Galvanic Isolators

[William Martin]

OK - Galvanic corrosion happens because of a tiny trickle of current over the earth wire. The earth wire has no purpose other than to drain away any big accidental spillages of current. A galvanic isolator disconnects the earth wire but hovers ready to reconnect it if a spillage takes place - which should happen hardly ever, even never. So a GI is an excellent solution to corrosion; so I bought one.

 

 

But!!! - now I hear; certain modern electrical appliances use a process called 'switched mode' which involves regular dumping of current down the earth wire as a normal part of their operation - so the GI now stays in a constant state of connection thus allowing my boat to disolve away.

 

 

 

Have I understood this correctly? And what are these devices that use 'switched mode' anyway?

[twoputtyrats]

OK - Galvanic corrosion happens because of a tiny trickle of current over the earth wire. The earth wire has no purpose other than to drain away any big accidental spillages of current. A galvanic isolator disconnects the earth wire but hovers ready to reconnect it if a spillage takes place - which should happen hardly ever, even never. So a GI is an excellent solution to corrosion; so I bought one.

But!!! - now I hear; certain modern electrical appliances use a process called 'switched mode' which involves regular dumping of current down the earth wire as a normal part of their operation - so the GI now stays in a constant state of connection thus allowing my boat to disolve away.

Have I understood this correctly? And what are these devices that use 'switched mode' anyway?

 

 

Desktop computers for one thing, laptops as well & televisions and lots of electronic devices such as phone chargers etc. I've a meter on my GI that shows when it's being effective, the laptop doesn't seem to trouble it.

 

From what I understand The more you have the worse it gets.....

 

Simes.

[blackrose]

OK - Galvanic corrosion happens because of a tiny trickle of current over the earth wire. The earth wire has no purpose other than to drain away any big accidental spillages of current. A galvanic isolator disconnects the earth wire but hovers ready to reconnect it if a spillage takes place - which should happen hardly ever, even never. So a GI is an excellent solution to corrosion; so I bought one.

But!!! - now I hear; certain modern electrical appliances use a process called 'switched mode' which involves regular dumping of current down the earth wire as a normal part of their operation - so the GI now stays in a constant state of connection thus allowing my boat to disolve away.

Have I understood this correctly? And what are these devices that use 'switched mode' anyway?

 

Yes, you've understood it perfectly. I thought we were on another thread discussing (even arguing!) about it.

 

If you are really worried about it buy an isolation transformer, but be prepared to pay large and more to have it fitted.

 

I have a galvanic isolator with a status monitor located in the cabin to let me know if the diodes are in circuit (and an earth path created). For this to happen the earth fault has to be at least 2.4v which I understand represents a lot of switched mode equipment.

 

Were this to happen I would switch off my tv, computer, etc to see if this was the cause, and if not I'd investigate further.

 

The two main drawbacks of a galvanic isolator compared to an isolation transformer are 1) It needs to be tested regularly. I use a 9v battery to first test that the status monitor is working, and then to test the diodes are working. (The status monitor lights dim as 9v from the battery is reduced to 2.4v due to the clamping action of the diodes).

 

2) More seriously, there is the possibility that the diodes in the galvanic isolator might fail in disconnected mode, leaving no path to earth. However I have been told (albeit by the manufacturer), that if diodes in a galvanic isolator fail which is rare, it's usually when they're in circuit, meaning there will be a path to earth but your boat will not be protected from galvanic action. That's no worse than a boat without a galvanic isolator and if you test it regularly you'd soon discover this so it wouldn't be a problem.

 

If you're not living on your boat & constantly plugged in to shore power then you don't need to bother with a galvanic isolator or an isolation transformer.

Edited November 8, 2006 by blackrose

[William Martin]

Great - thanks

 

Any idea what a Transformer costs?

[bottle]

Price list for Victron Isolation transformers Victron pdf

 

Second look I think the prices are given in Euros ex VAT.

