Maker Pro
Maker Pro

Galvanic Isolater

E

engsol

Has anyone worked out the size of diodes and caps
required to do the job? I plan to build my own....can't see
paying $100 plus for 4 diodes and a couple of caps.

The tricky part might be figuring out a way to monitor the
device for blown components. Anyone done that? Ides?

Thanks,
Norm B
 
E

engsol

Good observations. Thanks.
I agree that the jury is out on the value of capacitors.
I might install an isolation xformer later, but the cost is a bit
off-putting with so many other things to buy.
But that's boating, eh?
Norm B
 
M

Manlio Laschena

Followup to msg on Sun, 20 Feb 2005 17:51:50 -0800, engsol
Good observations. Thanks.

Hi,
I hope you don't feel offended if I copy a link where you can read
how the galvanic isolator works.
http://www.yandina.com/galvanicIsolator.htm

Galvanic isolator try to stop ONE of several causes of corrosion,
mainly that using the AC ground wiring to carry galvanic currents from
the harbour network .

From that point of view I get confused by your statements.

Hope it helps

Manlio


Manlio Laschena
s/y Amarose
 
H

halibutslayer

Your said:
As a follow-up, I should point out that this is a somewhat controversial
issue, this business of AC vs. DC currents. Many students of galvanic
corrosion believe that AC can and does cause corrosion. If you choose to
believe this, then you will want an isolation device that limits both AC
and DC voltages.

Mercury uses a capacitor in their isolators because a sufficiently high
AC voltage will forward-bias the diodes and allow them to conduct DC,
even at low voltages. But bypassing the diodes for AC means that the full
AC voltage (if any) will pass through the isolator. The net effect of
this may be a benefit in isolating some DC voltages, but at the expense
of allowing AC to pass through unaffected!

If there is a chance that you will need to protect against AC as well as
DC with your galvanic isolator, you will be disappointed with a
capacitor. The best course then, is to use an isolation transformer. It
will provide complete isolation.

Of course it can pass full AC thats the point. The whole idea is for a short in the
ungrounded
conductor (hot) to have a safe path to ground (hopefully tripping the breaker) The idea
behind
the isolator is to block DC currents caused when your aluminum outdrive and your
neihbors
stainless shaft become a battery because they are both connected electricaly and
emmerced in
an electrolyte. As far as small AC currents caused by leaky 48" shop lights, electric
motors etc.
you want the isolator to pass these. Its much better that these currents travel back
home via the
grounding conductor than through your through hull fitting into the water etc. This is
for the sake
of your boat and swimmers.
 
C

chuck

Hello there halibutslayer,

I think there is a little misunderstanding about what a
galvanic isolator does. If constructed the usual way with
four diodes, the isolator will conduct AC AND DC, except
that until the voltage exceeds about 1.5 volts, it doesn't
conduct at all (simplification of course). Depending on the
design rating of the diodes and heat sink, it may conduct 30
or 50 or 200 amps AC, DC, or both, forever! Now this is an
isolator without any capacitors.

Add a capacitor and what happens? It will still conduct 30
or 50 or 200 amps AC, DC, or both, forever. What changes? It
will now conduct AC without the 1.5 volt threshold kicking
in. But depending on the capacitor chosen, the AC voltage
drop could be even more than 1.5 volts. Could be less. Who
knows? What does it matter? AC voltage can and does vary by
a lot more than a few volts without jeopardizing the safety
of the vessel. So there simply does not seem to be a safety
issue associated with making sure there is not a 1.5 volt AC
threshold, while allowing some other unspecified AC voltage
drop. No difference will be seen in the way fuses and
breakers and GFIs work, with or without a capacitor.

So are there any benefits to not having a capacitor? Maybe.
There is no such thing as a galvanic AC current, I guess,
but there can be an AC electrolytic current. Research
suggests these currents may be even more damaging than DC
currents. So from a corrosion perspective, it would be good
to block them from getting into the boat's green wire. An
isolator without a capacitor would at least block the
lower-voltage AC, but would allow the higher-voltage AC to
pass (once the 1.5 volt threshold was exceeded).

So where is the benefit to adding the capacitor? None that I
can see.

The only ways to fully provide for onboard safety and also
eliminate galvanic and electrolytic currents from traveling
through the green wire are to use an isolation transformer,
or don't bring shore power aboard.

Usually, a simple galvanic isolator is sufficient.


