Microstrip advice

[Mosaic]

Hi All:

I purchased one of these items..note attached PDF wit all details. Proceeded to break the PCB due to a yanked cable and was in the process of recalculating the 'impedance' of the microstrip for 1.6mm FR4.

I realized that it adds up to several hundred ohms.
My question is this...why would a 50 ohm RF signal require a several hundred ohm microstrip (to couple some signal to a detector) and then connect to another 50 ohm output? See Note 4 on page 11.

PS: i asked the author...he does not know.

 

[crutschow]

Those microstrip impedance values shown in Note 4 are transmission line impedances.
They don't add together.
Ideally they should be all the same but there is some mismatch between them (apparently not enough to cause a problem).

Do you understand what the characteristic impedance of a transmission line is?

 

[Mosaic]

I was treating them as lumped element, but that's a wrong approach.
I have to do some more learning in this regard.
I think the stepped impedance.is deliberate to create some kind of Low pass arrangement, otherwise just make a single 50 Ohm microstrip for lowest VSWR and be done.
I am working on figuring it out. I am going to do some empirical VNA tests as well, I have a new PCB from the PDF to work with . Sometimes learning works better by doing.

 

[JimB]

I think that this may be a case of "over thinking" the problem.

The highest frequency involved here is about 1000MHz, wavelength is 30cm.
What are the line lengths involved here, 1cm maximum?

I think that this is going to have little effect on the signal.
Did Tektronix even design these as microstrip or are they just traces on a piece of exotic circuit material?

JimB

 

[Mosaic]

I have been advised that it is a meander line inductor.....

 

[JimB]

Yes, the squiggle on the FR4 board is an inductor, and is there to replace a straight trace on the ceramic(?) original.

I am just questioning whether there was any impedance matching or filtering designed in to the Tektronix original.

JimB

 

[Mosaic]

It is matched to replace a squiggle on the original , both stepped impedance and phase delay. It's length is adjusted for the different propagation/permittivity of FR4 vs Alumina. The designer of the substitute said he just duplicated the functionality without finding out why it's there. As I am only 3 months into RF (IoT motivated) I have a lot to do in the fundamentals and so here I am.

I gather the following:
Since the peak Freq is 1 Ghz the microstrip is well below 1/4 wavelength resonance and can be considered an inductor and not a tx line. VSWR then is not as significant.
A series inductor is a choke/LP pass filter, presumably to knock down harmonic spurs.

What makes this complicated for me is figuring out what is the inductance of this design.
Can you help with that?

I found this interesting:
https://www.qsl.net/va3iul/Microstrip_Stripline_CPW_Design/Microstrip_Stripline_and_CPW_Design.pdf

Edit:
using coil32 freeware I calc'd the individual inductor steps and came up with 7nH total for the original TEk strip and 9nH for the substitute. Thus it seems from a lumped element perspective the substitute is somewhat incorrect. From a tx line perspective (which appears invalid at the Frq range) the impedances and phase are correct.

 

[KeepItSimpleStupid]

Could it be a unbalanced to balanced conversion and back again?

See section #3 PDP page 4. http://www.atmel.com/images/atmel-42131-rf-layout-with-microstrip_application-note_at02865.pdf

Balanced lines are better for long distances on a PCB.