Edited November 8, 2006 by bottle

[GUMPY]

Great - thanks

 

Any idea what a Transformer costs?

 

Blatent plug time

Declares finacial interest

 

There will be an insulated one on the market soon for about £275 cant say more than that.

 

beware the victron one as its wired incorrectly and could cause the boat to go live....

 

see here for details

 

http://www.smartgauge.co.uk/iso_wire.html

 

Julian

[smileypete]

But!!! - now I hear; certain modern electrical appliances use a process called 'switched mode' which involves regular dumping of current down the earth wire as a normal part of their operation - so the GI now stays in a constant state of connection thus allowing my boat to disolve away.

The switching current in the earth connection are very very brief in duration, in no way is it constant!

 

Also even if the GI is biased into conduction when this happens, the current involved is not large. Averaged out over time it's almost certainly vanishingly small.

 

Have I understood this correctly?

Not quite, see above...

 

And what are these devices that use 'switched mode' anyway?

Anything switches of the mains current as part of it's operation, dimmers, computer power supplies, some battery chargers.

 

I posted about this on another thread here.

 

As I said in the above post, I'd like to see some real world measurements in order to decide if there really is a problem or not.

 

cheers,

Pete.

[Gibbo]

Desktop computers for one thing, laptops as well & televisions and lots of electronic devices such as phone chargers etc. I've a meter on my GI that shows when it's being effective, the laptop doesn't seem to trouble it.

 

From what I understand The more you have the worse it gets.....

 

Simes.

 

Well at last I am able to post again.

 

A quick word of warning.

 

The RFI (radio frequency interference) that can cause GIs to conduct thus rendering them useless is, of course, AC. It wouldn't be RFI if it wasn't AC.

 

The meters fitted to GIs to supposedly show when they are doing their job will not register this. They will show an average of the RFI which is, of course, zero. So your GI could be hard into conduction due to RFI and the meter will show precislely nothing giving the impression that all is fine.

 

Gibbo

[William Martin]

So; now the switching mode devices open the gate for mere miliseconds so there is nothing to worry about? The indicator LEDs are worse than useless because the cause a false sense of security?

 

 

 

Interesting - I started this thread to try and distill all the myriad of opinion down into simple facts - but it has become clear as mud again - methinks there is no real answer!

 

Is anyone on here a proper electronics boffin (qualified) who can answer this definitively?

Edited November 8, 2006 by William Martin

[jollyroger]

Just done a check, The TV, Sky Box, DVD Player, Laptop, Phone Charger, Sander, Drill and electric Razor are all Double Insulated, that means they don't have an earth wire, so they will have no effect on a GI.

[GUMPY]

Interesting - I started this thread to try and distill all the myriad of opinion down into simple facts - but it has become clear as mud again - methinks there is no real answer!

 

Is anyone on here a proper electronics boffin (qualified) who can answer this definitively?

 

I would say that Gibbo just did, and he is more qualified to answer that question than anyone I know including me.

 

GIs will switch on because of the RFI and the indicators wont tell you when they have.

 

Its simple if you want a total soloution get an isolating transformer that has the screen grounded to mains earth and NOT the way that Victron do it.

 

 

julian

[John Orentas]

Is anyone on here a proper electronics boffin (qualified) who can answer this definitively?

 

 

 

I am certainly not one, but I do know that you Must not put any components in series with an earth bonding cable. That is exactly what these things do.

[NB Willawaw]

http://www.canalworld.net/forums/index.php...amp;hl=galvanic

 

A lot of modern consumer electronics uses a switched mode power supply.

 

If the appliance has an earth wire (many metal cased items like washing machines, microwaves etc), then this is an IEC Class I device.

If the appliance is a IEC Class II device, it will be double insulated and will not have an earth wire to dump RFI to. This is normally indicated by a square within a square logo or icon on the casing - this covers most plastic cased appliances like drills and vacs.

