Alex_Blackwood
Well-known member
Quite!
DIY Galvanic isolator.
- Thread starter Vara
- Start date
Quite!
bedouin
Well-known member
I suppose that must be related to the issue of whether you connect your mains earth to your DC negative bonding.
I know that it is recommended that you do, but I am not sure of the arguments and whether having the GI in the circuit would impact those.
misterg
Active member
Firstly, can one keep the DC bonding separated from the 12V -ve? It presumably means an insulated prop-shaft coupling, etc.
If one can't achieve this, then we get the same scheme that VicS started a new thread on.
As regards fault currents, reliability, etc. it takes the GI out of the equation if a 'live'-'earth' fault develops within the mains wiring, or within something plugged into a mains socket. The GI still needs to do its stuff if a fault develops between the mains wiring and the ironmongery on the boat - *I* would have thought that the latter would be more likely (trapped or chaffed wiring; dropping the fan heater on the engine, etc...).
Success of this scheme also depends on the behaviour of any permanently installed battery charger:
This might connect DC -ve to mains earth by design or by accident (faulty, chafed wire, etc.) in which case the GI between the mains earth and the DC system is bypassed.
There's an interesting article by DEI (who make GIs) here (pdf file) written in 2006 when the requirement for GIs to contain status monitoring was introduced to the ABYC standard:
"When A-28 was {originally} written it did not adopt the long-standing, previously established criteria for an effective ground fault current path from the {U.S National Electrical Code (NFPA 70)}; hence, a marine galvanic isolator per the existing A-28 standard is not required to maintain continuity of the grounding conductor of the shore power cord - a potentially hazardous condition. To compensate for this concern, A-28 added a requirement for monitoring the galvanic isolator and to provide an alarm if the galvanic isolator failed open-circuit. DEI believes that this is a poor substitute for a galvanic isolator that meets the requirements of the previously referenced codes that assure safety grounding. The very event that would cause a galvanic isolator to fail open may well be the event that creates the potentially hazardous condition when grounding is lost, and which may also damage the monitor that is to provide the alarm. Even when the monitor alarm works properly, a potential hazard exists until the alarm is noted and the problem corrected. The most reliable galvanic isolator design is one that meets the code criteria referenced (i.e., assured grounding continuity at all times) and is not encumbered with the electronic components of a monitor that are much more subject to failure...
I note that the current version of ABYC A28 introduces a class of 'fail safe' GIs that are exempted from the requirement for integral monitoring. It also gives the test criteria for these devices, and it is in the 1000s of amps - the principle being that the earth lead melts before the GI fails.
A-28 Galvanic Isolators Standard (pdf file).
I note that these documents are US origin, where I believe they don't commonly use RCDs, but the fault current considerations are (IMHO) at least broadly valid in the UK. I would like to know what the equivalent UK standard on the effectiveness of earthing arrangements had to say - I guess it's part of the IEE regs...?
Interesting discussion, thanks.
Andy
halcyon
Well-known member
To try and explain earlier post, the problem is props and skin fitting corrosion due to galvanic action, mainly from shore based mains AC.
In the past, assuming prop, skin fittings etc, are bonded, then the GI was fitted between mains earth / DC neg and the bonding system. This protected against volt drop in shore side mains cabling, or any AC leakage to DC, passing to the bonding system. Unless more than 2.5 volt ? ( block 1.2 volt, conduct 2.2 or 2.5 volt years since I read the spec now ), thus allowing AC fault current a earth path.
In recent years the GI is being fitted in the incoming mains earth cable. This means AC leakage after GI ( quite possible )are allowed to pass to the bonding system and skin fittings, as long as the current is below the RCD trip current.
Thus with the new set-up we are defeating the reason for fitting the GI.
It is that obvious that it can't be right, there must be a very logical and obvious reason for fitting it there.
I'm looking at to deeply I think.
Must get back to the new website building, leave you all the question.
Brian
In the past, assuming prop, skin fittings etc, are bonded, then the GI was fitted between mains earth / DC neg and the bonding system. This protected against volt drop in shore side mains cabling, or any AC leakage to DC, passing to the bonding system. Unless more than 2.5 volt ? ( block 1.2 volt, conduct 2.2 or 2.5 volt years since I read the spec now ), thus allowing AC fault current a earth path.
In recent years the GI is being fitted in the incoming mains earth cable. This means AC leakage after GI ( quite possible )are allowed to pass to the bonding system and skin fittings, as long as the current is below the RCD trip current.
Thus with the new set-up we are defeating the reason for fitting the GI.
It is that obvious that it can't be right, there must be a very logical and obvious reason for fitting it there.
I'm looking at to deeply I think.
