Maker Pro
Maker Pro

ESR of Bypass caps

S

Sylvain Munaut

Jan 1, 1970
0
Hello,

I'd like to know if there is a big difference in using normal or low esr
caps for bypassing.

I use 10µF Tantalum caps near some group of chips ( in addition to
ceramic caps at each pins ), and I don't know wether to take low ESR (
100mOhm ) or normal ( 2500mOhm ) version of the caps. My application has
different "section", some digital, other analog, other audio. Maybe some
require low ESR and other don't. The low ESR part are four times more
expensive, so using them where not needed is a pretty big waste.

Thanks,
Sylvain Munaut
 
J

Jeroen Belleman

Jan 1, 1970
0
Sylvain said:
I'd like to know if there is a big difference in using normal or low esr
caps for bypassing.

I use 10µF Tantalum caps near some group of chips ( in addition to
ceramic caps at each pins ), and I don't know wether to take low ESR (
100mOhm ) or normal ( 2500mOhm ) version of the caps. My application[...]

Ordinary Ta caps have about 300mOhm ESR. If yours are 2.5 Ohms, they're
duds.

Jeroen
 
J

John Popelish

Jan 1, 1970
0
Sylvain said:
Hello,

I'd like to know if there is a big difference in using normal or low esr
caps for bypassing.

I use 10µF Tantalum caps near some group of chips ( in addition to
ceramic caps at each pins ), and I don't know wether to take low ESR (
100mOhm ) or normal ( 2500mOhm ) version of the caps. My application has
different "section", some digital, other analog, other audio. Maybe some
require low ESR and other don't. The low ESR part are four times more
expensive, so using them where not needed is a pretty big waste.

Thanks,
Sylvain Munaut

What is the largest current change your circuit can generate as a
step? That is what produces voltage drop across the bypass capacitor
ESR while the power supply lead inductance and regulator control loop
react to those current demand changes. Are you at all sure that those
tantalums are doing anything useful, at all?
 
T

Tim Wescott

Jan 1, 1970
0
Sylvain said:
Hello,

I'd like to know if there is a big difference in using normal or low esr
caps for bypassing.

I use 10µF Tantalum caps near some group of chips ( in addition to
ceramic caps at each pins ), and I don't know wether to take low ESR (
100mOhm ) or normal ( 2500mOhm ) version of the caps. My application has
different "section", some digital, other analog, other audio. Maybe some
require low ESR and other don't. The low ESR part are four times more
expensive, so using them where not needed is a pretty big waste.

Thanks,
Sylvain Munaut

Normal application of bypass caps is 100nF ceramics close to each chip,
with electrolytics somewhere on the board for overall power supply sag.
The low-ESR tantalums really came about because the LDO regulators
require them for stability (an LDO by itself is a current source,
without a low-ESR cap right at it's output it has very high gain at high
frequencies).

In theory you could use low-ESR tantalums, but you want to bypass _each_
chip, and you want the bypass chips to be _close_ -- like right next to
the chip, or on the back side of the board with multiple vias for low
inductance.
 
S

Sylvain Munaut

Jan 1, 1970
0
Thanks for your reply.

What is the largest current change your circuit can generate as a
step?

Well, that's a good question ...
How can I determine that ?

Basically, the circuits as some digital I2S interface to a
microcontroller board and route that to a optical tranceiver IC, an
optical receiver, and a 8channel DAC to have 4 stereo output ( not
amplified ).

That is what produces voltage drop across the bypass capacitor
ESR while the power supply lead inductance and regulator control loop
react to those current demand changes. Are you at all sure that those
tantalums are doing anything useful, at all?

Well, "all those" might be a little exagerated, I just have about 5 of them.

Two are near a group of 6 ampli ops and are on +12V and -12V. Since
theses are audio, I just thinked that be a good idea to add those.

One is near the DAC that drives these ampli ops and is for 5V.

The two others are at the 3.3V & 5V input near the alimentation connector.



Sylvain Munaut
 
J

John Popelish

Jan 1, 1970
0
Sylvain said:
Thanks for your reply.


Well, that's a good question ...
How can I determine that ?