 

If you have class I devices, they can produce RFI which if of sufficient magntitude, can open the diodes in the GFI/Galvanic Isolator. How often this happens and whether it will contribute significantly to stray earth currents is debatable. I don't think anybody can give you a cast iron guarantee either way. How often will you use these appliances onboard; how often is the boat sitting on the dock with mains connected on a shoreline and just the battery charger running ?? is the battery charger double insulated ?

 

I've seen boats with a Galvanic Isolator from new, that still had a pitted hull.

 

If you have any doubt, buy an isolation transformer.

 

OK - Galvanic corrosion happens because of a tiny trickle of current over the earth wire. The earth wire has no purpose other than to drain away any big accidental spillages of current. A galvanic isolator disconnects the earth wire but hovers ready to reconnect it if a spillage takes place - which should happen hardly ever, even never. So a GI is an excellent solution to corrosion; so I bought one.

But!!! - now I hear; certain modern electrical appliances use a process called 'switched mode' which involves regular dumping of current down the earth wire as a normal part of their operation - so the GI now stays in a constant state of connection thus allowing my boat to disolve away.

Have I understood this correctly? And what are these devices that use 'switched mode' anyway?

Edited November 8, 2006 by NB Willawaw

[blackrose]

Well at last I am able to post again.

 

A quick word of warning.

 

The RFI (radio frequency interference) that can cause GIs to conduct thus rendering them useless is, of course, AC. It wouldn't be RFI if it wasn't AC.

 

The meters fitted to GIs to supposedly show when they are doing their job will not register this. They will show an average of the RFI which is, of course, zero. So your GI could be hard into conduction due to RFI and the meter will show precislely nothing giving the impression that all is fine.

 

Gibbo

 

Can you explain this to me please Gibbo? I still don't understand.

 

The led indicators on my GI come on when the diodes are in circuit. How is the RFI going to do this without illuminating the leds?

Edited November 8, 2006 by blackrose

[Gibbo]

Can you explain this to me please Gibbo? I still don't understand.

 

The led indicators on my GI come on when the diodes are in circuit. How is the RFI going to do this without illuminating the leds?

 

No they don't. They indicate the voltage across the GI.

 

Ok Let's split the GIs into two types.

 

There are those that use an analogue meter with something like "safe" and "Not safe" written on them. These measure the DC voltage across the isolator. You could pass 20 amps of AC through one of these things (which of course would be clipped at the clamping voltage of the GI) and the meter would show exactly zero because the DC voltage of an AC signal is the average, which is zero. But the GI would still be permanently conducting (well 99% or so of the time depending on the AC waveshape an frequency etc).

 

The other types use LEDs. These can be made in various different ways. Some of them would show AC conduction of the GI. Others would not. Without seeing the circuit for the indicator (or actually trying the experiment) it would not be possible to know.

 

Gibbo

 

http://www.canalworld.net/forums/index.php...amp;hl=galvanic

 

A lot of modern consumer electronics uses a switched mode power supply.

 

If the appliance has an earth wire (many metal cased items like washing machines, microwaves etc), then this is an IEC Class I device.

If the appliance is a IEC Class II device, it will be double insulated and will not have an earth wire to dump RFI to. This is normally indicated by a square within a square logo or icon on the casing - this covers most plastic cased appliances like drills and vacs.

 

If you have class I devices, they can produce RFI which if of sufficient magntitude, can open the diodes in the GFI/Galvanic Isolator. How often this happens and whether it will contribute significantly to stray earth currents is debatable. I don't think anybody can give you a cast iron guarantee either way. How often will you use these appliances onboard; how often is the boat sitting on the dock with mains connected on a shoreline and just the battery charger running ?? is the battery charger double insulated ?

 

I've seen boats with a Galvanic Isolator from new, that still had a pitted hull.

 

If you have any doubt, buy an isolation transformer.

