EMC-PTSC Featured Discussion--March 2007

Electrolytic Capacitors in Primary Circuits

The IEEE Product Safety Engineering Society is seeking an author to write an article for our Newsletter as a follow-up to this discussion.  If you are interested please contact .  Refer to the information page or download the author's guide.

From: emc-pstc@ieee.org on behalf of Tarver, Peter [peter.tarver@sanmina-sci.com]
Sent: Friday, March 16, 2007 3:52 PM
To: emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

> From: Jody Leber
> Sent: Friday, March 16, 2007 11:38 AM
>
> The electrolytic capacitor that sits between the rectifier
> and transformer in SMPS is typically not an agency approved component.
> Where is this exemption documented in 60950-1, Clause 1.5.6?

Jody -

§1.5.6 looks at X and Y capacitors, and those connected between the primary circuit and earth.
The cathode of the electrolytic capacitor might not connect to earth, unless the bottom end of
the rectifier is earthed. If it doesn't connect to earth, no exception is needed.


Regards,

Peter L. Tarver, PE
ptarver@ieee.org

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Friday, March 16, 2007 4:23 PM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<BE3336BE85968D49BE01E66D6E365B1E01B5A784@SJC1AMFPEW01.am.sanm.corp>,
dated Fri, 16 Mar 2007, "Tarver, Peter" <peter.tarver@sanmina-sci.com>
writes:

>§1.5.6 looks at X and Y capacitors, and those connected between the
>primary circuit and earth. The cathode of the electrolytic capacitor
>might not connect to earth, unless the bottom end of the rectifier is
>earthed. If it doesn't connect to earth, no exception is needed.
>

I think a little clarification is necessary. Between the mains conductors
and the filter capacitor are, typically, a fuse, a common-mode choke, a
rectifier diode or a bridge rectifier and a resistor to limit inrush
current. In no case that I can envisage is the filter capacitor
connected to both mains conductors. In fact, that would probably cause
it to explode.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of pat.lawler@slpower.com
Sent: Friday, March 16, 2007 4:29 PM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

Hi Jody:

UL 60950-1 1st Ed., clause 1.5.6 ('Capacitors in primary circuits') says:
'A capacitor connected between two line conductors of the PRIMARY CIRCUIT,
or between one line
conductor and the neutral conductor, shall comply with IEC 60384-14:1993,
subclass X1 or X2.'

The electrolytic capacitor you mentioned is connected between the
rectifier '+' and '-', not between line conductors.

I always though it interesting that a 0.1uF film cap connected across the
line conductors going into a rectifier had to be safety-rated. But if you
move it to the output side of the rectifier, there's no safety rating
required.

Pat Lawler
SL Power Electronics Corp.

From: emc-pstc@ieee.org on behalf of Ted.Eckert@apcc.com
Sent: Friday, March 16, 2007 4:43 PM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

I have to agree with Mr. Woodgate that the components between the capacitor
and the AC mains make a difference. The available fault current on the AC
mains can be very high. It may be 1 kA, 10 kA or even more. A component
failure due to a short circuit can be very dramatic. The rectifier alone
will likely limit the fault current. If nothing else, the diodes will act
as fuses. You could design a rectifier circuit that would allow a high
fault current, but you would have to set out to do so and you would spend a
lot of money in the process.

Besides, proper abnormal condition testing will involve simulating a short
circuit on the electrolytic capacitor. The purpose of the test is to
verify that the system fails gracefully when the capacitor shorts out. You
can't easily do this test on a capacitor directly across the line. The
results of shorting out the X-capacitor are heavily dependent on the supply
circuit.

Ted Eckert
American Power Conversion/MGE
http://www.apc.com/

From: emc-pstc@ieee.org on behalf of Fred Townsend [fred@DCTOLIGHT.NET]
Sent: Saturday, March 17, 2007 2:55 AM
To: Ted.Eckert@apcc.com
Cc: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

See comments below.

