Grounding a chassis for line voltage safety

[Chris Carlen]

Greetings:

Usually I tie the ground wire from the line input to the metal chassis.
But sometimes the chassis has several different metal components
screwed together, and perhaps with painted surfaces that may interfere
with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?

I'd like to make sure things are done right in a project that will be
handed over for completion by another technologist.



Thanks for comments.



Good day!



--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crcarle@sandia.gov

 

[Terry Given]

"Chris Carlen" <crcarle@BOGUS.sandia.gov> wrote in message
news:c4ad3s01dla@enews2.newsguy.com...

Greetings:

Usually I tie the ground wire from the line input to the metal chassis.
But sometimes the chassis has several different metal components
screwed together, and perhaps with painted surfaces that may interfere
with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?

I'd like to make sure things are done right in a project that will be
handed over for completion by another technologist.



Thanks for comments.



Good day!



--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crcarle@sandia.gov

Hi Chris,

the "rules" invariably depend on the product, its application etc. My
particular field is power electronics, and this is what we do:

ALL exposed metalwork must be SECURELY earthed, such that the "rated fault
current" (whatever that means) can flow through said earth connection for 1
minute without causing damage. The last couple of places I worked (designing
motor controllers & UPS) we simply used the rated current - a 600A drive
with 2x overload therefore had to support a 1200Arms fault current through
our grounding straps, for 1 minute (we used 50mm braid originally, which
glowed red-hot with 1200Arms flowing through it - it was replaced with a
large piece of cable, 100mm^2 IIRC). The rationale behind using maximum
rated current is of course the device fuse ratings.

most electrical earthing requirements tend to require a total earth
impedance of < 1 Ohm, which is a good start, but clearly it depends on your
fault current rating - 1200^2*1 = 1.44MW into 1 Ohm, leading to instant
destruction of earth conductor.......

If you meet the fault current spec, you will most likely meet any other
requirements, but it depends on the standards you must meet - for example
some UL standards require GREEN screws for earth connections (usually this
only applies to user terminations; you can normally do as you choose inside
the box)

Hope this helps

Terry Given
Domes Engineering
109 Centennial Avenue
Te Aroha
New Zealand
ph +64 7 8844 596 (voicemail)
cell +64 21 422 400 (no voicemail)
terry_given@ieee.org

 

[John Larkin]

On Mon, 29 Mar 2004 15:56:44 -0800, Chris Carlen
<crcarle@BOGUS.sandia.gov> wrote:

Greetings:

Usually I tie the ground wire from the line input to the metal chassis.
But sometimes the chassis has several different metal components
screwed together, and perhaps with painted surfaces that may interfere
with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?

A few UL/CE rules are...

The grounding wire must be green with yellow stripe

The wire must be grounded to the same piece of metal that the AC
enters through.

The grounding stud may not be shared with any other connection or
function; you can't use, say, a transformer mounting bolt as the
ground connection.

Crimped ring lugs are OK. We usually solder a wire to a Corcom AC
power inlet ground tab, with a crimped ring lug on the other end.


We've been OK (five or so boxes passed CE so far) with the various
metal pieces being bolted together, with no specific ground straps
between different parts of the box.

John

 

[Mac]

On Mon, 29 Mar 2004 15:56:44 +0000, Chris Carlen wrote:

Greetings:

Usually I tie the ground wire from the line input to the metal chassis.
But sometimes the chassis has several different metal components
screwed together, and perhaps with painted surfaces that may interfere
with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?

I'd like to make sure things are done right in a project that will be
handed over for completion by another technologist.



Thanks for comments.



Good day!

I don't actually know regulatory requirements, but I've noticed that some
of the really well built industrial computers we've bought have a
dedicated grouding lug near the power entry module. Other than that, they
seem to rely on bare metal connections held together with machine screws.

Oh, the hard drive sub-chassis is usually grounded with a metal braid,
because it is shock mounted on rubber bushings and would not otherwise
make good contact with the chassis.

