rf impedance transformers

[tsp]

Hi,

I have a question regarding impedance transformers in rf circuits.
Why would you use lumped elements for impedance transformation? could
we not simple use transformers???

tspjor

 

[Udo Piechottka]

tsp schrieb:

Why would you use lumped elements for impedance transformation? could
we not simple use transformers???

The losses are lower when using lumped elements. Ordinary transformers can
only be used for low frequencies. Lumped elements change their
magnetization with fewer losses.

- Udo

 

[Tam/WB2TT]

"tsp" <tspjor@yahoo.com> wrote in message
news:8efbd3ef.0403241250.29529a89@posting.google.com...

Hi,

I have a question regarding impedance transformers in rf circuits.
Why would you use lumped elements for impedance transformation? could
we not simple use transformers???

tspjor
Transformers are commomly used. The most obvious example would be the 300:75

Ohm transformer used with TV antennas. A lumped LC circuit only works over a
narrow range of frequency but is easier to design (in my opinion) and can be
made adjustable.

Tam

 

[tsp]

Can you explain why the loss is lower for lumped elements? why can
transformers be only used for lower frequencies???

tspjor

 

[John Woodgate]

I read in sci.electronics.design that tsp <tspjor@yahoo.com> wrote (in
<8efbd3ef.0403241250.29529a89@posting.google.com&gt about 'rf impedance
transformers', on Wed, 24 Mar 2004:

I have a question regarding impedance transformers in rf circuits. Why
would you use lumped elements for impedance transformation? could we
not simple use transformers???

There may be many reasons, depending on what the circuit is intended to
do. But r.f. transformers CAN be used for impedance transformation, and
very often are. Usually, an r.f. transformer is arranged to form a
fairly narrow-band bandpass filter, so if that is not what is wanted, a
lumped-element impedance conversion may be preferred.

Sometimes a transformer may be used WITH a lumped-element converter. For
example, a double-tuned transformer feeding a common-base stage would
require either a secondary winding with very few turns or a tap at very
few turns from the earthy end. Instead, the secondary winding could be
tuned with, for example, a 100 pF capacitor in series with a 1 nF
capacitor, the output being taken from the junction, thus effectively
creating a 10% tap on a winding with only eight turns!
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Udo Piechottka]

tsp schrieb:

Can you explain why the loss is lower for lumped elements? why can
transformers be only used for lower frequencies???

Due to electrical conductivity the losses of iron cores are much higher
compared to non-contuctive ferrites.
(I think that transformers with ferrite cores are what you call "lumped
elements" / i'm no native english writer)

- Udo

 

[Jeroen]

tsp wrote:

I have a question regarding impedance transformers in rf circuits.
Why would you use lumped elements for impedance transformation? could
we not simple use transformers???

Either method works. Lumped-element transformers, in the form of
one or more cascaded L, T or Pi sections are usually restricted
to relatively narrow bandwidths. Say, an octave or so.

Classical flux-coupled transformers, using wire or coax wound on
ferrite or amorphous NiFe alloy cores as the magnetic material,
can be made for bandwidths up to three decades of frequency,
anywhere in the range from a few kHz up to several hundred MHz.

Transmission line transformers, often with no magnetic material
at all, work far into the GHz range.

Google for 'Ruthroff transformer' and 'Guanella transformer'.
That will certainly bring up enough to keep you busy.

Jeroen

 

[Frank Raffaeli]

tspjor@yahoo.com (tsp) wrote in message news:<8efbd3ef.0403241250.29529a89@posting.google.com>...

Hi,

I have a question regarding impedance transformers in rf circuits.
Why would you use lumped elements for impedance transformation? could
we not simple use transformers???

tspjor

RF transformers are useful into the 10s of Ghz. These are often
elements printed into the PCB. Transformers may have an advantage when
you can control the mutual coupling (K). With a transformer, you can
create a great bandpass filter with a flat amplitude response with a
relatively flat group delay, with excellent out-of-band rejection. For
example, a "transformer" operating at 2.44 GHz can pass the entire
2400-2483.5 Ghz ISM band with less than 0.5 dB passband ripple while
at the same time rejecting the PCS bands (1800-1990 MHz) by more than
75 dB. Unfortunately, this example has 6.5 dB of insertion loss if
implemented on low cost PCB laminate.

If you need a relatively lossless match, use "lumped" elements on the
smith chart. Find a freeware software tool such as Smith.exe -
available at http//:www.rfglobalnet.com/

Use the smith chart to visualize how the matching elements (L's, C's,
R's and transmission lines) operate on the path between the source and
the load. A wideband matching network can be very tricky. You will
probably need to make some compromises. If you can afford some loss
(R's in the mix), this usually helps a broadband match.

I didn't even touch on stability, but that's another thread ....

Frank Raffaeli
http://www.aomwireless.com/