Why Do Edge Coupled Microstrip Impedance Calculators Give Different Results for USB 2.0?

i need to calculate line widht and gap for USB 2.0 use, but all the calculators give significantly different results with same input. i also tried the web calculator eeweb has
http://www.eeweb.com/toolbox/edge-coupled-microstrip-impedance/

and tried to calculate with mathcad using formulas provided at
IPC-2141A (2004) “Design Guide for High-Speed Controlled Impedance Circuit Boards”

doing that i noticed that formulas brought out in the standard are different than ones on
http://www.eeweb.com/toolbox/edge-coupled-microstrip-impedance/

for example:
in Z0odd from formula by the standard, Q10 is under the sqrt with Ereff (resulting in Z0odd being a complex value) at the web calculator Q10 is outside the square root resulting in scalar value

in Z0surf formula by the standard there is sqrt(2pi) in the web calculator there is pi*sqrt(2)

in A0 formula there is no sqrt by the standard, there is one at the web calc

by the standard u = w / h in the web calc explenation this formula is left out(a minor problem)

i cant find the Z0even formula presented at web calc in the standard at all and i dont understand how replacing Q10 with Q4 in Z0odd formula suddenly calculates Z0even

eta0(free space wave impedance) is not defined in standard nor at the web calc, i use value 376.73031ohms from http://en.wikipedia.org/wiki/Impedance_of_free_space what value does the web calculator use?


even if i calculate in these deviations from standard i still dont get the same results as the web calculator and calculations by the standard are different from web calulator results as well

could someone point me towards reliable and proven method to calculate impedances for edge coupled differential microstrip?

To answer some of your questions:

The EEWeb calculator uses 377 for the impedance of free space.

Make sure you look at the errata for the IPC-2141A. Here is the "link":http://www.ipc.org/4.0_Knowledge/4.1_Standards/2141A_Errata.pdf

I believe that the committee that put IPC-2141A referenced the Transmission Line Design Handbook. I have attached a couple pages from the book on Edge Couple Microstrip Lines.

I hope this helps.

Thank you for your answer, i would never had figured to question accuracy of IPC standard myself(naive me), this clears up most of my questions and i now get much closer results to EEWeb calculator

input data:
w=0.3mm
s=0.2mm
h=0.18mm
t=0.018mm
eta0=377
Er=4.4

my calculations give:
Zodd = 49.739 ohm
Zeven = 72.606 ohm
Zdiff = 99.479 ohm
Zcommon = 36.303 ohm

EEWeb calculator gives:
Zodd = 51.6 ohm
Zeven = 54.3 ohm
Zdiff = 103 ohm
Zcommon = 27.2 ohm

so Zodd/Zdiff are quite close to the calculator result but Zeven is still quite off, i still cant figure out why i am not getting exactly the same results for Zodd, rounding errors in web calculator? i use MathCAD so i figure that might not have these problems? still the difference is small, close to negligible

also the IPC-2141A errata does not clear up the question about A0 formula, why does EEWeb page has the sqrt at last term? IPC does not specify that with errata or without, and neighter does Transmission Line Design Handbook
I will now try to calculate based solely on Transmission Line Design Handbook, and see what results i get there seem to be a bit more accurate formulas there, will post if i get any results

i also attached my calculation:

okey dont mind the Z0even formula, there is my mistake there i used Ereffo instead of Ereffe there

Hi
i have some further results, if i calculate Zodd by IPC standard and Zeven by Transmission Line Design Handbook except using Ereff formula from IPC standard i get Zeven exactly the same result as EEWeb calculator, and Zodd pretty close to it

same inputs as before
my calc
Zodd = 49.739 ohm
Zeven = 54.329 ohm
EEWeb calc
Zodd = 51.6 ohm
Zeven = 54.3 ohm

however if i replace Ereff formula from IPC standard with formula from Transmission Line Design Handbook i get results pretty similar to http://web.awrcorp.com/Usa/Products/Optional-Products/TX-Line/
while the Ereff value differs somewhat, 4.4 by Transmission Line Design Handbook(no change from initial data Er value), 3.294 by IPC standard
with Ereff at 4.4 i get
Zodd = 46.367 ohm
Zeven = 55.625 ohm
TxLINE 2003 gives:
Zodd = 45.2797 ohm
Zeven = 55.8782 ohm

the difference in Zodd is significant enough to make the difference in being within limits of USB 2.0 standard or being outside the limits

so the question is in the Ereff formula, is IPC standard correct or is Transmission Line Design Handbook correct?

My guess is that the IPC standard based their work off of the Transmission Line Design Handbook and that the equations in the handbook is what they intended.