--Mac

 

[John Fields]

On Tue, 30 Mar 2004 07:26:45 -0800, Chris Carlen
<crcarle@BOGUS.sandia.gov> wrote:

Sounds like I'm on the right track. Our "in-house" instrumentation
never sees UL/CE approval testing, so I usually just use "common sense,"
which seems to be agreeing quite well with the usual ways of doing things.

---
One thing you might watch for is that if you bolt two aluminum
surfaces together and you want them to be electrically connected to
each other you'll need to make sure you break through the ever-present
oxide layer to the parent metal. One way to do this is to use
internal-tooth or external-tooth lockwashers against the aluminum
surfaces so that when the nut an bolt are tightened they (the washers)
bite into the surfaces of the aluminum panels as well as into the bolt
and nut.

--
John Fields

 

[John Woodgate]

I read in sci.electronics.design that John Fields <jfields@austininstrum
ents.com> wrote (in <406f9d31.599983984@news.texas.net&gt about
'Grounding a chassis for line voltage safety', on Tue, 30 Mar 2004:

you'll need to make sure you break through the ever-present
oxide layer to the parent metal.


Have you ever found that to be actually necessary? I can take a couple

of bits of really dull Al sheet, press them together and they contact
down in the milliohm region.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Walter Harley]

"Terry Given" <the_domes@xtra.co.nz> wrote in message
news:Tm3ac.4405$Tf3.72350@news.xtra.co.nz...

most electrical earthing requirements tend to require a total earth
impedance of < 1 Ohm, which is a good start, but clearly it depends on
your
fault current rating - 1200^2*1 = 1.44MW into 1 Ohm, leading to instant
destruction of earth conductor.......

That would seem to suggest a second criterion, that Rg < Esource / Ifuse.
That is, you'd better be able to blow the fuse by faulting to ground; if the
ground termination is too resistive to blow the fuse at the expected
voltage, then the ground won't do its job.

If the 600A device is powered by, say, 240V, then to achieve 1200A to trip
the fuse, you'd need 0.2 ohms or less. Of course, you still have to cope
with 288kW at that point.

Boy, that's a pretty beefy widget you've got there, Terry! (I know, I bet
you hear that all the time.)

 

[Active8]

On Tue, 30 Mar 2004 07:26:45 -0800, Chris Carlen wrote:

John Larkin wrote:
On Mon, 29 Mar 2004 15:56:44 -0800, Chris Carlen
crcarle@BOGUS.sandia.gov> wrote:

Greetings:

Usually I tie the ground wire from the line input to the metal chassis.
But sometimes the chassis has several different metal components
screwed together, and perhaps with painted surfaces that may interfere
with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?


A few UL/CE rules are...

The grounding wire must be green with yellow stripe

Oh well, I don't do that, just plain green.

The wire must be grounded to the same piece of metal that the AC
enters through.

Yeah, I do that.

The grounding stud may not be shared with any other connection or
function; you can't use, say, a transformer mounting bolt as the
ground connection.

Hmm. I've gotten lazy a few times lately and done that.

Crimped ring lugs are OK. We usually solder a wire to a Corcom AC
power inlet ground tab, with a crimped ring lug on the other end.

We've been OK (five or so boxes passed CE so far) with the various
metal pieces being bolted together, with no specific ground straps
between different parts of the box.

Interesting.


John

Sounds like I'm on the right track. Our "in-house" instrumentation
never sees UL/CE approval testing, so I usually just use "common sense,"
which seems to be agreeing quite well with the usual ways of doing things.

Good day!

Electronic Instrument Design - Fowler.

Covers grounding and a pile of other stuff that you might find
interesting. Check out the table of contents at your favorite online
bookstore.
--
Best Regards,
Mike

 

[John Fields]

On Tue, 30 Mar 2004 19:09:15 -0800, Chris Carlen
<crobc@BOGUS_FIELD.earthlink.net> wrote:

John Woodgate wrote:
I read in sci.electronics.design that John Fields <jfields@austininstrum
ents.com> wrote (in <406f9d31.599983984@news.texas.net&gt about
'Grounding a chassis for line voltage safety', on Tue, 30 Mar 2004:

you'll need to make sure you break through the ever-present
oxide layer to the parent metal.


