??? 12/05/04 16:18 Read: times Msg Score: +1 +1 Informative |
#82560 - Highcap ceramics against paralleling Responding to: ???'s previous message |
Joseph said:
For really broadband protection, it is a good idea to use two or three caps in parallel, each performing best in a different range. Unfortunately, this paralleling of decoupling capacitors is very likely to result in unwanted resonances, which can cause relevant impedance peaks: The equivalent series inductance of a larger capacitor reacts with the equivalent series capacitance of a smaller capacitor. The resonance is the more sharp the more the equivalent series inductance of involved capacitors differ and the lower the equivalent series resistance of the larger capacitor is. Unsuited is, for instance, a combination of big fat tantal (10µF, leaded) in combination with tiny SMD ceramic capacitor (10nF, X7R, 0603). A better combination, which prevents resonance is 10µF/16V electrolytic in combination with 100nF, X7R, 0805. Here the rather high equivalent series resistance of electrolytic very effectively dampens any resonance. Unfortunately, at the same moment this rather high series resistance of 10µF/16V electrolytic degrades the effect of decoupling. So, this combination might be good enough for an OPamp circuit, but for a microcontroller circuit something better is needed, which shows very low impedance at higher and at lower frequencies, free of impedance peaks. Since some years highcap ceramics are available, which offer an equivalent series resistance of about 0.01Ohm in a very broad range of frequencies. I would recommend the use of 4.7µF/10V/X5R in 0805 case or 1µF/10V/X5R in 0603 case. By this you have both, the much larger capacitance compared to 10...100nF ceramics and the very low equivalent series inductance of SMD case. And this combined with an equivalent series resistance of 0.01Ohm and the absence of any impedance peak. These highcap ceramics provide an outstanding decoupling performance! Kai |
Topic | Author | Date |
EMI/RFI noise | 01/01/70 00:00 | |
Keep looking | 01/01/70 00:00 | |
Brownout? | 01/01/70 00:00 | |
It is Brownout | 01/01/70 00:00 | |
brownout protection | 01/01/70 00:00 | |
EMI/ RFI solutions. | 01/01/70 00:00 | |
Archives | 01/01/70 00:00 | |
0.1uF | 01/01/70 00:00 | |
Caps | 01/01/70 00:00 | |
Highcap ceramics against paralleling | 01/01/70 00:00 | |
Highcap Ceramic. | 01/01/70 00:00 | |
Thanks!![]() | 01/01/70 00:00 | |
.01uf | 01/01/70 00:00 | |
decoupling | 01/01/70 00:00 | |
EMI problem solution. | 01/01/70 00:00 | |
is it ? | 01/01/70 00:00 | |
That's all | 01/01/70 00:00 | |
to Alberto | 01/01/70 00:00 | |
To Erik | 01/01/70 00:00 | |
Electrically Noisy Environments | 01/01/70 00:00 | |
brownout - no, everything else - maybe | 01/01/70 00:00 | |
Problem solved | 01/01/70 00:00 |