Ted.Eckert@apcc.com wrote:

I have to agree with Mr. Woodgate that the components between the capacitor
and the AC mains make a difference. The available fault current on the AC
mains can be very high. It may be 1 kA, 10 kA or even more. A component
failure due to a short circuit can be very dramatic. The rectifier alone
will likely limit the fault current. If nothing else, the diodes will act
as fuses. You could design a rectifier circuit that would allow a high
fault current, but you would have to set out to do so and you would spend a
lot of money in the process.



Ted:
I have seen literally thousands of diode failures, mostly from lightning damage. I have never
seen a silicon power diode fuse (open). They always fail shorted at about two ohms. When they
fail the surge limiting resistor becomes very important. That's where they sometimes use
'fusible resistors'. Shorted diodes often cause secondary failures instead of protecting
anything.

Fred Townsend
DC to Light

From: emc-pstc@ieee.org on behalf of Robert Johnson [johnson@ITEsafety.com]
Sent: Saturday, March 17, 2007 2:04 PM
To: Ted.Eckert@APCC.COM; emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

The most dependable limiting factor in short circuit currents is usually the power cord. A six
foot 16 AWG cord on a 120 volt source will only allow 2500 amps and most branch circuit
impedances will lower that further to 1000 amps. Not that these are small numbers. They can
cause a lot of damage, but they are nowhere near the 10 kA the branch circuit breakers are
expected to limit to.

Bob Johnson
ITE Safety <http://www.itesafety.com>

From: emc-pstc@ieee.org on behalf of Ted.Eckert@apcc.com
Sent: Sunday, March 18, 2007 11:27 AM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In regard to Fred Townsend's comment: I am aware that diodes will normally
fail shorted. However, even in this state, they provide some current
limiting. The internal construction can only pass a given amount of
current before the diode will open. I have, however, seen a few poorly
designed circuits where there was no fuse on the input. The available
fault current was high enough that after the diode shorted out from the
overcurrent, it suffered from thermal damage and became an open circuit.
Even when the diode fails shorted, there is some resistance as noted which
will provide at least some current limiting.

In regards to Robert Johnson's comments: I work in an industry where we
have numerous field wired ITE products. I have products that are on a 50 A
branch circuit. The available fault current can be very high. The
selection of fuse for the switch-mode power supply is critical. Even with
a common mode choke providing some protection, I have seen fault currents
over 200 A when there is a component short after the diode bridge. Some of
the small, circuit board mounted fuses have an AIC rating of only 50 A and
they do not fail gracefully under worse conditions. I agree that most ITE
will not be exposed to high fault currents, but this is not always the
case.

Ted Eckert
American Power Conversion/MGE
http://www.apc.com/

From: emc-pstc@ieee.org on behalf of Fred Townsend [fred@DCTOLIGHT.NET]
Sent: Sunday, March 18, 2007 1:06 PM
To: Ted.Eckert@apcc.com
Cc: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

A few clarifications:

1. I will agree with Ted that many diode packages will fuse (sometimes explosively... I
once had a diode embed itself in my safety glasses) but always as a secondary failure. In other
words the diode shorted first and started passing AC. Then the electrolytic looked like a short
and then things start popping.
2. In any switching power circuit the EMI filter will be the controlling current limiter.
The inductance in the filter reacts to the leading edge of any surge, limiting current and in
many cases opening up before the fuse, or any other component, can react.


Fred Townsend
DC to Light

From: emc-pstc@ieee.org on behalf of Leber Jody-G19980 [jody.leber@motorola.com]
Sent: Monday, March 19, 2007 8:48 AM
To: pat.lawler@slpower.com
Cc: emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

Pat,

I am having trouble convincing an agency that the other side of the
rectifier is not "between line conductors". Are you or is anyone else
on the list aware of any actual documents that state this more
explicitly, provided everyone else agrees with this analysis.

Jody

From: emc-pstc@ieee.org on behalf of Tyra, John [John_Tyra@bose.com]
Sent: Monday, March 19, 2007 9:06 AM
To: Leber Jody-G19980; pat.lawler@slpower.com
Cc: emc-pstc@ieee.org
Subject: RE: Electrolytic Capacitors in Primary Circuits

Hello Jody,

I have a copy of the CENELEC Committee OSM decisions for the IEC60065
standard 6th/ 7th editions where the various CENELEC member Agencies
agreed that capacitors after a bridge rectifier in a SMPS do not need to
be approved. This would be in clause 14.2. I have a meeting in a few
minutes but will e-mail you a copy after the meeting.