If all else fails, you might put a low value resistor in series with
the common lead of the chip and measure the instantaneous voltage
drop across it as the chip operates. It is certainly a good idea to
connect the scope directly across ant sensitive or noisy chip and see
what kind of ripple voltage it is generating/experiencing with a given
bypass arrangement. If you can't see any significant different in the
supply peak ot peak voltage when you remove a given cap (or substitute
a cheaper one) then the cap is not paying its way. I would also
consider substituting a multi layer ceramic for any non timing
application of tantalums for a cost/failure mode improvement. A
couple microfarads at 25 volts does not cost much.
Basically, the circuits as some digital I2S interface to a
microcontroller board and route that to a optical tranceiver IC, an
optical receiver, and a 8channel DAC to have 4 stereo output ( not
amplified ).


Well, "all those" might be a little exagerated, I just have about 5 of them.

Two are near a group of 6 ampli ops and are on +12V and -12V. Since
theses are audio, I just thinked that be a good idea to add those.

If you want ot keep high frequency noise out of opamps (which reject
low frequency variations pretty well) you might consider adding a low
value series resistor between them and their bypass caps and the noisy
power rail. This rolls off the high frequency noise more.
One is near the DAC that drives these ampli ops and is for 5V.

I would consider a bigger ceramic cap, here, instead.
The two others are at the 3.3V & 5V input near the alimentation connector.

Try aluminum electrolytics and see if you can see any degradation in
the supply noise level. Larger capacitance will cost less money.
 
T

The other John Smith

Jan 1, 1970
0
Sylvain Munaut said:
Hello,

I'd like to know if there is a big difference in using normal or low esr
caps for bypassing.

I use 10µF Tantalum caps near some group of chips ( in addition to
ceramic caps at each pins ), and I don't know wether to take low ESR (
100mOhm ) or normal ( 2500mOhm ) version of the caps. My application has
different "section", some digital, other analog, other audio. Maybe some
require low ESR and other don't. The low ESR part are four times more
expensive, so using them where not needed is a pretty big waste.

Thanks,
Sylvain Munaut


Tuesday night I attended a seminar hosted by Texas Instruments for the IEEE
Dallas EMC Society on bypassing. See http://www.sigcon.com/ for background.
Dr. Howard Johnson, author of "High Speed Signal Propagation" was the
speaker.

He made the following points on bypassing:

1. Use the smallest ceramic capacitor with the lowest parasitic inductance
you can find.

2. Put the plated-through holes as close to the body as possible to one
side.

o.-.
| |
o'-'

3. Don't be much concerned with the ESR (ceramic caps). Too low an ESR can
cause ringing.

All of this, he said, is covered in his book.

Of course, he was addressing high speed logic design. If you're doing audio
stuff, you may not be concerned with MHz to Ghz bypassing.

Anyway, that's the best I can remember. Good luck.

John
 
S

Sylvain Munaut

Jan 1, 1970
0
Hello
Tuesday night I attended a seminar hosted by Texas Instruments for the IEEE
Dallas EMC Society on bypassing. See http://www.sigcon.com/ for background.
Dr. Howard Johnson, author of "High Speed Signal Propagation" was the
speaker.

He made the following points on bypassing:

1. Use the smallest ceramic capacitor with the lowest parasitic inductance
you can find.

2. Put the plated-through holes as close to the body as possible to one
side.

o.-.
| |
o'-'

3. Don't be much concerned with the ESR (ceramic caps). Too low an ESR can
cause ringing.

All of this, he said, is covered in his book.

Thanks, I'll have a look, given the title it may have some good advice
on other matters I'm interested in, like high speed bus layout and
termination.

Of course, he was addressing high speed logic design. If you're doing audio
stuff, you may not be concerned with MHz to Ghz bypassing.

Well, I'm doing audio but the audio is being fetched/decoded by a 400
Mhz PowerPC so I also have some bus at such speed.


Sylvain Munaut
 
S

Sylvain Munaut

Jan 1, 1970
0
Well,

thank you very much for all theses advices, I'll see what I can come up
with ;)

If you want ot keep high frequency noise out of opamps (which reject
low frequency variations pretty well) you might consider adding a low
value series resistor between them and their bypass caps and the noisy
power rail. This rolls off the high frequency noise more.

