A 4-way power divider suggestion without resistor

[reza064]

i want to design an equal 4-way power divider without resistor that can handle high power up to 40w and small size for an amplifier.
i find some papers,but i want to all output ports be equal.
do you suggest anything that can satisfy my conditions?:grin:
please help me with details for simulation and drawing.:?::?::?:
thanks

 

[BigBoss]

I think "Double Wilkinson Divider" will serve to your requirements.You can design it by yourself on a appropriate PCB substrate.

 

[reza064]

yes i know it,but because of high temperature and burning possibility i want to use none resistors.
please help me in a way that need no transistor and handle high power and equal divide.
thanks.:?:

 

[chapt]

If you don't use resistors in Wilkinson divider, you don't have isolation between outputs. But VSWR and insertion loss does not change.

 

[DDavid]

1. u can use resistors for high temp for example:
www.ims-resistors.com/resistors.html
2. if u can't use with resistor because of your layout, than the only problem that u will have not isolation... IL, RL at all ports will be the same.

 

[volker_muehlhaus]

IL, RL at all ports will be the same.

Only the return loss measured from the input side is the same.
The return loss measured from the output side will change if you remove the resistors.

 

[DDavid]

Not correct
If u looking at simulators like: ADS, AWR etc, so u right
BUT in real live take and perform measurements (before and after removing the resistor) with network and u will see that RL at output side will be practically the same , with small degradation

 

[volker_muehlhaus]

BUT in real live take and perform measurements (before and after removing the resistor) with network and u will see that RL at output side will be practically the same , with small degradation

Interesting. Maybe your measurement ignores physics and math basics?

For the power divider that I developed and measured, the isolation resistors DID change BOTH isolation and output matching, as expected.

 

[DDavid]

Pls try it...

 

[volker_muehlhaus]

Pls try it...

As said: I designed and measured many power dividers.

The only explanation that I have for your measurement result is this: was your divider built on very lossy substrate like FR4? Then, because of high insertion loss, return loss for the no-resistor case isn't too bad.

 

[DDavid]

Believe me I do designed also a lot of this, mostly for amplifiers, dummy loads and so on. I used to work with : FR4; Rogers 6010, 5880, 4350; Taconic CER10, etc
First I saw this affect at R6010 (by mistake)...

 

[volker_muehlhaus]

It does not matter if I believe you. As for all internet forum posts, the reader should not trust any information given here, and better double check himself.

 

[reza064]

pay attention:: I talk about a new way to this. not wilkinson or ...
I found 2 papers that show new dividers with new shapes.
1- Design technique and performance assessment of new multiport multihole power divider suitable for M(H)MIC's
2- Multiport Power Divider-Combiner Circuits Using Circular-Sector-Shaped Planar Components.

first papers simulation with cst shows S-parameters difference in about 2dB of dividing.2-port by 2-port
but other one in a wide band divide to 4 port with max 1-dB difference.
when i divide with sectoraial divider,amplify with selected 4-transistors, and combine all with sectoraial divider i found S21 between 2dB-5dB amplifying but my transistors have a gain of 7dB.
please say why this happened? and was my work correct?
and help me to do it better.
thanks

- - - Updated - - -

**broken link removed****broken link removed**

- - - Updated - - -

**broken link removed****broken link removed**

 

[DDavid]

Can't open your attachment

 

[reza064]

what is your idea for these 2 papers for amplifier project?