Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Velocity factor of PCB stripline / antennas

Status
Not open for further replies.

lambdaanttech

Newbie level 6
Joined
Jan 12, 2011
Messages
14
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,283
Activity points
1,403
Hello All ,

I want to design a Franklin antenna from PCB material .
First I designed the antenna with a NEC based calculator to obtain impedance and radiation pattern .
Afterwards I scaled the model down from free space to FR4 substrate .

As far as I knew , the velocity factor is calculated as 1/sqrt εr .
The VF value was 0,466 .

After milling out the antenna I did some field strength test to compare the 2 element Franklin antenne versus a simple split dipole .
The split dipole had more gain then the Franklin ...
Measuring and calculating and remaking the antenna didn't solved my problems .

Finally I took 2 pieces of FR4 strip PCB to make a dipole and measured the resonance frequency on the VNA .
I found out that way that my calculated VF for FR4 was not correct .
Based on my measurements the VF factor of FR4 should be 0,84 - 0,86 .

Searching the net for some formulas how to calculate the VF factor from the dielectric constant didn't solved my problem .

Comparing my calculated/measured VF factor with the calculations with AppCAD confirmed that the VF factor was right , but I don't know how to calculate it .

Could someone help me about the formula to calculate the VF factor from dielectric constant for PCB's.

Thank you,
Kenny
 

RCinFLA

Advanced Member level 2
Joined
Aug 31, 2010
Messages
590
Helped
187
Reputation
378
Reaction score
186
Trophy points
1,323
Activity points
5,698
Measuring velocity factor should be done with a time domain reflectometer or with network analyzer. It is 1/sqrt(Er). There is sizable variation in Er on FR4 material between different manf. and lot to lot mix variation.

A dipole antenna (or resonator) will give you a very wrong result due to stray capacitance loading on dipole. The fields will not be confined to the board material. You will be measuring more air then FR4.

For a microstrip transmission line Vp, much of the fields are between the strip and ground plane within the substrate. Even more so for a stripline encapsulation.
 
Last edited:

lambdaanttech

Newbie level 6
Joined
Jan 12, 2011
Messages
14
Helped
2
Reputation
4
Reaction score
2
Trophy points
1,283
Activity points
1,403
I found that the 1/sqrtEr formula is meant for waves travelling trough a surrounding medium , like coax , where the dielectric surrounds the conductor completely .

However when recalculating with the VF given from my measurements I achieve resonant antennas .

I work with a PCB without a groundplane .
Even with a groundplane the stripline sees the FR4 material and air so its mixed .

Even when calculating with appcad a 1/4 lambda stripline calculated with the 1/sqrtEr formula is not right , you are getting another value .
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
47,995
Helped
14,149
Reputation
28,557
Reaction score
12,841
Trophy points
1,393
Location
Bochum, Germany
Activity points
278,361
I work with a PCB without a groundplane .
Even with a groundplane the stripline sees the FR4 material and air so its mixed .

Even when calculating with appcad a 1/4 lambda stripline calculated with the 1/sqrtEr formula is not right , you are getting another value .
Right. You get the correct propagation speed as v = 1/√LC (for lossless transmission lines). If C is completely dependant on Er, then the √Er factor applies.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top