Regards,

John

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 9:20 AM
To: emc-pstc@ieee.org
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<768EE6AB7D56D54BB5000EC2DD113E71016BE62E@de01exm61.ds.mot.com>, dated
Mon, 19 Mar 2007, Leber Jody-G19980 <jody.leber@motorola.com> writes:

>I am having trouble convincing an agency that the other side of the
>rectifier is not "between line conductors". Are you or is anyone else
>on the list aware of any actual documents that state this more
>explicitly, provided everyone else agrees with this analysis.

Look at definition 1.2.8.3. Unfortunately, IEC 60950-1 doesn't define
'directly connected', but the examples indicate what is meant. IEC 60065
does define 'directly connected', in 2.4.3.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 9:35 AM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<B11802460B4F4B4E963B51ADF2FAE08B04C4A86F@USMAFREXMB02.bose.com>, dated
Mon, 19 Mar 2007, "Tyra, John" <John_Tyra@bose.com> writes:

>I have a copy of the CENELEC Committee OSM decisions for the IEC60065
>standard 6th/ 7th editions where the various CENELEC member Agencies
>agreed that capacitors after a bridge rectifier in a SMPS do not need
>to be approved. This would be in clause 14.2. I have a meeting in a few
>minutes but will e-mail you a copy after the meeting.

Even so, I think that this matter is formally unclear in the context of
IEC/EN 60950 and action is needed to either amend the standard or have
an official interpretation issued by IEC TC108, not CENELEC.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of Tyra, John [John_Tyra@bose.com]
Sent: Monday, March 19, 2007 10:36 AM
To: John Woodgate; emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

Here is what the last OSM decisions I have state for both 60950 and
60065 for caps after a rectifier:

"Capacitors which are connected after a rectifier in a primary circuit
of a switch mode power supply unit need not be separately approved."

"In a primary circuit before a rectifier there is installation category
III for Permanently Connected equipment, therefore class X1 capacitors
must be used.
In a primary circuit before a rectifier there is installation category
II for Pluggable equipment Type A and Pluggable equipment Type B,
therefore minimum class X2 capacitors must be used. The use of a mains
fuse, a mains filter or a varistor cannot be a method to reduce
installation category."

"Secondary circuits are normally in installation category I when the
primary is in installation category II. However, a floating secondary
shall be subject to the requirements for primary circuit in table III
unless separated from primary circuits by an earthed metal screen."

Seems pretty clear to me but, since these are not legal modifications to
the standard(s) in questions, it depends on the Agency you are dealing
with as to whether they are willing to accept these decisions.

I agree it would be better to have the standard amended for
clarification or an official interpretation from TC108 but when you are
in the middle of an Agency submittal there is not usually time for this
so hopefully the Agency in question will consider the OSM decision and
change their interpretation favorably...

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 10:56 AM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<B11802460B4F4B4E963B51ADF2FAE08B04C4A871@USMAFREXMB02.bose.com>, dated
Mon, 19 Mar 2007, "Tyra, John" <John_Tyra@bose.com> writes:

>Seems pretty clear to me but, since these are not legal modifications
>to the standard(s) in questions, it depends on the Agency you are
>dealing with as to whether they are willing to accept these decisions.

OSM interpretations are valid in Europe unless, in a particular case, a
test house can give valid reason(s) to reject it. And many are adopted
by CENELEC as official interpretations by CENELEC TC108.

However, this particular interpretation can be a bit misleading insofar
as it concentrates on reasoning related to overvoltage (installation
categories) rather than to single-fault input current, which is what the
agency may have in mind.

