Transmission Line Calculator Tool - Synthesized Line Length Too Long

[Matti Ukkola]

Hello!

Yesterday I was redoing some transmission line calculations in Keysight ADS Linecalc that I had previously done in the KiCad "Transmission Lines" calculator tool. I noticed a difference in the results during physical parameter synthesis.

All other results seemed fine (apart from minor deviation in decimal places), however the length seemed to be twice as long in KiCad (6.87 mm in Linecalc, 12.92 mm in KiCad). Screenshots of both tools are in the two attachments.

First possibility that came to my mind was a user error. I assume Ang_l in the KiCad Transmission Line tool meant the phase difference between the beginning of the line and the end of it. If this is not the case, the rest of this post may be disregarded.

I was going through the source code, but could not immediately spot an error in the source code, the formula inside void MICROSTRIP::calcSynthesize() seemed fine at first glance.

Something I found interesting is that the effective epsilon_r_eff  is almost the same in both tools: 3.53.

I used this to calculate the quarter wave-length of the transmission line:

lambda/4 = (c_0/f_r/eps_r_eff^0.5)/4

(3e8/5.8e9/3.53^0.5*1e3)/4
= 6.88

(3e8/5.8e9/1^0.5*1e3)/4
= 12.93

The upper value matches ADS Linecalc, the lower one matches KiCad Transmission Lines.

This may indicate that eps_r_eff is accidentally left to 1.

Is this a known issue, if it is one at all? If this is an issue, what are the required steps to get it fixed?

Best Regards,

Matti

[Seth Hillbrand]

Hello Matti-

Thank you for finding and investigating this issue.

It looks like this may have been fixed at some point in the v8 cycle and not yet cherry-picked to v7.  We'll make sure to integrate it for 7.0.10

Seth

Seth Hillbrand

Lead Developer

+1-530-302-5483‬

Long Beach, CA

www.kipro-pcb.com    in...@kipro-pcb.com

--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/1fad19df-3f4e-4de1-84c2-7c0d4e4aef27n%40kicad.org.

[Seth Hillbrand]

Oops.  Looking at the commit list (should have done that first), it appears that JP corrected it based on your message.

Thank you for helping to improve KiCad!

Seth

Seth Hillbrand

Lead Developer

+1-530-302-5483‬

Long Beach, CA

www.kipro-pcb.com    in...@kipro-pcb.com