Must get back to the new website building, leave you all the question.
Brian
misterg
Active member
I looked at the PBO article again this morning (as part of my daily 3S routine 
)and to be fair to the original author, it does say, in very small letters in the annotation on the circuit diagram that each bridge must have a continuous rating of at least 50 amps - One would expect these to have an adequate surge rating, but I haven't done any checking.
Brian - the problem is differences currents flowing due to the difference in potential between the local 'ship's earth' (by virtue of the metal bits swimming in the sea) and the 'earth' supplied by the shore power wire (as opposed to dealing with leakage to earth from on-board 240v stuff, which could still be an issue).
Andy
)and to be fair to the original author, it does say, in very small letters in the annotation on the circuit diagram that each bridge must have a continuous rating of at least 50 amps - One would expect these to have an adequate surge rating, but I haven't done any checking.Brian - the problem is differences currents flowing due to the difference in potential between the local 'ship's earth' (by virtue of the metal bits swimming in the sea) and the 'earth' supplied by the shore power wire (as opposed to dealing with leakage to earth from on-board 240v stuff, which could still be an issue).
Andy
nigelmercier
RIP
- Joined
- 20 Jun 2007
- Messages
- 16,234
- Location
- Live in Kent, boat in Canary Islands
For the record, the device in the article is NOT using underpowered diodes in parallel. It is simply using one half of a bridge rectifier rated at the correct current, with the added bonus of a second pair as a failsafe. The electonics are 100% sound.
However, there is a serious error in the article. It states that the metal box should be grounded to the dockside cable. On no account should this be done, as the box would then short out the diodes if in contact with any metalwork on the boat. It should be grounded to the output side.
Kev
New member
Hi all,
I have missed this article by David Berry about making a DIY galvanic isolator. Is there a way to read and/or order it online for non subscribers ?
Many thanks, Kev
I have missed this article by David Berry about making a DIY galvanic isolator. Is there a way to read and/or order it online for non subscribers ?
Many thanks, Kev
Matthew Millard
New member
Galvanic Isolation
Hello all
I've been reading the thread on PBO's galvanic isolator and felt compelled to join the forum.
Im an electrician and a boater of 15 years.
NICEIC registered 17th edition etc etc
And employed by a leading German manufacturer of marine diesel heaters and aircon equipment. There are only tow to choose from so you decide!
Preamble over.
I will get to the point and make it simple, forget Galvanic Isolators and fit an isolation transformer.
I see many GI's on ebay and on the web that rate there internal components, (diodes, bridge rectifiers) at 100Amps.
Ok fair enough you might say as normally as shore power connection on the pontoon is 16A.
So for the average joe with no electrical knowledge could think the 100Amp diodes or rectifiers should be man enough to withstand the alledged maximum current from the pontoon bollard of 16Amps.
Thing is though under fault conditions i.e. a phase to earth dead short the current flowing with be a little bit greater than 16Amps. It could be up to a few thousand amps depending on the time taken for the protective devices to cut the power. And not forgetting the PSCC value at the pontoon.
PSCC is the Prospective Short Circuit Current, The PSCC will be around 4kA, (4000Amps).
The PSCC could be greater depending on the nature of the supply!
The RCD must trip in less than 40ms.
And as all electricians know... nobody tests there RCDs on a regular basis.
If you have a look at your consumer unit at home, if its a modern one and in good order you will see a label that tells you to press the 'T' Test button quarterly to test the operation of the device. Now then, who bothers to test the RCD fitted to the pontoon bollard??? Answer not many people at all.
So, with an RCD on the pontoon and on the boat, (if fitted!) thats not been tested for a while could well and truly goose you and your boat with around 4000Amps.
The trip time as stated is <40ms, for every millisecond, or second above this the current that could flow increases lograthmically upto if unchecked the PSCC value.
At this point the GI is fried, and theres no way of knowing as most GI's are encapsulated in a box, the only way to be sure is to test the device with a multimeter. Otherwise after a trip / incident the boat could be without any earth protective if the GI is open circuit.
The 100Amp GI no longer looks attractive!
Its a no brainer, forget the GI's and fit an isolation transformer.
You can replace anodes and props but sadly not people.
Hello all
I've been reading the thread on PBO's galvanic isolator and felt compelled to join the forum.
Im an electrician and a boater of 15 years.
NICEIC registered 17th edition etc etc
And employed by a leading German manufacturer of marine diesel heaters and aircon equipment. There are only tow to choose from so you decide!
Preamble over.
I will get to the point and make it simple, forget Galvanic Isolators and fit an isolation transformer.
I see many GI's on ebay and on the web that rate there internal components, (diodes, bridge rectifiers) at 100Amps.