 

If you want an idea of how quickly your GI will be rendered useless simply disconnect the earth lead from a class I type apparatus and measure the voltage between neutral and the now disconnected earth on the equipment. You will almost certainly find the earth at around half mains potential. i.e. around 115 volts. That will *certainly* switch on a GI!

 

I don't recommend anyone does this unless they have the knowledge and experience to avoid killing themselves.

 

You're absolutely right, if you have any doubt, use an isolation transformer.

 

Gibbo

 

 

 

Is anyone on here a proper electronics boffin (qualified) who can answer this definitively?

 

Me. And I think I have answered it.

 

Gibbo

[Callum]

OK, I understand that you could have sizeable AC current whizzing back and forth across the GI diodes and that itself would not register on a dc meter type monitor. Now the next is hypothesis that I'd appreciate confirmation or correction of (purely as an academic exercise, btw );

 

Since both sets of diodes are alternatively conductive there is in fact always a dc connection available between boat & shore. Therefore any dc galvanic current could be flowing through the boat & the GI, i.e. dc superimposed on the ac current, causing corrosion. If a net dc current is passing through the GI, is the GI's meter not sensitive enough to detect it (no idea what ampage a galvanic current would be!)?

[blackrose]

.

 

 

 

Ok Let's split the GIs into two types.

 

There are those that use an analogue meter with something like "safe" and "Not safe" written on them. These measure the DC voltage across the isolator. You could pass 20 amps of AC through one of these things (which of course would be clipped at the clamping voltage of the GI) and the meter would show exactly zero because the DC voltage of an AC signal is the average, which is zero. But the GI would still be permanently conducting (well 99% or so of the time depending on the AC waveshape an frequency etc).

 

The other types use LEDs. These can be made in various different ways. Some of them would show AC conduction of the GI. Others would not. Without seeing the circuit for the indicator (or actually trying the experiment) it would not be possible to know.

 

Gibbo

 

Mine claims to show AC conduction by means of 2 leds.

Edited November 8, 2006 by blackrose

[Gibbo]

OK, I understand that you could have sizeable AC current whizzing back and forth across the GI diodes and that itself would not register on a dc meter type monitor. Now the next is hypothesis that I'd appreciate confirmation or correction of (purely as an academic exercise, btw );

 

Since both sets of diodes are alternatively conductive there is in fact always a dc connection available between boat & shore. Therefore any dc galvanic current could be flowing through the boat & the GI, i.e. dc superimposed on the ac current, causing corrosion. If a net dc current is passing through the GI, is the GI's meter not sensitive enough to detect it (no idea what ampage a galvanic current would be!)?

 

You're halfway to answering your own question. In fact I think you actually know the answer

 

Assume the galvanic cell made by your hull and the quayside produces 0.3 volts. With no AC messing up the GI this will register on the GI meter as 0.3 volts. The GI meter shows the voltage across the GI. It does *not* show the current. But the diodes in the GI will not be conducting so no current flows, therefore no galvanic erosion. With a direct connection (i.e. no GI) this will produce a current that depends upon the resistance of the complete circuit, i.e. from hull through earth bond, back the neutral/earth bond (possibly at substation), back through the ground to the quayside. It is the current that produces the erosion.

 

So with a GI there would be no erosion, as there is no current flow, as the galvanic cell made by the hull, quayside and water doesn't produce enough voltage to "breakdown" the GI.

 

The meter reads the nett DC voltage of 0.3 volts.

 

Now if there is sufficient AC voltage (be it 50Hz or RFI or whatever) so that the GI is forced into almost permanent conduction, the nett DC voltage across the GI remains at 0.3 volts. But as you quite rightly state this is now superimposed on the AC signal. And as the GI is now conducting (due to the AC), the superimposed DC voltage causes a current to flow, which is the galvanic corrosion current. The meter continues to show 0.3 volts which is correct. But there is now a current flowing at DC.

