Microstrip transmission line teardrop-shape power divider question

  • Thread starter yefj
  • Start date
  • Tags Tags
    Network
10 replies · 3K views · last reply
yefj
Messages
166
Reaction score
2
Hello,I have built a microstrip line structure as shown below In photo and article 1.
my lines are er=9.4 h=0.433 line width W=0.015mm spacing between lines 0.065mm and 2 micron thickness.
The line impedance properties is 140 Ohm.
However I need also to build a feeding network for these lines.
I was suggested to use this nature article with the intrested shape.
Howeven i dont see the impedance power dividing logic in this tear drop shape.
I know its supposed to be a transformer of some sort,but there is a huge discontinuety in the width between W1(tear drop shape) and Wt(line width).
What is the impedance logic in this huge transition?
Thanks.
https://www.nature.com/articles/s41598-022-17212-0
https://www.nature.com/articles/s41598-022-17212-0/figures/2


Left: 3D drawing of a microstrip teardrop-shaped power divider on RT5880 substrate, with Port #0 feeding a tapered grey metal section that splits into Ports #1 to #4. Right: graph of amplitude in dB versus frequency from 26 to 30 GHz for S00, S10, S20, S30 and S40, with S00 dipping to about -30 dB near 28 GHz.

3D render and top-view schematic of a microstrip structure with four parallel lines, a central sample strip and a teardrop-shaped feed; labels read Length = 5650 um, Spacing = 65 um, Width = 15 um, and Microwave Magnetic Field Amplitude with B0 along one axis.
 
Attachments
  • 1.pdf
    1.pdf
    474.5 KB · Views: 87
Last edited:
That is a tapered transmission line impedance transformer.
Your input port #0 has an impedance of Z0
Your output ports #1,2,3,4 have equal impedances Z1,2,3,4
The output ports are connected in parallel to the wide end of the transformer.
The wide end of the transformer will have an impedance of Z1/4.
The width of the track, and droplet, at the input end, will be matched to Z0 over the ground plane.
The width of the taper at the wide end will be the same as a transmission line over ground of Z1/4.
Halfway along the taper, the width of the taper will be for an impedance, being the geometric mean of the input and output.
Z0% = Z0
Z100% = Z1/4
Z50% = √(Z0 * Z1/4) = GM(Z0, Z1/4); the geometric mean.
Z25% = GM(Z0%, Z50%)
Z75% = GM(Z50%, Z100%) ; and so it continues until you have the transformer.
The length of the taper is not critical, but must be over a quarter of a wavelength long at the longest wavelength.
 
Last edited:
Hello Balune core, my output impedance is 140 Ohm
The wide end of the tear drop is Z100%=140/4
few questions:

1.In the article section shown below they use quarter wave type transformer formula for Zt part below.
we have single transformer line between output and Z1 why did they use coeffient 4 in the formula?
In the regular formula its Zt=sqrt(z0z1)

3. what is the logic in putting this quarter wave transformer Lt(Zt) when we already have the tear drop as a transformer?
Thanks.
Layout of a microstrip teardrop-taper power divider with Port #0 input, a tapered metal section tapering to four output ports, and labelled dimensions L_t, W_0, D_t, W_1, W_t, L_1, L_tap and t_sub.

Printed journal page excerpt with a red hand-drawn underline under the formula Zt = sqrt(4Z0Z1) and a red curved mark in the right margin; text describes quarter-wave impedance transformers and a microstrip divider.
 
update :
my substrate er=9.4 h=0.43mm w=0.015mm gives each line output is 140 Ohm(I have 4 such lines with 0.065um spacing between lines).
the total tear drop wide side need to be 255um which is 63Ohm.
how do I calculate the transformer inpedance for such case?
Thanks.
 
If you design for a single frequency, or for very narrowband, then a quarter wave step transformer can be made from a transmission line, with a GM impedance.

