Designer’s Guide Community

Fred New Member Offline Posts: 6	Caps Feb 19^th, 2004, 1:07am Hi, Can anyone please let me know what's the difference between bypass cap and coupling cap? thanks,
Back to top	IP Logged

Cao-Thong Tu

Guest

Re: Caps
Reply #1 - Feb 19^th, 2004, 5:11am

Fred,

I think that the word coupling cap is more for AC
coupling of signals. For example, if you have a
resistive input impedance of 100kOhms, using a
capacitor will make an AC coupling and we might therefore use the terminology of coupling cap.

A bypass cap is a cap you typically use to shunt
sources and avoid AC current flowing into a particular
path. For example, imagine a NPN with resistor in
the emitter and a load resistor at the collector. At
high frequency, you can add a shunt capacitor in parallel with the emitter resistor to increase the gain.
Such a cap, deviates the current into the shunt capacitor instead of flowing into the resistor, and this
is called "bypassing".

Regards,

Cao-Thong

Back to top

IP Logged

Mighty Mouse

Community Member

Offline

Here I come to save
the day!

Posts: 75
Fantasyland

Re: Caps
Reply #2 - Feb 19^th, 2004, 10:13am

Fred,
It is common to confuse decoupling capacitors (or d-caps) with bypass capacitors. In both cases the capacitors being referred to are the capacitors that shunt the power supply and are used to quiet the supplies. However, the term "decoupling capacitor" is wrong; the correct term is "bypass capacitor". The word decoupling implies the isolation of two circuits, which inherently requires a serial component, such as an inductor. What these capacitors do is shunt the high frequency currents away from the supply lines, hence the term bypass. Kundert has a nice paper that describes bypassing and decoupling on this website. It is called Power Supply Noise Reduction and you can find it at www.designers-guide.com/Design.

- MM -

Back to top

IP Logged

Fred

New Member

Offline

Posts: 6

Re: Caps
Reply #3 - Feb 19^th, 2004, 3:58pm

Thanks to you both Cao-Thong and Mighty mouse.

I have another question about Cao-Thong's post:

Suppose we have a common source amplifier with
a 100k ohm in the gate and the voltage source at
the gate has an AC and DC component.

does an AC current flowing thru a 100k ohm resistor
produce a dc voltage?

If it doesn't, is that why we need a cap for coupling
so the ac current will produce a dc voltage for VGS to bias the transistor? What about the dc voltage
that's already being supplied by the voltage source?

What about small signal AC voltage that was applied
at the gate in order to amplify? After we couple this
ac signal with the cap in front of the gate, will the
transistor still amplify this ac small signal?

thanks a lot,

Back to top

IP Logged

yeow

Guest

Re: Caps
Reply #4 - Feb 21^st, 2004, 2:26pm

Thanks to you both Cao-Thong and Mighty mouse.

I have another question about Cao-Thong's post:

Suppose we have a common source amplifier with
a 100k ohm in the gate and the voltage source at
the gate has an AC and DC component.

does an AC current flowing thru a 100k ohm resistor
produce a dc voltage?
--it will, if you think the roo-mean-square voltage.
--> if you considers, ac signal as the noise.

If it doesn't, is that why we need a cap for coupling
so the ac current will produce a dc voltage for VGS to bias the transistor?

---> We need coupling cap for blocking dc signal

What about the dc voltage
that's already being supplied by the voltage source?

--> You need coupling cap to block the dc signal

What about small signal AC voltage that was applied
at the gate in order to amplify?

--->I need to ask you back, are you going to ac gnd the
source and drain and also the substrate.

--> Since we have the junction, parasitic and fringing
cap, the ac will flow thru to each individual junction.

After we couple this
ac signal with the cap in front of the gate, will the
transistor still amplify this ac small signal?

---> it depends how you bias the transistor, oxide
cap and coupling cap value.

Thanks

Back to top

IP Logged