What I want to avoid at all cost is noise in the hearing range. Anything
above 50kHz won't be heard anyway ( and filtred by some later filter ).

As english is not my native language, just to be sure, it's :

Noisy Rail -----|---\/\/\/-----> Chips
|
---
---Bypass
|
###
AGND


Again, thanks for your time.


Sylvain Munaut
 
S

Sylvain Munaut

Jan 1, 1970
0
Hi,

Normal application of bypass caps is 100nF ceramics close to each chip,
with electrolytics somewhere on the board for overall power supply sag.
In theory you could use low-ESR tantalums, but you want to bypass _each_
chip, and you want the bypass chips to be _close_ -- like right next to
the chip, or on the back side of the board with multiple vias for low
inductance.

Yes, I have those as well. 100nF ceramic for all chips, excepts the
critical ones where I have 100nF and 3.3nF.

The main processor also have other caps/inductors but there, I'm just
following the datasheet that dictates exactly how it should be so ...

The low-ESR tantalums really came about because the LDO regulators
require them for stability (an LDO by itself is a current source,
without a low-ESR cap right at it's output it has very high gain at high
frequencies).

Good to know.


Thanks,

Sylvain Munaut
 
J

John Popelish

Jan 1, 1970
0
Sylvain said:
John Popelish wrote:

What I want to avoid at all cost is noise in the hearing range. Anything
above 50kHz won't be heard anyway ( and filtred by some later filter ).

As english is not my native language, just to be sure, it's :

Noisy Rail -----|---\/\/\/-----> Chips
|
---
---Bypass
|
###
AGND

My english was pretty poor. The resistor goes between the rail and
the parallel combination of chip and capacitor (to form a low pass
filter in the supply line and to make the Aground cleaner.

10-220 ohms (a few % DC supply loss)
Noisy Rail --\/\/\/--+--> Chips
|
---
---Bypass
|
###
AGND

This dumps less high frequency current into the Aground than just
using capacitors. It also provides a loss to absorb ringing that is
bouncing around on the board between high Q capacitors and trace
inductance. This approach is only useful on low current loads (where
the resistor can be high enough (relative to the cap ESR) to help the
lower the roll off frequency.

I have also used ferrite beads on leads instead of resistors (works
for higher current loads).

Some analog chips are nonlinear with respect to high frequency noise
and you hear the envelope of the noise pulse density. If the number
of spikes per second varies (because of changes in the operation of
logic), you hear a low frequency AM demodulation of that noise. So
the filter may need to only be effective well above the audio
frequency range to reduce such nonlinear effects and have a noticeable
benefit. This, of course depends on the particular chips used.
 
I

Ian Buckner

Jan 1, 1970
0
Sylvain Munaut said:
The two others are at the 3.3V & 5V input near the alimentation connector.



Sylvain Munaut

I like that - "alimentation connector" ;-)

On the decoupling, put some 0.1uF ceramics close to the chips.
They are very low ESR, and should allow you not to care about
the electrolytic ESR (although 2.5 ohms sounds very high).

Regards
Ian
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Ian Buckner
I like that - "alimentation connector" ;-)

It's French. German has 'Speisespannung' - somewhat similar. We have
'coaxial feeders' and 'feed wires' etc.
 
S

Sylvain Munaut

Jan 1, 1970
0
"High-Speed Digital Design: A Handbook of Black Magic" Howard W. Johnson &
Martin Graham, Prentice-Hall 1993, ISBN 0-13-395724-1

this book is fabulous, and dead useful! Bob Pease praised it in EDN in the
early nineties, so I bought a copy; I browsed thru it at lunchtime, and
found a great way to measure ESR & ESL of caps, which I used that afternoon!
Its not too heavy on the theory, has useful chapter summaries, and he gives
you all of the equations he uses.

Yes, by looking at the "High speed signal propagation" on Amazon, I
viewed this one on the recommended and my university library had a copy,
I had a quick look and it seemed interesting, I think I'll buy myself a
copy for a more in depth reading.

Sylvain Munaut
 
J

Joerg

Jan 1, 1970
0
Absolutely agree, "High Speed Digital Design" is a book that any EE who does fast
designs should have. I paid $50 for it but that was I believe in 1993 right when
it came out. Just couldn't wait. My copy sits right next to "The Art of
Electronics", another "must have".

