0402 capacitor self resonance
This information is part of the catalog data. You can back out the series inductance. Then you will have to include the traces as transmission lines and the via inductance.
see it at the data sheet
A historical note. Using series resoanace of bypass capacitors goes back many decades. Eimac power grid company sold sockets for their power valves/tubes with build in screen grid bypass capacitors of a size required to resonate the grid lead inductance at the higher end of the frequency operating range of the device. This insured a very good ground of that grid which is required for proper, non-oscillating operation of the device.
Another possible method to determine the self-resonant frequency of a cap is to make use of a network analyzer. If you sweep from a lower freq (say for example 50MHz) to a higher (than what you require e.g. 3GHz) you can easily determine where the self resonance occurs. Connect the cap in shunt as a 1 port network. Plotting on a Smith Chart (the Impedance Locus), you can see where the self resonance occurs. Where ever the locus crosses the 'resistance' line (i.e. the cap represents that it is purely resistive, this is where one of the self resonant frequencies occur). Remember, the capacitor is a theorectical series resonant circuit, so the impedance prsented by a series resonant circuit is purely resistive at resonance. As an example, a 120pf disc ceramic capacitor is self-resonant at 133MHz. Another piece of info to remember is that for every 1mm of lead length represents 1nH of inductance at 1GHz. I hope this helps.
A further follow-up to your question. You could instead use a software package such as ADS by Agilent. You can simulate a capacitor response (by sweeping the applied signal frequency, and checking the magnitude of the impedance). The response you will get should be identical to that if you were to use a network analyzer, just without having to use the physical capacitor.
With standard 0402 size capacitors you can approximate the model with a series inductance around 0.7nH. This is a raw approximation, but usually good for 1GHz range power decoupling.
Personnally I use 33pF for 900MHz, 10pF for 1900MHz and 6pF for 2.4GHz.
In order add some compensation for components leads, track/wire inductance, you might need lower value.
Murata Web site has a very good interactive tools to calculate the impedance and S parameters of their capacitors. I use it a lot.
Hi Brian39,
You can download from the Murata website an utility(called Chip S parameter and Impedance Library),
http://www.murata.com/designlib/mcsil.html
this is for their range of capacitors.
Pick up a capacitor, and visualise the impedance of the capacitor. You will see at the self resonant frequency (SRF) the minimum of the graph.
At this freq. the parasithic inductance became resonant with the capacitor. Above this freq. the capacitor act as an inductor. Be aware, what capacitor are u using for bypassing
From UHF range to 2.4 GHz I use several capacitors in paralel (22pF, 100pF,1nF and 1uF/tantal
Cheers, Al
Look at:
http://www.avx.com/SpiApps/default.asp
here you will find some interesting tools, such as programs that show impedance vs frequency charts of the capacitor you select.
Thanks !
- How the output power changes by varying the capacitor connected in series with gate
- DC blocking capacitors or interdigital capacitors at 24 GHz
- Why do company sites have inductors and capacitors in an s2p file format?
- DC block connector vs capacitor
- How to choose the right capacitor?
- What type of resistors, inductors, and capacitors to use for 2 GHz LNA design?