Ok fair enough you might say as normally as shore power connection on the pontoon is 16A.
So for the average joe with no electrical knowledge could think the 100Amp diodes or rectifiers should be man enough to withstand the alledged maximum current from the pontoon bollard of 16Amps.
Thing is though under fault conditions i.e. a phase to earth dead short the current flowing with be a little bit greater than 16Amps. It could be up to a few thousand amps depending on the time taken for the protective devices to cut the power. And not forgetting the PSCC value at the pontoon.
PSCC is the Prospective Short Circuit Current, The PSCC will be around 4kA, (4000Amps).
The PSCC could be greater depending on the nature of the supply!
The RCD must trip in less than 40ms.
And as all electricians know... nobody tests there RCDs on a regular basis.
If you have a look at your consumer unit at home, if its a modern one and in good order you will see a label that tells you to press the 'T' Test button quarterly to test the operation of the device. Now then, who bothers to test the RCD fitted to the pontoon bollard??? Answer not many people at all.
So, with an RCD on the pontoon and on the boat, (if fitted!) thats not been tested for a while could well and truly goose you and your boat with around 4000Amps.
The trip time as stated is <40ms, for every millisecond, or second above this the current that could flow increases lograthmically upto if unchecked the PSCC value.
At this point the GI is fried, and theres no way of knowing as most GI's are encapsulated in a box, the only way to be sure is to test the device with a multimeter. Otherwise after a trip / incident the boat could be without any earth protective if the GI is open circuit.
The 100Amp GI no longer looks attractive!
Its a no brainer, forget the GI's and fit an isolation transformer.
You can replace anodes and props but sadly not people.
VicS
Well-known member
It is a requirement of iso13297 that:
Does that not cater for your 4000amp fault current ?
Galvanic isolators shall be designed to
withstand the application of power from a short-circuit test from a source capable of delivering 5 000 A r.m.s.
symmetrically to its output test terminals for the time required for the circuit-breaker in the test circuit to trip. After
three applications of the short-circuit test, the electrical and mechanical characteristics of the isolator shall be
unchanged.
withstand the application of power from a short-circuit test from a source capable of delivering 5 000 A r.m.s.
symmetrically to its output test terminals for the time required for the circuit-breaker in the test circuit to trip. After
three applications of the short-circuit test, the electrical and mechanical characteristics of the isolator shall be
unchanged.
Does that not cater for your 4000amp fault current ?
lw395
Well-known member
The single sine rating of a diode with a 15A continuous rating is typically pretty big.
However, being one who distrusts delicate mechanical mechanisms in the same box as salt water, I think there is a case for having some fast blow fuses too.
RCD's do need frequent test. You cannot emphasise that enough. You are unlikely to wear it out and you might stop it from seizing on.
I'm not sure an isloation transformer helps with earth potential issues.
You need to bond the boat to within a few volts of ground earth (particularly when it's ashore).
Double insulated might be the way to go? But you have two problems a) to prevent electricution/stray currents, and b) to meet the regulations. Some solutions to a are shot down by b.
However, being one who distrusts delicate mechanical mechanisms in the same box as salt water, I think there is a case for having some fast blow fuses too.
RCD's do need frequent test. You cannot emphasise that enough. You are unlikely to wear it out and you might stop it from seizing on.
I'm not sure an isloation transformer helps with earth potential issues.
You need to bond the boat to within a few volts of ground earth (particularly when it's ashore).
Double insulated might be the way to go? But you have two problems a) to prevent electricution/stray currents, and b) to meet the regulations. Some solutions to a are shot down by b.
Matthew Millard
New member
Hello
I think not on two counts:
1 The PSCC value of 4ka was a ball park figure and could well be much greater depending on the Zs value at the pontoon.
2 You mention 'the time taken for the circuit breaker to trip...'. I think your pretty buggered if its not been physically operated for a while and has become stuck in the closed position or worst still, with the contacts fused together.
So no, GI's are not the one size fits all cheap solution. The design on paper is excellent but we dont live in a perfect environment on a boat with sticking breakers and RCDs and variable power supply characteristics depending on where your moored.
Unlike the likes of 'Safeshore Marine' who peddle their garden shed black boxes on ebay I dont have a vested interest in flying the flag for GI's or transformers, im just a boater who happens to be an electrician who knows when the wools being pulled over ones eyes.
I dare say there will be some people on here who do have a vested interest in singing there praises, but take a moment to remove the blinkers and consider this. You get what you pay for in boating, i.e. 50 quid GI ebay special made in a garden shed or a £400 isolation transformer from a reputable company.
To insert some two bob bridge rectifier in the protective conductor and have no fail safe method of automatically switching the supply off in the event of the GI failing is just madness.