Have you ever found that to be actually necessary? I can take a couple
of bits of really dull Al sheet, press them together and they contact
down in the milliohm region.

The current should tunnel through such thin layers, no?

---
I've never had a failure doing it the "right" way, but who knows what
would have happened if I'd been sloppy? It's only a couple of
washers, for God's sake.

Time and moisture play strange tricks, and it's better to head them
off at the pass when we can instead of having to clean up afterwards,
wouldn't you agree?

"Should"? Yeah, right...

--
John Fields

 

[John Woodgate]

I read in sci.electronics.design that Chris Carlen <crobc@BOGUS_FIELD.ea
rthlink.net> wrote (in <c4dcor0hkl@enews4.newsguy.com&gt about 'Grounding
a chassis for line voltage safety', on Tue, 30 Mar 2004:

John Woodgate wrote:
I read in sci.electronics.design that John Fields <jfields@austininstrum
ents.com> wrote (in <406f9d31.599983984@news.texas.net&gt about
'Grounding a chassis for line voltage safety', on Tue, 30 Mar 2004:

you'll need to make sure you break through the ever-present
oxide layer to the parent metal.


Have you ever found that to be actually necessary? I can take a couple
of bits of really dull Al sheet, press them together and they contact
down in the milliohm region.

The current should tunnel through such thin layers, no?


I don't think it's tunnelling, just that the oxide film breaks down at a

volt or two.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Paul Burke]

John Woodgate wrote:

Have you ever found that to be actually necessary? I can take a couple
of bits of really dull Al sheet, press them together and they contact
down in the milliohm region.

Every time, except the one time you really need it. Plain aluminium is
ususally OK, but don't rely on contact if it's anodised. This only makes
contact if you DON'T want it to.

Paul Burke

 

[John Woodgate]

I read in sci.electronics.design that Paul Burke <paul@scazon.com> wrote
(in <c4dq39$2hd054$1@ID-128611.news.uni-berlin.de&gt about 'Grounding a
chassis for line voltage safety', on Wed, 31 Mar 2004:

John Woodgate wrote:


Have you ever found that to be actually necessary? I can take a couple
of bits of really dull Al sheet, press them together and they contact
down in the milliohm region.

Every time, except the one time you really need it. Plain aluminium is
ususally OK, but don't rely on contact if it's anodised. This only makes
contact if you DON'T want it to.

Deliberate anodizing makes a much thicker oxide film than forms

naturally. The subject is *natural* oxidation.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[John Woodgate]

I read in sci.electronics.design that mike <spamme0@juno.com> wrote (in
<406AF871.40206@juno.com&gt about 'Grounding a chassis for line voltage
safety', on Wed, 31 Mar 2004:

It's good to seek opinion. But in the end there's only one person's
opinion that matters. The certification agency. And there's probably
more than one. Get the written test criteria from every agency who needs to
approve your product and try to find a solution that makes them
all happy.

Doesn't apply in Europe. No 'certification'.

At the final certification test is NOT the time to find out that the guy
you met on the internet had an obsolete standards document.

Yes, current editions of standards are essential. But some of us already
don't give advice based on old standards. Don't tar everyone with the
same brush.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Chris Carlen]

mike wrote:

Chris Carlen wrote:

Greetings:

Usually I tie the ground wire from the line input to the metal
chassis. But sometimes the chassis has several different metal
components screwed together, and perhaps with painted surfaces that
may interfere with an electrical connection.

What are "the rules" about grounding chassis with multiple metal
components assembled together? Must a ground wire connect via a
"standard" electrical termination like a ring terminal to every metal
component? Or if the chassis elements contact each other tightly and
there is bare metal to metal contact, is that sufficient?

I'd like to make sure things are done right in a project that will be
handed over for completion by another technologist.



Thanks for comments.



Good day!


It's good to seek opinion. But in the end there's only one person's
opinion that matters. The certification agency. And there's probably
more than one. Get the written test criteria from every agency who
needs to approve your product and try to find a solution that makes them
all happy.

At the final certification test is NOT the time to find out that the guy
you met on the internet had an obsolete standards document.

mike

Yes, I should probably get familiar with a specific cert. agency. I
haven't done so yet since nothing we do here ever gets used other than
internally as custom lab instrumentation.

