|
 |
Caps in IC design - RF Cafe Forums
|
marko
Post subject: Caps in IC design Posted: Sun Feb 26, 2006 2:00
am
Lieutenant
Joined: Sun Feb 26, 2006 1:56 am
Posts:
3
hello....someone asked me this question and i wasn't sure about
the answer. for a dc block why not pick the smallest possible capacitance
value because at dc no matter what the value is the cap will be an open.
the only answer i could think of was that if you pick a very small cap
value and depending on your frequency of operation you have unwanted
ac voltage drop across the blocking cap because small cap. value will
give rise to higher impedance. is there any other reason?
marko
Top
IR
Post subject: Posted: Sun Feb 26, 2006 3:05
am
Site Admin
Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373
Location: Germany
Hello marko,
For DC block
capacitors, the desired thing would be a low insertion loss for the
AC signal. Therefore if you will choose the lowest possible value you
will get the exact opposite.
The reactance value is given by:
Xc=1/(2*pi*f*C)
For most applications a reasonable value
of reactacne is between 3 to 5 ohms. When you choose the capacitor value,
it is of course most important to consider the frequency of operation,
ESR and the SRF of the capacitor. You should be below the SRF of the
capacitor, or else the capacitor might act as inductor.
There
are many technologies of capacitors: Y5V, X7R... They differ from each
other by temperature range, ESR value, SRF etc.
_________________
Best regards,
- IR
Top
marko
Post subject:
Now, about inductorPosted: Sun Feb 26, 2006 4:26 pm
Lieutenant
Joined: Sun Feb 26, 2006 1:56 am
Posts: 3
thanks IR for your
reply....one more question about passives in IC design. i know the definition
of Q in English but don't completely understand the concept, especially
this point: lower Q means more lossy inductor--does this mean that if
you have a very low Q then the current entering one port of inductor
won't equal current at the output because the lower Q means higher resistance
and capacitive coupling to substrate (current going into the substrate)?
is that essence of Q: that there would loss of current?
marko
Top
IR
Post subject: Posted: Sun Feb 26, 2006 4:41
pm
Site Admin
Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373
Location: Germany
Hello marko,
Just a quick
overview about Q Quality Factor:
There are 2 definitions:
1. Q defined as the ratio between the energy dissipated as heat
to the total energy stored in the sine wave.
2. The more modern
definition, which is valid for band-pass and band-stop filters:
Q=Fo/BW
The first definition is more of the unloaded Q and
can be used to answer your question:
Very low Q means that the
ratio of the energy being transferred through the inductor to the energy
being dissipated as heat in the inductor is high. That menas that the
inductor dissipates the AC energy in the form of heat due to the internal
DC resistrance of the wire.
Q=Xl/Rdc
If Rdc (DC resistance)
is high then Q is low. The same applies for capacitors only, instead
of Rdc there is ESR.
_________________
Best regards,
- IR
Top
marko
Post subject: Posted: Sun
Feb 26, 2006 5:53 pm
Lieutenant
Joined: Sun Feb 26, 2006
1:56 am
Posts: 3
hi IR...sorry i'm bugging you...so would you
say there's loss of current? i mean the heat dissipated has to be because
current is turned into heat from the series resistance and hence there
will loss of current...is that right?
Top
IR
Post subject: Posted: Mon Feb 27, 2006 4:03 pm
Site Admin
Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373
Location:
Germany
Hello marko,
There is no such thing loss of current.
There is loss of Enrgy and power. Energy is being dissipated in the
coil because of its associated Rdc.
By adding this resistance
ot the circuit, the current through the circuit is being reduced. If
the Q is high, then the power dissipation through the coil will be reduced
and more current will flow through it.
_________________
Best
regards,
- IR
Posted
11/12/2012
|
|
|
|
|