Regards,

Chuck
 
H

halibutslayer

chuck said:
So are there any benefits to not having a capacitor? Maybe.
There is no such thing as a galvanic AC current, I guess,
but there can be an AC electrolytic current. Research
suggests these currents may be even more damaging than DC
currents. So from a corrosion perspective, it would be good
to block them from getting into the boat's green wire. An
isolator without a capacitor would at least block the
lower-voltage AC, but would allow the higher-voltage AC to
pass (once the 1.5 volt threshold was exceeded).

So where is the benefit to adding the capacitor? None that I
can see.

The only ways to fully provide for onboard safety and also
eliminate galvanic and electrolytic currents from traveling
through the green wire are to use an isolation transformer,
or don't bring shore power aboard.

Usually, a simple galvanic isolator is sufficient.

Regards,

Chuck

The AC currents are currents are all ready in your green wire from leaky
equipment in your boat. These currents are making a complete circuit to
where the grounding wire is connected to neutral on shore. The question
is whether or not this happens through the grounding wire via the
capacitor in the isolator or through a metal fixture on your boat,
through the water, to ground, grounding rod, and then to neautral.
The down side is if you have a capacitor and your neigbor doesn't than
her AC currents might use your underwater parts and capacitor in your
isolator as the shortest path to the neutral/ground connection point on
shore. This is why ALL isolators should have a capacitor and I think
ABYC may require one. If they don't they should. Or you could just spend
the money and valuable space for an isolation transformer and not worry
about it.

Hope my point is better made than in the previous posting.

Eric
 
C

chuck

Hello Eric,

Thanks for clarifying.

I believe you are correct. You are talking about
high-resistance onboard leakages that generate currents too
small to be detected by the GFI circuit or the breakers. The
isolator diodes would probably not conduct under those
circumstances and a capacitor would help.

UL requires the GFI to trip at a 5 ma unbalance, so 24,000
ohms of leakage would trip it. Actually, the isolator diodes
would probably pass 5 ma in that circuit without a
capacitor. The capacitor would be necessary when the leakage
resistance was in the megohms and the currents in the
microamps.

Would rather not have that stuff flowing through my ground
connections through the water to adjacent boats, even at
those low current levels. This underscores the importance of
making sure you don't have dangerous leakages onboard in the
first place. Easy enough to check, but how many regularly
test their GFIs?

We sure agree on the isolation transformer, too.

Thanks again, Eric.

Chuck
 
J

John Proctor

Hello Eric,

Thanks for clarifying.

I believe you are correct. You are talking about high-resistance
onboard leakages that generate currents too small to be detected by the
GFI circuit or the breakers. The isolator diodes would probably not
conduct under those circumstances and a capacitor would help.

UL requires the GFI to trip at a 5 ma unbalance, so 24,000 ohms of
leakage would trip it. Actually, the isolator diodes would probably
pass 5 ma in that circuit without a capacitor. The capacitor would be
necessary when the leakage resistance was in the megohms and the
currents in the microamps.

Would rather not have that stuff flowing through my ground connections
through the water to adjacent boats, even at those low current levels.
This underscores the importance of making sure you don't have dangerous
leakages onboard in the first place. Easy enough to check, but how many
regularly test their GFIs?

We sure agree on the isolation transformer, too.

Thanks again, Eric.

Chuck

Chuck,

I am puzzled by your UL 5 ma rating.

Here in Australia we have basically two ratings. The first is for most
domestic installations an is 30 ma. The second is for hazerdous
locations including hospitals where the rating is 10 ma. There are
other higher rating RCD (Residual Current Devices) which are used in
industrial contexts (eg 60 ma) but the majority of installations are
either 30 ma. or 10 ma.

BTW I am an EE as well as a licensed Electrical Contractor.
 
S

Sir Spamalot

Chuck,

I am puzzled by your UL 5 ma rating.

Here in Australia we have basically two ratings. The first is for most
domestic installations an is 30 ma. The second is for hazerdous
locations including hospitals where the rating is 10 ma. There are
other higher rating RCD (Residual Current Devices) which are used in
industrial contexts (eg 60 ma) but the majority of installations are
either 30 ma. or 10 ma.

BTW I am an EE as well as a licensed Electrical Contractor.


