Reducing the amount of Ferrite in Power Transformers

Welcome and feel free to ask questions about class E, mosfet's, and AM.
You must be registered to post.

Reducing the amount of Ferrite in Power Transformers

Postby WA2WVL » Wed Feb 01, 2006 12:22 pm

One of the expensive parts needed in Push/Pull Class E Amplifiers is the Ferrite cores. (Amidon FB-1020 @ $2 ea) A transformer running in the 50-100v range would need 8-16 cores at 160 meters.

The obvious answer is to use more than one turn on the primary.
The type of transformer which uses copper or brass tubeing for the primary is the best design since all of the flux from the inner wire must couple to the primary (leakage at the ends). What can be done is to extend the tubeing by one turn at each end thus making a 3 turn primary.
A 3 turn winding thru the hole forms a 1:1 transformer.
A 1:1 transformer ratio appears to be optimum for high power Class E amplifiers with the output network doing the stepup to 50 ohms. A benefit of this is smaller shunt C being required across the Fets.


How well does this work? The answer is very well as I will show.
A small 2 core assembly was built up and measured about 68 ohms @ 1800Khz. A single turn of # 12 teflon wire was added to each end of the tubeing and the primary impedance was measured to be 1200+j300.
A secondary of 3 turns was wound and terminated with a resistor that measured 8+j2. Now looking into the primary the impedance was 9+j5.
The reactance just becomes part of the output L which is tuned out.

Next the core loss was measured. The termination was removed and a matching network was built to match 1200+j300 to 50 ohms. Applying RF power to the matching network the following measurements were made:

Primary RF Volts (rms)------Loss/core-----------Xmtr pwr at 12.5 ohms (when used in an amp)

77.7v--------------------------2W-------------------483W

93.3v--------------------------3W-------------------697W

120v---------------------------5W-------------------1155W

This was with only 2 cores. Too be more conservative 4 could be used.

Floyd

PS: Core loss may actually be lower than measured ( 22uh used to match transformer impedance).
Last edited by WA2WVL on Thu Feb 02, 2006 6:00 pm, edited 4 times in total.
WA2WVL
Class E Enthusiast
 
Posts: 167
Joined: Thu Dec 22, 2005 1:40 pm
Location: Homosassa, Fl.

Postby NE4AM » Wed Feb 01, 2006 1:04 pm

For my p/p Class E rigs, I have been using Amidon FT-240-43 toroids, and winding them with 14g. Romex wire. These transformers have higher leakage inductance than those wound on short, fat cores, but as Floyd observes, this inductance is absorbed by the output network. The large diameter cores tend to radiate heat better, as well.
73 - Dave NE4AM
NE4AM
Member
 
Posts: 24
Joined: Fri Mar 11, 2005 1:24 pm
Location: Chatham, Illinois

ferrite cores

Postby N9NEO » Wed Feb 01, 2006 5:10 pm

Good discussion Floyd, Hi Dave,

I run push-pull rig like Frank. I get by ok using coax for primary and secondary. I use 4 cores and wind the coax thru the cores twice so flux cut in half. Braid is primary and inner is secondary. My output network is antenna tuner. I don't have to mess around with filter design and tuner can adjust swr and power level too. I can actually drive my modulator into current limit at just unmodulated carrier with the tuner. Nice clean signal too. It looks like a nice sinewave by eye - no spectrum analyzer yet.

I design some small chokes at work that they like to hang on the motor side of servo drives. Nobody knows exactly what these things are supposed to do. Servo folklore says they are necessary for 'controlability issues'. It seems the same exact cores I like to use for my radios show up on the servo drives. Heheh.

73
Bob
N9NEO
User avatar
N9NEO
Class E Enthusiast
 
Posts: 322
Joined: Tue Apr 08, 2003 11:52 am
Location: Uxbridge MA

Postby frank carcia » Thu Feb 02, 2006 7:51 pm

Bob,
The motor guys downstairs visited looking for a 300 amp inductor. I suggested they string some beads on their cables. They wanted to try the same trick. I showed them the core we all love to use. Hope they buy a crate of them.
A transformer wound with coax is a good way to get broadband operation.
Actually you rip the guts out and install a center conductor setting up the cable Z for the transformation. I bet that floyd has used that trick. I have also seen it used with metal tubing terminated to the shield in various spots.
I'm thankful Floyd shares his transformer information, good stuff.
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Re: Reducing the amount of Ferrite in Power Transformers

Postby VK3KRI » Sat Feb 04, 2006 7:32 am

WA2WVL wrote:One of the expensive parts needed in Push/Pull Class E Amplifiers is the Ferrite cores. (Amidon FB-1020 @ $2 ea) A transformer running in the 50-100v range would need 8-16 cores at 160 meters.