Regards, Joerg
 
T

Terry Given

Jan 1, 1970
0
Sylvain Munaut said:
Thanks, I'll have a look, given the title it may have some good advice
on other matters I'm interested in, like high speed bus layout and
termination.


"High-Speed Digital Design: A Handbook of Black Magic" Howard W. Johnson &
Martin Graham, Prentice-Hall 1993, ISBN 0-13-395724-1

this book is fabulous, and dead useful! Bob Pease praised it in EDN in the
early nineties, so I bought a copy; I browsed thru it at lunchtime, and
found a great way to measure ESR & ESL of caps, which I used that afternoon!
Its not too heavy on the theory, has useful chapter summaries, and he gives
you all of the equations he uses.
 
T

Terry Given

Jan 1, 1970
0
John Popelish said:
My english was pretty poor. The resistor goes between the rail and
the parallel combination of chip and capacitor (to form a low pass
filter in the supply line and to make the Aground cleaner.

10-220 ohms (a few % DC supply loss)
Noisy Rail --\/\/\/--+--> Chips
|
---
---Bypass
|
###
AGND

This dumps less high frequency current into the Aground than just
using capacitors. It also provides a loss to absorb ringing that is
bouncing around on the board between high Q capacitors and trace
inductance. This approach is only useful on low current loads (where
the resistor can be high enough (relative to the cap ESR) to help the
lower the roll off frequency.

I have also used ferrite beads on leads instead of resistors (works
for higher current loads).

Some analog chips are nonlinear with respect to high frequency noise
and you hear the envelope of the noise pulse density. If the number
of spikes per second varies (because of changes in the operation of
logic), you hear a low frequency AM demodulation of that noise. So
the filter may need to only be effective well above the audio
frequency range to reduce such nonlinear effects and have a noticeable
benefit. This, of course depends on the particular chips used.

Some very good points here.

The general idea is to critically/heavily damp every LC circuit.

Your power supply LC circuit will be the power supply trace inductance and
the decoupling caps (which no doubt completely swamp your trace
capacitance).

An 0V plane with PSU traces will have a moderate inductance, whereas 0V and
PSU planes will have an extremely low inductance (sub 100nH) and therefore
lower characteristic impedance (conversely a psu trace + badly routed 0V
trace will have a very high L and hence high Zo).

Because C is high, Fres is fairly low (even though L can be quite low) and
so it looks like a lumped circuit. This means that your ac-coupled shunt
damping resistor (ie electro/tant cap + ESR) can be pretty much anywhere opn
the PCB. Because it is a shunt damping resistor, a lower value R will
increase damping (to a certain extent).

If you measure a few boards, and do a few calculations, you'll find the
damping resistor tends to work out at a few ohms. One test I like to do is
stuff caps onto an otherwise blank PCB, then do step-response tests and
measure inductance. You can design it like an RC snubber - using no damping
R+C, masure Fres. Add shunt (lossless) capacitance Cx until Fres halves - Cx
= 3Cres, and can now calculate Lres = 1/[(2piFres)^2*Cres] and Rdamp = Zo =
sqrt(Lres/Cres) and Cdamp >= 3Cres. Unitrode have a nice app note on how to
design a snubber.

depending on what you are doing, you may or may not get away with cap ESR to
damp your psu - ESR can increase dramatically at low temperatures, may vary
widely from part-to-part and likely changes significantly with age. When I
damp SMPS filters I tend to use low-ESR caps + series resistors (caveat:
peak pulse power at startup) when looking for long-term reliability. Im not
sure about tantalum ESR behaviour though, every time I consider them I can
do it cheaper with electrolytics/smt ceramics.....

And of course if you dont excite the resonant circuit, it wont ring - but
square waves are rich in harmonics, an edge rise/fall of 10ns has a "knee"
frequency of around 32MHz, and plenty thereafter.....very fast transitions
can excite the resonances between paralleled bypass caps (eg
100nF||3.3nF)...the plot thickens.....and a crappy layout can pick up noise
(outisde or self-generated) that excites PSU resonance......

cheers
Terry
 
Top