Chuck,

Now **I** am puzzled by the Australian ratings, especially since my
sister lives "across the pond" in NZ. 30ma has been known to kill,
hence I question the 30ma "domestic installation" rating. And you
guys are on 240V, correct?

Are we discussing semantics here? In the US, a GFI (120V) is supposed
to protect the unknowing public from faults that could channel more
than 5ma through the body. UL/CSA/others test to that 5ma
specification.

SS


PS: Me too on the EE business.
 
C

chuck

Hello John,

Thanks for the info. My first reaction is to suggest that we
use 5 ma because we Yanks are a less hardy lot and need that
extra protection.

But if you look at the link:

http://hypertextbook.com/facts/2000/JackHsu.shtml
Electric Current Needed to Kill a Human

you'll see that 60 - 70 ma. may be a lethal dose.

My guess is that the 5 ma was chosen to be below the general
human threshold of sensation. But I have no idea why
Australia has chosen so high a threshold.

And did I read somewhere that in Europe the residential
distribution lines are nowhere grounded? A two-wire system,
I seem to recall. You don't have that kind of system in
Australia, do you John?

Regards,

Chuck
 
J

John Proctor

Chuck,

Now **I** am puzzled by the Australian ratings, especially since my
sister lives "across the pond" in NZ. 30ma has been known to kill,
hence I question the 30ma "domestic installation" rating. And you
guys are on 240V, correct?

Are we discussing semantics here? In the US, a GFI (120V) is supposed
to protect the unknowing public from faults that could channel more
than 5ma through the body. UL/CSA/others test to that 5ma
specification.

SS


PS: Me too on the EE business.

Curiouser and curiouser...

In Australia we have generally what is called a multiply earthed
neutral system (MEN). There are other types of distributution systems
used in remote areas but MEN system is found in all urban areas. In
this system n+1 service conductors (where n = number of phases) are
brought into the consumer's premises although if available on the
distributiuon system 3 phase power is much easier to obtain here than
in NA (I originally came from Canada). This phase active and netral in
a single phase supply is the normal residential. The neutral is boded
to earth at each consumer switchboard and the earth is bonded to a
ground system which is normally a grounding rod of reasonable size.
Some allowances are made for installations where a ground rod is not
practical or where it cannot be of sufficient size. It has not been
acceptable to ground to utility services (gas or water) for some number
of years due to lack of guaranteed continuity (gas reticulation is via
plastic pipes and water meters har generally non metalic internally).

I have generally seen in the literature that 60 ma is generally
considered the level at which statistically fibrillation is almost
guaranteed when the current passes through the chest cavity. Some of
the references given seem to indicate much lower levels. However when
standards are set they must take into account the statistical nature of
some of the effects they are trying to deal with. Hence I think the 30
ma residential rating here in Australia. The AS/NZS 3000:2000 Wiring
Rules specifies an maximum RCD rating of 30 ma for residential circuits
and I don't believe there are any addendums which mandate a lower
maximum. As I indicated in my previous post the 10 ma level is required
in hospitals or other high risk application areas. However you need to
balance emi filtering requirements with leakage levels so that you
don't get nuisance tripping.

I don't know about New Zealand but given the nature of trans-Tasman
standards activities I would be surprised if their wiring rules were
significantly different from ours. After all we use the same wiring
rules!

In residential installations an RCD (residual current device) is
required on general purpose outlets and lighting circuits but does not
need to be applied to direct connection appliances like cooktops or
stoves. However practice has evolved so that 'safety switches' (RCDs)
are placed on the whole installation (for older installations) as these
are easier to wire into old and cramped switchboards. New installations
segregate the power, lighting and direct connected appliances and have
an RCD on the appropriate circuits. Some contractors also will provide
unprotected circuits for essential circuits (refrigerators, freezers
etc.) to prevent food spoilage from nuisance tripping where the outlet
is set up for fixed use.

In the case of marinas there is usually a 16A branch circuit protected
by a 30 ma RCD supplying shore power to the vessel. Unfortuanately, I
don't have any experience with larger vessels where 3 phase power from
shore is delivered but I would expect the situaltion to be similar with
n+1+1 conductors being delivered to the boat (where n = number of
phases and the extra +1 is the earth). It is mandatory that earth and
neutral be kept separate on the vessel. Installations where there are
changeover facilities to a genset require the earth and neutral to be
bonded only when the genset is the power source.

Sorry about the length but that is generally how things work downunder.
 
Top