 

So, although you didn't actually say it in your post, yes, you are correct

 

Gibbo

 

 

 

.

Mine claims to show AC conduction by means of 2 leds.

 

That's perfectly feasable and possible. If it does indeed do so then you can rely on them with confidence. But I seriously would have my doubts without making tests and measurements or seeing the schematic.

 

Gibbo

[tomsk]

You're halfway to answering your own question. In fact I think you actually know the answer

Assume the galvanic cell made by your hull and the quayside produces 0.3 volts. With no AC messing up the GI this will register on the GI meter as 0.3 volts. The GI meter shows the voltage across the GI. It does *not* show the current. But the diodes in the GI will not be conducting so no current flows, therefore no galvanic erosion. With a direct connection (i.e. no GI) this will produce a current that depends upon the resistance of the complete circuit, i.e. from hull through earth bond, back the neutral/earth bond (possibly at substation), back through the ground to the quayside. It is the current that produces the erosion.

 

So with a GI there would be no erosion, as there is no current flow, as the galvanic cell made by the hull, quayside and water doesn't produce enough voltage to "breakdown" the GI.

 

The meter reads the nett DC voltage of 0.3 volts.

 

Now if there is sufficient AC voltage (be it 50Hz or RFI or whatever) so that the GI is forced into almost permanent conduction, the nett DC voltage across the GI remains at 0.3 volts. But as you quite rightly state this is now superimposed on the AC signal. And as the GI is now conducting (due to the AC), the superimposed DC voltage causes a current to flow, which is the galvanic corrosion current. The meter continues to show 0.3 volts which is correct. But there is now a current flowing at DC.

 

So, although you didn't actually say it in your post, yes, you are correct

 

Gibbo

 

This is a speciality of Rosies

[Gibbo]

PS. I'd like to make it quite clear that I do not sell galvanic isolators. I do not sell isolation transformers. I do not make either. I do not install either. I am in no way connected with either equipment or their distributors, agents, dealers etc etc etc. I have nothing to gain by whichever method is best.

 

This whole subject first came to light to me because my company was contracted by a.n.other to carry out a totally independant study of GIs with regard to modern switchmode power supplies. The results surprised me to say the least. So much so that I now wouldn't touch one with a barge pole.

 

Gibbo

[blackrose]

This is a speciality of Rosies

 

That wasn't a response to my question. Do try to keep up dear boy!

 

That's perfectly feasable and possible. If it does indeed do so then you can rely on them with confidence. But I seriously would have my doubts without making tests and measurements or seeing the schematic.

 

Gibbo

 

Thanks, for the info, I will ask if the manufacturer if he will send me the schematic. May I pass it onto you if he does?

Edited November 8, 2006 by blackrose

[tomsk]

That wasn't a response to my question. Do try to keep up dear boy!

 

Cool.. .. Couldn't help myself...All this Cragdale celebaration goings on has put me in a particualy frivolous mood.....Have a beeeeerrr

[blackrose]

Cool.. .. Couldn't help myself...All this Cragdale celebaration goings on has put me in a particualy frivolous mood.....Have a beeeeerrr

 

Has it been recovered?

[tomsk]

Has it been recovered?

 

Are you seriously telling me you havn't checked that thread today...... I know I'm a gullable Womble, but really....... Have another Beeeer

Edited November 8, 2006 by tomsk

[Gibbo]

That wasn't a response to my question. Do try to keep up dear boy!

Thanks, for the info, I will ask if the manufacturer if he will send me the schematic. May I pass it onto you if he does?

 

I'd be interested to see it. Just out of curiosity. And of course I can let you know whether what they say is correct (which I can only assume it is in these days of litigation for everything).

 

But I doubt they'll give it to you! Thing is, unless the monitoring circuit has an external power supply, it's actually far from a trivial task to light an led from 0.8 volts. It can be done, obviously, but it's nowhere near as simple as it at first appears.

 

Gibbo