

I have had a lot of requests for the article and I am glad to see there is still a lot of interest in building solid state amplifiers.

There have been a couple of hams who sounded serious about implementing TLT's in their EB104 amplifiers and I will like to see some more experimentation with this.

This posting has had a lot of people reading it, but very little feedback.

73 de Don, K9AQ
Logged
M0HCN
Member

Posts: 473




Ignore
« Reply #6 on: July 11, 2012, 03:43:52 PM »

There is that design in the latest ARRL book using source degeneration that looks kind of interesting, maybe a variant using SD2933 or something would be worth trying.

The other interesting feature of that design is the use of the DC injection bifilar as a 4 port hybrid to direct most of the even harmonics into a dump resistor, and the winding of the two TLTs on a single pigs nose.

The article claims 250W from a pair of VRF151 with IMD3 around -40dbC,  and unfiltered second very nearly good enough even without any filtering (third is another matter of course). 

I do wonder if there is enough inductance on top band to get the TLTs to work well, guess I will have to build one and see.

73 Dan.
Logged
K9AQ
Member

Posts: 54




Ignore
« Reply #7 on: July 11, 2012, 06:29:33 PM »

Dan,

I tested the TLT on 160 meters on a test jig and even with #61 material cores, I was able to get a flat match from 160-10 meters.  I didn't have any #43 material but it might be better.

I have studied the amplifier design in the ARRL handbook, and I am using the same TLT core as that article.  I am also building the control circuit from that amplifier. 

If you look at the schematic for the Elecraft KPA500, which uses SD2933;s, it is very similiar to the ARRL amplifier.

73 de K9AQ, Don
Logged
NX7U
Member

Posts: 56


WWW

Ignore
« Reply #8 on: July 13, 2012, 06:41:32 AM »

Shooting from the hip...
Do you think that this design modification would apply to BJT-based amplifiers as well?
Logged

Scott NX7U in DM43
K9AQ
Member

Posts: 54




Ignore
« Reply #9 on: July 13, 2012, 12:27:29 PM »

this really will be from the hip, as I have no experience with bipolar transistor high power amplifiers.  The output impedance of a solid state amplifier is primarily determined by the supply voltage and the output power. Since the supply voltage is usually much lower, ie. 12 volts instead of 50 volts, the output impedance is much lower and would require a transformer ratio from 16:1 to 25:1 depending on the output power and supply voltage.  This would be much more difficult to achieve with a transmission line transformer.

You are probably better off staying with a conventional tube and sleeve transformer.  A good explanation of the output impednance calculation is in Motorola's AN762 application note.
K9AQ
Don
Logged
4Z4RB
Member

Posts: 5




Ignore
« Reply #10 on: August 05, 2012, 08:46:35 AM »

using  a TLT for BJT is common practice, But at 12V it is rather complicated when compared to sleeve and tube type of trafo
Logged
K9FV
Member

Posts: 478




Ignore
« Reply #11 on: August 07, 2012, 09:28:24 AM »

using  a TLT for BJT is common practice, But at 12V it is rather complicated when compared to sleeve and tube type of trafo

You know - I'll bet the BJT is why so many designs are using tube 'n sleeve for MOSFET - getting the start in BJT amps.  Baruch is one of the "real experts" in MOSFET design with all the work he's done with them.

Interesting point on TLT - I've been testing a 4:1 TLT for T1 input RF transformer this morning. Looking at the phase shift with a dual channel O'scope.  On the mid ranges (10 to 30mhz range) the phase shift shows to be pretty good at 180ยบ.  At 50mhz chnl 1 (Q1 side) will lead chnl 2 (Q2 side) by a small amount, but at 160 meters, chnl 2 will lead chnl 1 side.  Does anyone find that nsame results in their tests?  I'm using a "2 holer" PCB (based on KPA500 and ARRL 250 watt schematics) with a 10 ohm resistor to ground at each gate.

73 de Ken H>
Logged
4Z4RB
Member

Posts: 5




Ignore
« Reply #12 on: August 09, 2012, 10:52:05 PM »

what is the length of the line and what is the velocity factor of that wire?
what is the lambda/8 design frequency?
these factors are basic
Logged
K5JZS
Member

Posts: 9




Ignore
« Reply #13 on: October 20, 2012, 02:07:51 PM »

good reading on this and related posts.

do want to say that the Heatsink in my "kit" is not flat across the surface, enough so to require machining.  Just something to be aware of.
Logged
K9AQ
Member

Posts: 54




Ignore
« Reply #14 on: October 20, 2012, 03:47:22 PM »

My heatsink did not come from CCI but it also needed machining.  I had a local machine shop mill both the heatsink and the copper heat spreader.  Let me know how your EB104 amp works out.  I use mine every day and just love it.

73 de K9AQ, Don
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!