What is the frequency, wavelength and bandwidth of the signal?

yefj said:
how do I calculate the transformer inpedance for such case?
Which transformer?
Maybe.
63R * 4 = 252R, each will become 140R, so Zgm = Sqrt( 252 * 140 ) = 188R.

Use R as the text form of omega.
Write the SI unit "ohm" with all lower case. The man was "Ohm", capitalised.
 
Hello Balune core, so if I have a situation as shown below W1 has impeadnce of Z1 and W2 has ipedance of Z2, then on the connection point between the two sections ,W2 line will feel an impedance of 4*Z1?
and W1 line will feel W2/4? correct?

if I understnad you correctly why does each W2 line will have 4*Z1 impedance connected to it from the left?

Diagram of one wide rectangle labelled W1 on the left feeding four narrow parallel strips labelled W2 on the right, with a red arrow pointing up at the junction.
 
I think you should analyse it as resistors in parallel, by looking at the voltage and the currents. Assume the transition node is matched. The voltage, v, at the transition between w1 and w2, is the same on both sides of the junction. A current of, 4i, that flows in w1 is divided into four lines of, i, in the four w2.
Z1 = v / (4 i ) ;
Z2 = v / i .
 
Hello Baluncore, In the article photo shown below they say that dufferent amplitues distribution gives us different field profile. Is there a way to understand from the field photo what amplitude distribution we are going to have?

another question is how did they know where to place the VIAs in order to controll the amplitude?

Thanks.
https://www.nature.com/articles/s41598-022-17212-0
Figure 5 with two rows of three panels: 3D renderings of a teardrop-shaped microstrip divider with ports 0 to 4, colour field maps with a 0 to 1 scale, and amplitude-versus-frequency plots of S00, S10 and S20 from 27.5 to 28.5 GHz.
 
yefj said:
another question is how did they know where to place the VIAs in order to controll the amplitude?
The power divider was first modelled without vias.
A fixed patch of vias were then added, and modelled, then machine learning was used to decide which vias should be kept.
 
Hello Baluncore , from this idea I have built a 64 gap coupled line resonator which is based on two teardrop feeding networks as shown below.
I got a very intresting resonance situation
5.56 one half wavelegth
6.09 three half wavelegth
6.53 five half wavelegth as shown below.
the five half wavelegth is a mode I want to enchance its coupling.
Is there some physicalnintution I could use to increase its coupling?
maybe some intution regarding line distance from each other etcc.?
Thanks.

CST simulation screenshot of a microstrip layout: a cyan ground-plane rectangle with two yellow teardrop-shaped feed flares meeting a black rectangular coupled-line region, with U, V, W axes, labelled points P1 and P2, and an 8.00 mm dimension.

Simulation screenshot of a row of interdigitated yellow microstrip fingers on a cyan background, with a tapered teardrop-shaped feed at the bottom left joining the fingers.

Line graph of S-parameter magnitude in dB versus frequency from 4 to 8 GHz, with four numbered markers: 1 at 5.556 GHz (-17.59 dB), 2 at 6.093 GHz (-10.07 dB), 3 at 6.635 GHz (-1.25 dB), 4 at 7.040 GHz (-0.27 dB).

Simulated H-field magnitude color map in A/m on a microstrip structure with a teardrop-shaped feed, peak near 147 A/m, with a legend from 0.00161 to 147 and a field-info panel listing 5.56 GHz.

Simulated h-field magnitude plot in A/m on a microstrip board showing a teardrop-shaped coupled-line resonator; colour scale 0.00161 to 147 A/m, with the bright field concentrated in the central cells and feed tips.

CST simulation screenshot of an h-field magnitude plot at 6.78 GHz on a blue plane, showing a teardrop-shaped microstrip structure with a colour scale in A/m from about 7.36e-05 to 220.
 
yefj said:
Is there some physical intuition I could use to increase its coupling?
Maybe some intuition regarding line distance from each other etc?
Thanks.
I have no idea how you will apply this two terminal resonator.
You only plot S22, which is the reflectance from the output.
How is port 1 connected?