Hi Everybody!

I'm looking for some help. Referring to a manual on the 8510, I learn that the Time Domain Low Pass can provide the complex impedance of a device under test (DUT) at the reflection points. The manual says that one can use the "Circuit Modeling Program" supplied with the Time Domain option (described at the end of the Time Domain Section). I cannot find that discussion - can someone point me to this? I am somewhat familiar with the time domain operation and can do the simple measurements.

My ultimate problem is the improvement of the match of a standard stripline launcher like an OmniSpectra surface launch jack part number 2066-1401-02 (for those of you that are familar with this problem). I'm having troubles around 6 GHz and am hopeful of using an 8510C to find the complex impedance and then develop a matching circuit in Ansoft Designer.

I may also get access to a more modern PSA or ESA analyzer - is the measurement significantly easier on those units than on the 8510???

Any thoughts?

Cheers

I'm looking for some help. Referring to a manual on the 8510, I learn that the Time Domain Low Pass can provide the complex impedance of a device under test (DUT) at the reflection points. The manual says that one can use the "Circuit Modeling Program" supplied with the Time Domain option (described at the end of the Time Domain Section). I cannot find that discussion - can someone point me to this? I am somewhat familiar with the time domain operation and can do the simple measurements.

My ultimate problem is the improvement of the match of a standard stripline launcher like an OmniSpectra surface launch jack part number 2066-1401-02 (for those of you that are familar with this problem). I'm having troubles around 6 GHz and am hopeful of using an 8510C to find the complex impedance and then develop a matching circuit in Ansoft Designer.

I may also get access to a more modern PSA or ESA analyzer - is the measurement significantly easier on those units than on the 8510???

Any thoughts?

Cheers

For connector modelling, I think it is best to look in low pass mode for peaks and valleys, but you need to look at very high frequencies (like 26 GHz) to find why responses at lower frequency are not good.

Newer analyzers, such as the PNA and ENA, provide nice advantages such as multiple traces, but the basic time domain response is almost the same. There is a newer program called "PLTS" that has some nice fixture de-embedding functions and might provide more capability than the native time domain response.

I used to give a demo to our internal engineers on using time domain to analyze the response of connetors to improve the response, but it takes an hour or so, so too much to put in the forum. Look at this application note:

http://cp.literature.agilent.com/litweb ... -5329E.pdf