Unfortunately, I have found some very incorrectly assembled line powered
devices here produced in the past. One had no connection of the metal
chassis and frame to earth, and had the hot line very sloppily bare
wired to various components inside the metal box. There is a
consequence to the management not knowing the difference between
technologists and engineers, and assuming that technologists "can do
anything." No one even realizes that there exists vast bodies of
knowledge outside their awareness. But that is a gripe for another day.

What I am hesitant to do though, is to propose that we have all
internally used line powered stuff certified. There have been efforts
to make this happen in the past, by the safety folks, and vigorously
opposed by techs. and management. Considering what I said about having
found badly wired chassis, it might actually by the most ethical and
moral thing to do to advocate requiring UL or other cert. for line
powered instruments produced by techs., since I have some evidence and
strong opinions that knowledge about how to do things correctly is not
as ubiquitous as assumed.

But that would make me the least popular tech. around here. :-(

Not sure what to do about this. Have to think on that.

But I may heed your advice and seek the current printed standards from
an agency. Would you think that UL would be the first choice?


Thanks for input.




--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crcarle@sandia.gov

 

[John Woodgate]

I read in sci.electronics.design that Chris Carlen
<crcarle@BOGUS.sandia.gov> wrote (in <c4f0060eh3@enews2.newsguy.com&gt
about 'Grounding a chassis for line voltage safety', on Wed, 31 Mar
2004:

Not sure what to do about this. Have to think on that.

There are lots of guys who run training course on making safe products.
You are too far away for me, though.

But I may heed your advice and seek the current printed standards from an
agency. Would you think that UL would be the first choice?

For USA, and a good guide for Canada and Mexico, yes. Not for Europe,
for which you probably need IEC/EN 61010-1.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Frank Bemelman]

"Chris Carlen" <crcarle@BOGUS.sandia.gov> schreef in bericht
news:c4f0060eh3@enews2.newsguy.com...

[snip]

But that would make me the least popular tech. around here. :-(

Not sure what to do about this. Have to think on that.

Go for popular And apply some common sense, here and
there.

--
Thanks, Frank.
(remove 'x' and 'invalid' when replying by email)

 

[John Woodgate]

I read in sci.electronics.design that Walter Harley
<walterh@cafewalterNOSPAM.com> wrote (in <c4f4rt$k4q$0@216.39.172.65&gt
about 'Grounding a chassis for line voltage safety', on Wed, 31 Mar
2004:

"Chris Carlen" <crcarle@BOGUS.sandia.gov> wrote in message
news:c4f0060eh3@enews2.newsguy.com...
Yes, I should probably get familiar with a specific cert. agency. I
haven't done so yet since nothing we do here ever gets used other than
internally as custom lab instrumentation.

And is probably one-off construction in many cases? I've heard that UL
certification is not only expensive (>$10k), they require several units to
destructively test. Very impractical for a lab situation, even one with
defense funding, I'd think. Also I'm not sure how long the process takes.

But that's just hearsay, not knowledge. Perhaps someone with real
information can correct me.


It's substantially true. Certification is inappropriate for such one-off

equipment. What is by NO means inappropriate is *making the equipment
safe according to applicable standards*. And that is important for
health and safety at work purposes, which are really rather important.

You can make equipment 'safe' in this context by attending to design
issues; it certainly isn't necessary to carry out destructive testing if
the design is done with proper insight and a thorough understanding of
the requirements of the applicable standard.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[Chris Carlen]

John Woodgate wrote:

I read in sci.electronics.design that Walter Harley
walterh@cafewalterNOSPAM.com> wrote (in <c4f4rt$k4q$0@216.39.172.65&gt
about 'Grounding a chassis for line voltage safety', on Wed, 31 Mar
2004:

"Chris Carlen" <crcarle@BOGUS.sandia.gov> wrote in message
news:c4f0060eh3@enews2.newsguy.com...

Yes, I should probably get familiar with a specific cert. agency. I
haven't done so yet since nothing we do here ever gets used other than
internally as custom lab instrumentation.