The agency in question should take into account that every power supply
made has an electrolytic capacitor after the rectifier and there are no
safety-related 'approvals' for these capacitors when subjected to
alternating supply.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of Tarver, Peter [peter.tarver@sanmina-sci.com]
Sent: Monday, March 19, 2007 11:26 AM
To: John Woodgate; emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

> From: John Woodgate
> Sent: Friday, March 16, 2007 1:23 PM
>
> Tarver, Peter writes:
>
> >§1.5.6 looks at X and Y capacitors, and those connected between the
> >primary circuit and earth. The cathode of the electrolytic capacitor
> >might not connect to earth, unless the bottom end of the rectifier is
> >earthed. If it doesn't connect to earth, no exception is needed.
> >
>
> I think a little clarification is necessary. Between the mains
> conductors and the filter capacitor are, typically, a fuse, a
> common-mode choke, a rectifier diode or a bridge rectifier
> and a resistor to limit inrush current. In no case that I can
> envisage is the filter capacitor connected to both mains
> conductors. In fact, that would probably cause it to explode.

If the post-rectifier filter capacitor had one terminal tied to earth, it would need to comply
with Y1, Y2 or Y4 requirements, regardless of surges being impedance limited.

In 60950-1, §1.5.6, the term, "primary," is used throughout. "Mains," is not used. The concept
of, "direct connection," though not explicitly stated, is found in the use of the term, "line
conductor." "Line conductor," is not defined in 60950-1 (though it's used in more than just §
1.5.6) and it may be a failing in the standard to not define that term or to not include the
term, "direct connection." (Inference of the meaning of "line conductor" might be taken from §
1.2.1.1, but that's not very rigorous.)

All that aside, Jody's concern that someone was misreading the requirements in §1.5.6 for the
post-rectifier filter capacitor is well founded, unless one side of the capacitor is earthed
(I've never seen this in an SMPS, but that doesn't mean someone hasn't designed one that way).

I suspect the experience level of the engineer Jody's working with is on the low end of the
scale.

I have seen post-rectifier primary circuits in SMPSs connected to earth by small disk capacitors
that I doubt were Y capacitors and the voltages across them were only a few Volts. Moving
further and further away from the ac mains in the primary circuit would limit surges currents to
almost negligible levels, but §1.5.6 still requires such capacitors meet Y capacitor
requirements. Maybe the SMPS manufacturers should lobby TC108, MT2, to make more than one
clarification.


Regards,

Peter L. Tarver, PE
ptarver@ieee.org

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 11:44 AM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<BE3336BE85968D49BE01E66D6E365B1E01B5A8DB@SJC1AMFPEW01.am.sanm.corp>,
dated Mon, 19 Mar 2007, "Tarver, Peter" <peter.tarver@sanmina-sci.com>
writes:

>If the post-rectifier filter capacitor had one terminal tied to earth,
>it would need to comply with Y1, Y2 or Y4 requirements, regardless of
>surges being impedance limited.

In that case, it could not be an electrolytic capacitor. A 470 uF 375 V
Y-class capacitor would be very large and costly. In THAT case, I don't
suppose anyone would ever do it!
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of Don_Borowski@selinc.com
Sent: Monday, March 19, 2007 12:02 PM
To: emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

If the power supply in question has a bridge rectifier connected to the
mains of a power system with hot and neutral conductors, neither side of
the capacitor on the DC side of the bridge can be connected to earth --
there would be a connection from hot to earth every half-cycle through the
bridge rectifier.

I have seen probes connected to an earthed oscilloscope go up in a cloud of
smoke when the user forgot this in his haste to troubleshoot a problem on a
power supply or other non-transformer-isolated piece of equipment.

The only way this could possibly work (without letting out the smoke) would
be a single rectifier diode on the hot line, with neutral carried through
to the other side of the capacitor, and then that side of the capacitor
connected to earth. Of course, this will immediately actuate a ground-fault
protector. And there might be some regulatory agencies unhappy about the DC
load current presented to the AC power line.

