working above the series resonance frequency of a cap....
cap. in 0603 SMD packages ,
100 PFarad ,
they used i think as a coupling capacitors... suggested by RF amp datasheet...
theyre just mony ...
everyone has a series parasitc inductance of 0.7 nh....
so eveyone has a SRF of 601 MHZ
im working in 1.3 GIGA ....
in that range the capacitors wd work as coils ? am i wrong ?
is that dangerous ? should i replace them ? or they gonna work just fine ?
and how can i order a better performance cap has less parastiic ?
and is there any suggested type of cap to work in that range ?
thanks in advance...
For bypassing it is the impedance that counts. You can use a lower value capacitor so that the series resonance is at your operating frequency. In the old days of valves/tubes high power types had special sockets with built in bypass capacitors that made the lead inductance be series resonant at the usual range of operating frequencies.
For signal filters you need to use a capacitor with less inductance. For coupling capacitors you can use a smaller value.
first im not sure if these capacitors are for coupling ...or not...
theyre suggested by the DATAsheet of the RF AMP...
ithink they use them to suppress the DC...
and they suggest different values for every frequency range....
Added after 1 minutes:
another comment to what u said...
u said " ... so that the series resonance is at your operating frequency ..."
if i choose the cap. to work in the SRF it would be just a short circuit ? ithink am i right or wrong ?then it will has no effect at all ?
You have to consider two cases. When the capacitor is used as decoupling capacitor, the best choice is to use a capacitor which has the SRF identical with the frequency you want to reject from the power supply lines. Any digital circuit will generate noise into the power lines, and this noise have nothing to do with the frequexcy of a certain digital signal. The thing that matters is the rise/fall times of that signal. If your noise have significant power around 1GHz for example, than the best choice is to decouple it with a cap with the SRF at 1GHz. This is because in that point, the impedance of the cap will be at its minimum, and equal with the ESR of the cap. Indeed, in that point you need an AC short circuit. In the second case, when the cap is a part of a filter/matching netwrork, then you must be carefull and use a capacitor with the SRF much higher than the working frequency. The cap will act as a coil if you work above its SRF. From your comment, I am almost sure that the cap is a part of an impedance matching network. There are components especially designed for microwave, with SRF up to many GHz, check for example www.avx.com .
/pisoiu
If the purpose of the capacitor is to be a low impedance then being at self resonance will make it a low impedance.
There is one trick you can use. Put N capacitors of value 1/N of the desired value in parallel. This will reduce the total inductance by N and increase the self resonant frequency of the group by the square root of N. An added bonus is that each of the smaller capacitors may have a lower series inductance than the larger value capacitor.