And is probably one-off construction in many cases? I've heard that UL
certification is not only expensive (>$10k), they require several units to
destructively test. Very impractical for a lab situation, even one with
defense funding, I'd think. Also I'm not sure how long the process takes.

But that's just hearsay, not knowledge. Perhaps someone with real
information can correct me.



It's substantially true. Certification is inappropriate for such one-off
equipment. What is by NO means inappropriate is *making the equipment
safe according to applicable standards*. And that is important for
health and safety at work purposes, which are really rather important.

Wowy-dowy! If that is what is involved in UL certification, then this
makes me very curious about what one does who wishes to sell a
commercial instrument (in the US primarily, but interested in gloal as
well) that is line powered but is also a "one-off."

It would appear totally impractical to UL test in this case, and so
mustn't there be some exemption from requiring a cert. to sell the item
legally? Yet, how does one protect themselves from liability (other
than by designing well so that it *is* as safe as possible to begin with)?

You can make equipment 'safe' in this context by attending to design
issues;

Yes.


Thanks for the input.


Good day!







--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crcarle@sandia.gov

 

[John Woodgate]

I read in sci.electronics.design that Chris Carlen
<crcarle@BOGUS.sandia.gov> wrote (in <c4fsap02g4q@news1.newsguy.com&gt
about 'Grounding a chassis for line voltage safety', on Wed, 31 Mar
2004:

Wowy-dowy! If that is what is involved in UL certification, then this makes me
very curious about what one does who wishes to sell a commercial instrument (in
the US primarily, but interested in gloal as well) that is line powered but is
also a "one-off."

It would appear totally impractical to UL test in this case, and so mustn't
there be some exemption from requiring a cert. to sell the item legally? Yet,
how does one protect themselves from liability (other than by designing well so
that it *is* as safe as possible to begin with)?

This is precisely one reason why Europe has abandoned safety
certification for most products. It is up to the manufacturer to ensure
that his product is safe, and he decides how he does that.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[John Larkin]

On Wed, 31 Mar 2004 17:47:05 -0800, Chris Carlen
<crcarle@BOGUS.sandia.gov> wrote:

John Woodgate wrote:
I read in sci.electronics.design that Walter Harley
walterh@cafewalterNOSPAM.com> wrote (in <c4f4rt$k4q$0@216.39.172.65&gt
about 'Grounding a chassis for line voltage safety', on Wed, 31 Mar
2004:

"Chris Carlen" <crcarle@BOGUS.sandia.gov> wrote in message
news:c4f0060eh3@enews2.newsguy.com...

Yes, I should probably get familiar with a specific cert. agency. I
haven't done so yet since nothing we do here ever gets used other than
internally as custom lab instrumentation.

And is probably one-off construction in many cases? I've heard that UL
certification is not only expensive (>$10k), they require several units to
destructively test. Very impractical for a lab situation, even one with
defense funding, I'd think. Also I'm not sure how long the process takes.

But that's just hearsay, not knowledge. Perhaps someone with real
information can correct me.



It's substantially true. Certification is inappropriate for such one-off
equipment. What is by NO means inappropriate is *making the equipment
safe according to applicable standards*. And that is important for
health and safety at work purposes, which are really rather important.

Wowy-dowy! If that is what is involved in UL certification, then this
makes me very curious about what one does who wishes to sell a
commercial instrument (in the US primarily, but interested in gloal as
well) that is line powered but is also a "one-off."

It would appear totally impractical to UL test in this case, and so
mustn't there be some exemption from requiring a cert. to sell the item
legally? Yet, how does one protect themselves from liability (other
than by designing well so that it *is* as safe as possible to begin with)?

UL is not a law-enforcement agency. Some states have laws that require
consumer items to be UL (or equivalent) certified, but seldom address
scientific or industrial items. UL and FCC both have explicit
exemptions for many classes of "test equipment".

In short, there's no enforcement. Even "CE" is said to mean Can't
Enforce.

My company designs everything to conform to UL and CE safety
standards, but we don't formally test or certify. This is good for the
products and good for the health of our customers.

John