Don Borowski
Schweitzer Engineering Labs
Pullman, WA, USA

From: emc-pstc@ieee.org on behalf of Tarver, Peter [peter.tarver@sanmina-sci.com]
Sent: Monday, March 19, 2007 12:15 PM
To: emc-pstc@IEEE.ORG
Subject: RE: Electrolytic Capacitors in Primary Circuits

> From: John Woodgate
> Sent: Monday, March 19, 2007 8:44 AM
>
> Tarver, Peter writes:
>
> >If the post-rectifier filter capacitor had one terminal tied
> to earth,
> >it would need to comply with Y1, Y2 or Y4 requirements,
> regardless of
> >surges being impedance limited.
>
> In that case, it could not be an electrolytic capacitor. A
> 470 uF 375 V Y-class capacitor would be very large and
> costly. In THAT case, I don't suppose anyone would ever do it!

We've been on the same page all along, John. My approach to Jody's post
has been to look at what the standard says and let any failure of logic
or design sense fall out from the readers' thoughts.

That's the way the standard is written. Having been involved in
standards writing as long as you have, I'm sure this isn't the first
failure of logic you've seen in the process.

Regards,

Peter L. Tarver, PE
ptarver@ieee.org

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 12:52 PM
To: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message
<BE3336BE85968D49BE01E66D6E365B1E01B5A90B@SJC1AMFPEW01.am.sanm.corp>,
dated Mon, 19 Mar 2007, "Tarver, Peter" <peter.tarver@sanmina-sci.com>
writes:

>We've been on the same page all along, John. My approach to Jody's
>post has been to look at what the standard says and let any failure of
>logic or design sense fall out from the readers' thoughts.

Indeed. I just wanted to spell it out in all its horror for those not
easily convinced!
>
>That's the way the standard is written. Having been involved in
>standards writing as long as you have, I'm sure this isn't the first
>failure of logic you've seen in the process.

Not by any means, and the time pressure now applied by the top
managements of standard bodies, the failure of industry to support
standards terminology and editorial work and the increasing number of
standards writers who were never taught English properly are combining
to make the situation rapidly deteriorate.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

From: emc-pstc@ieee.org on behalf of Robert A. Macy [macy@CALIFORNIA.COM]
Sent: Monday, March 19, 2007 1:38 PM
To: John Woodgate
Cc: emc-pstc@IEEE.ORG
Subject: Re: Electrolytic Capacitors in Primary Circuits

This question is not meant to sound argumentative, but
really is a sincere question. Your other two points are
spot on, therefore I assume this point is also. but I did
not understand it.

What do you mean "...failure of industry to support
standards terminology..."?

and what would it take to change that?

Robert

From: emc-pstc@ieee.org on behalf of John Woodgate [jmw@jmwa.demon.co.uk]
Sent: Monday, March 19, 2007 2:43 PM
To: Robert A. Macy
Cc: emc-pstc@ieee.org
Subject: Re: Electrolytic Capacitors in Primary Circuits

In message <web-141102949@california.com>, dated Mon, 19 Mar 2007,
Robert A. Macy <macy@california.com> writes:

>This question is not meant to sound argumentative, but really is a
>sincere question. Your other two points are spot on, therefore I
>assume this point is also. but I did not understand it.
>
>What do you mean "...failure of industry to support standards
>terminology..."?

IEC has a terminology committee, TC1, which shows how important the
founders of IEC rated the subject. But over the last two decades, as the
'old timers' who were members of the supporting Working Groups
affiliated to the product committees retired (some well into their
eighties) to the Great Plenary Meeting in the Sky, they have not been
replaced, in spite of repeated pleas. As a result, many of these WGs
have had to be disbanded, and TC1 is thus deprived of a large amount of
input from specialists in particular subjects.
>
>and what would it take to change that?

A miracle? Well, making available people who are allowed to devote
enough time to the job, who are good at precise language without being
overly pedantic, preferably know English and French (German, Spanish
and/or Russian would be bonuses) and, above all, are not assigned the
task because they are too unpredictable to be allowed anywhere near
product development.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
There are benefits from being irrational - just ask the square root of 2.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
 

###

The IEEE Product Safety Engineering Society is seeking an author to write an article for our Newsletter as a follow-up to this discussion.  If you are interested please contact .  Refer to the information page or download the author's guide.

Back to The Product Safety Engineering Society