I like the way you think! $16 worth of cores is en exessive expense .

The obvious answer is to use more than one turn on the primary.
The type of transformer which uses copper or brass tubeing for the primary is the best design since all of the flux from the inner wire must couple to the primary (leakage at the ends). What can be done is to extend the tubeing by one turn at each end thus making a 3 turn primary.

I think this was mentioned in the classic motorola broadand transformer & combiner app note. I know I read it some where as I have used it with 2 extra 1/2 turns to to get a 2 turn primary. leakage inductance must be a bit wose but it got the B in the core down a bit.

Also with DC feed on a primary centre tap the current thru each 1/2 of the primary winding directly causes magenetic field in each core, rather than only one core in the case of a 1 turn primary. I'm not sure if this actually has an effect but I think its one of the reasons for using a seperate bifilar type DC feed for broadband amps.

These discussions are quite illuminating for those of us who don't quite know whet were doing !
Ian VK3KRI
VK3KRI
Member
 
Posts: 44
Joined: Fri Jan 31, 2003 8:32 pm
Location: Melbourne, Australia

Postby frank carcia » Sat Feb 04, 2006 9:03 pm

Ian,
My push pull class D driver has 2 cores and the 2 turn secondary connects directly to the 2 phases of the final gates with the center tap connected to the final sources. The load was so heavy on the driver (14 IRF 840s) that the waveform fell apart above 6 volts on the driver drains. I needed more power and voltage swing but couldn't easily add more cores. I changed the primary to 2 turns and doubled the vcc to 12 volts.
Cores still nice and cool with a clean signal. The driver only puts out about 30 watts. I had a bit of extra voltage peak so I just added 3-47 ohm 2 watt resistors per phase gate source on the final. This limits the peak voltage to 15 volts. A transorb is included on each phase just in case.
I saw a commerical amp with 2 primary turns once. It used brass tubes of two different diameters with teflon tubing between them. I used tubing for 1 turn and #14 teflon for the other.
I collect BB transformers and cores. I have a bunch of old Harris stuff
so have a good idea of how the smart guys do things.
A couple extra cores is cheap money to insure safe operation. fc
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby VE7KHz » Thu Feb 16, 2006 10:45 am

Hi Frank, Bob et al

I have purchased many years back, several spools of coaxial wire from the Boeing surplus store in Seattle. (what a great place to shop) This isn't actually "50 ohm" type coax, at least the mfr doesn't list it as such in their spec sheets. It's very expensive stuff (at least it was when Boeing bought it) silver plated stranded core, solid Teflon dielectric, then silver plated copper braid, then Teflon wrap for the outer jacket. Very nice indeed. I have a couple of different sizes and configurations and would be willing to send some "samples" to those of you who are actively doing design work for the collective group. I have spec sheets for a couple of them which give wire gauge, insulation thickness etc. I will have to actually find the wire again as it is still packed from our move a few years ago. IIRC, they core was 12 gauge in one of them, maybe 16ga in another.??

Anyway, this broadband xfmr designing sounds like lots of fun. Reminds me of Helge Granberg's work when he was designing for /\/\. What a great engineer.

Cheers

Paul
VE7KHz
I used to keep an open mind.......but my brain kept falling out.
User avatar
VE7KHz
Member
 
Posts: 23
Joined: Fri Jan 31, 2003 3:01 pm

Postby frank carcia » Thu Feb 16, 2006 10:51 am

Wow Paul,
Great find. That stuff is quite rare. I have a stash of the smaller stuff.
I have twisted pair shielded of various sizes also. It makes great mic cable too.
I built a lot of Helge projects but learned BB transformers from a German guy I worked for back in the '70s. fc
frank carcia
Senior Contributer
 
Posts: 765
Joined: Fri Jan 31, 2003 10:48 am
Location: Enfield Ct. / Niantic Ct.

Postby VE7KHz » Thu Feb 16, 2006 4:28 pm

After a few hours of digging through boxes in my storage shed, I have found all the wire I mentioned above. I am heading out to the shop to strip and take pictures....of the wire.

Let me know your thoughts, usable?? etc.

Here they are. From about 20 ga up to 12 ga all with braid except the 4 wire which is just Teflon coated silver plated and the big 12ga stranded one in the middle. I put in a piece of house wire, 14ga for comparison.

Image

And some mini rigid coax, 50 ohm and maybe some 25 ohm.

Image

Paul
VE7KHz
I used to keep an open mind.......but my brain kept falling out.
User avatar
VE7KHz
Member
 
Posts: 23
Joined: Fri Jan 31, 2003 3:01 pm


Return to Class E Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron