difference between using 3 capacitors in parallel and 1 equi

[kris]

Hi,

I have seen a biasing configuration for an amplifier which has 3
capacitors in parallel from vcc to gnd which are probably acting as
blocking capacitors.
The values I have seen are as follows: 1uF closest to vcc, then 1000pF
,
then 100 pF. Is there any difference between connecting like this and
use only one capacitor with an equivalent value?

any help is greatly appreciated.

thanks
hk

 

[Michael Black]

kris (harihk@lycos.com) writes:

Hi,

I have seen a biasing configuration for an amplifier which has 3
capacitors in parallel from vcc to gnd which are probably acting as
blocking capacitors.
The values I have seen are as follows: 1uF closest to vcc, then 1000pF
,
then 100 pF. Is there any difference between connecting like this and
use only one capacitor with an equivalent value?

any help is greatly appreciated.

thanks
hk
Yes, there is a difference.

The three capacitors are not in parallel to get an exact value. INdeed,
for bypassing work, the exact value of a capacitor is fairly unimportant.

The capacitors are in parallel so there will be effective bypass at
all frequencies.

You need the 1uF to handle those low frequencies. You don't want audio
frequencies roaming the power buss, so you need a relatively high value.
Theoretically, at higher frequencies that same capacitor would still be
effective, because its impedance at higher frequencies would be even lower.
But that's theory. The 1uF capacitor is not a perfect capactior. It's
effectiveness at a higher frequency will not be so great, because the
very scheme to make that larger value capacitor means there is a fair
amount of inductance in series with that capacitor. At low frequencies,
that inductance is irrelevant. But at higher frequencies it becomes
significant, and so kills the effect of the capacitor at those higher
frequencies. The circuit, at higher frequencies, will not be seeing
1uF, so in effect there will be little bypass at higher frequencies.

So you put the 1000pF in parallel. That is too low a value for effective
use at those lower frequencies. But it will be effective at the higher
frequencies, since the lower value does not does not use the same
design as the 1uF, and hence will have lower internal inductance.

In some ways, it's seems puzzling to have the 100pF in parallel also,
because while there is a vast difference between the 1uF and the 1000pF,
there is much less so between the 1000pF and the 100pF. The reasoning
is the same, the 100pF will remain effective at higher frequencies
still, but the 100-pF will still be relatively effective.

Michael

 

[hk]

Michael,

Thanks a lot for the quick response. your response provided a lot of
insight beyond the theory I studied in textbooks.

regards,
hk

Yes, there is a difference.

The three capacitors are not in parallel to get an exact value. INdeed,
for bypassing work, the exact value of a capacitor is fairly unimportant.

The capacitors are in parallel so there will be effective bypass at
all frequencies.

You need the 1uF to handle those low frequencies. You don't want audio
frequencies roaming the power buss, so you need a relatively high value.
Theoretically, at higher frequencies that same capacitor would still be
effective, because its impedance at higher frequencies would be even lower.
But that's theory. The 1uF capacitor is not a perfect capactior. It's
effectiveness at a higher frequency will not be so great, because the
very scheme to make that larger value capacitor means there is a fair
amount of inductance in series with that capacitor. At low frequencies,
that inductance is irrelevant. But at higher frequencies it becomes
significant, and so kills the effect of the capacitor at those higher
frequencies. The circuit, at higher frequencies, will not be seeing
1uF, so in effect there will be little bypass at higher frequencies.

So you put the 1000pF in parallel. That is too low a value for effective
use at those lower frequencies. But it will be effective at the higher
frequencies, since the lower value does not does not use the same
design as the 1uF, and hence will have lower internal inductance.

In some ways, it's seems puzzling to have the 100pF in parallel also,
because while there is a vast difference between the 1uF and the 1000pF,
there is much less so between the 1000pF and the 100pF. The reasoning
is the same, the 100pF will remain effective at higher frequencies
still, but the 100-pF will still be relatively effective.

Michael

 

[Jamie]

it is very common to use a Disc Type cap or some very low inductive type
cap around the same area as you would see the large electro type/
the reason for this is due to problems with the electro type and their
construction make it hard to absorb high frequencies due to the coiling
of the foil in side and material used. thus the end results would cause
an high Lx value the be generated how ever, using a very low inductive
type cap you can solve the problem of the higher freq's
as far as having 2 larger types on the same circuit?. that is also
common in places where there may be a thin or low current foil run from
the original source for the main Cap.. this rung could case some
resistance and get flutter on the VCC.


kris wrote:

Hi,

I have seen a biasing configuration for an amplifier which has 3
capacitors in parallel from vcc to gnd which are probably acting as
blocking capacitors.
The values I have seen are as follows: 1uF closest to vcc, then 1000pF
,
then 100 pF. Is there any difference between connecting like this and
use only one capacitor with an equivalent value?

any help is greatly appreciated.

thanks
hk