Hello,

I have a circuit working at roughly 13MHz and have several IC's within the circuit...However I'm having trouble understanding DC Blocking and Bypass Capcitors and the values I should be using.

I notice a common value for a bypass cap is .1uF which turns out to be .1175 Ohms impedance at 13MHz and the lower I go in cap value the larger the impedance gets and vice versa. So in order to bypass the AC signals from the DC lines going to the IC should I choose the smallest impedance possible < 1 Ohm or does it not matter...Also with bypass capcitors if I know the frequency I'm running at do I need multiple caps to work/bypass at a wider range of frequencies just incase.

The same question goes for the DC blocking cap...should I be using a value cap that gives an impedance lower than 1 Ohm or does the high impedance not affect the ac signal going throught he capacitor and only the DC? Or does the impedance/value of the capacitor not matter because no matter the value of the cap...all capacitors block DC?

I was thinking of just using a 10nF or .1uF capactor for DC blocking and a 10nF and .1uF capacitor for the shunt bypass capcitors...would this be a safe bet or should the values be somewhat differet.

Thank you for all your help

__________________
1001110010101100100110110011101100101101

 

Keep in mind that at RF frequencies (you stated 13mhz) that the inductive value of caps can have a large impact on the effectivness for both a bypass and coupling cap. Not all caps of the same capacitance value will work effectivly as others when it comes to RF applications.

Lefty

__________________
Measurement changes behavior

 

As Lefty said the type of capacitor you're using is almost more important than it's value, generically higher values are better for coupling as a higher capacitance will have a lower impedance at any given frequency. For DC block you need to define exactly what you're going to block. ANY capacitor will block straight DC, but in the real world there is almost no such thing as true DC, all signals invariably have an AC component, primarily ripple from power supplies and load changes. Switch modes use a higher frequency than mains power upwards of 100-500khz so it's ripple will couple better through a capacitor rather than the 60hz from a wallwart supply which pretty much any capacitor will block. Bypass is just as tricky as you want something that will shunt as much of the undesired frequency as possible while keeping insertion loss for your desired pass signal as low as possible. It's really a complex subject, especially if you want sharp band cuts or narrow pass bands. Filter topology is like black voodoo to me =)

I know ceramics are generically used for lower values and higher frequencies, electrolytics are primarily power hold up and coupling capacitors for mid power audio, I think they have a high inductance and their impedance is very non-linear over a certain range. Don't know much about the internal structure of tants, other than they're voltage sensitive and are in between ceramic and electrolytic, usually being of a high value but with a small package size, I don't think they're particularly thermally stable though. Thermal stability can be a major determining factor in choosing a capacitor type. Different chemistries even in the same capacitor class can provide positive and negative temperature coefficients.

__________________

 

The desired value of the DC blocking cap in the signal path is determined by the load the capacitor sees at its output. The capacitor impedance acts like a voltage divider with its load.

For example, if the capacitor sees a 1k ohm load then you would usually want the capacitor impedance at the frequency of interest (in this case 13MHz) to be much less than 1k ohm, say 10 ohms or less which would require 1.2nF or larger. As noted, larger capacitors tend to have more inductance, so small ceramics (with short lead lengths) or, preferably, surface mount devices should be used at higher frequencies.