Good luck to all the boaters who are fooled by the garden shed boffins and the bankside experts.
misterg
Active member
A properly installed isolation transformer is the dog's danglies at solving all these problems. You do need to make arrangements for an earth connection if you're using the installation with the boat ashore.
Shame they're so big, heavy and expensive, though
(Not sure where you're proposing to put fuses, or what they'll achieve over and above the circuit protection that should already be in place.)
Andy
Elessar
Well-known member
Crikey what did you start?
All the talk about peril, disconnect the earth on the boat and you'll probably never know.
It is guarding against an electrical fault that allows the live wire to touch a bit of casing on an appliance that you can touch, on the same day that your RCD fails to work.
If your home made isolator happens to fail at the same time as all of the above, before your circuit breakers trip, you are so unlucky you should never go out in a boat.
Having said that I bought one off ebay that included warning lights - they indicate if there is leakage current that the isolator is blocking. This is useful, before I fitted it I lost all my anodes and both props in just a fortnight due to a fualt on a nearby boat. The warning light would have me disconnecting shorepower and investigating.
It wasn't expensive, was in a robust case that was easy to fit, so it wasn't worth the hassle of getting the soldering iron out.
But no there is no compelling reason stopping you IMHO.
VicS
Well-known member
OK fair enough. We have been talking about the RCD not tripping. Suppose there is also a separate over-current circuit breaker. That is more likely to have been operated from time to time as the main isolator and therefore less likely be seized on . I know there are different types but can I assume that which ever type is used it will be even slower to trip than the RCD?
This all raises another interesting aspect to the value of an RCD at all. You are saying that it probably won't have been tested regularly and could be seized in the on position. So it won't function when called upon to protect against electrocution! It is becoming apparent then that RCDs are a pretty useless bit of kit to have installed ... but you'd still install them even with an isolating transformer.
To get back to the GI.
I believe GIs in the USA are now required by the ABYC regulations to be fitted with with indicating and testing electronics.
Is this not the way forward with GIs?
I have wondered in the past why GIs are installed as they are.
If they were fitted not in the incoming earth conductor but in the bonding connection to the DC negative etc there would be less likelihood of them being subjected to these big fault currents. Even if they were and they failed they would not disable the safety earth connection, merely the bonding to the DC negative .. that in itself something that many people wont accept anyway. I think the diagram below illustrates what I am asking about. Fig 1 shows a normal installation, Fig 2 show the GI just in the bonding connection to the DC negative etc
Can you explain what the objection to Fig 2 is .
VicS
Well-known member
Crums do you have any idea what the fault was ? It must have applied the 12 volt positive to the shorepower earth while leaving the negative still connected to his own sterngear or whatever. I expect he was wondering why his battery kept going flat!
Bobobolinsky
New member
Hm, watching this thread, I do believe that it is about time PBO put down the malfuction lifejackets and did a GI test under fault conditions.
For my two pennorth, I am always suspicious of manufacturers who "pot" everything, so you can't tell what a load of manure they are selling you.
For my two pennorth, I am always suspicious of manufacturers who "pot" everything, so you can't tell what a load of manure they are selling you.
Elessar
Well-known member
never found out. just lucky to have seen it before I lost my drives. The props were irrepairable but the drives un scathed.
F hammerhead, Shamrock quay, sterngear 4ft from a pile don't know if that made things worse. I know it was only a fortnight because it was when I first moved there, the anodes were checked on the Thames only 3 weeks before I moved in, they were reduced to a few specs around the bolts, not just eroded but gone.
I removed the mains lead, then bought and fitted the GI shortly afterwards and all was well - including no indication from the fault lights so the source of the problem had gone I presume, but have now moved berth too. Or perhaps the fault lights don't work I've never seen them on.
The annoying thing was I assumed the boat would have had a GI so didn't worry. My previous 2 boats were American and had them as standard.
Last edited:
Matthew Millard
New member
OK fair enough. We have been talking about the RCD not tripping. Suppose there is also a separate over-current circuit breaker. That is more likely to have been operated from time to time as the main isolator and therefore less likely be seized on . I know there are different types but can I assume that which ever type is used it will be even slower to trip than the RCD?
This all raises another interesting aspect to the value of an RCD at all. You are saying that it probably won't have been tested regularly and could be seized in the on position. So it won't function when called upon to protect against electrocution! It is becoming apparent then that RCDs are a pretty useless bit of kit to have installed ... but you'd still install them even with an isolating transformer.
To get back to the GI.
I believe GIs in the USA are now required by the ABYC regulations to be fitted with with indicating and testing electronics.
Is this not the way forward with GIs?
I have wondered in the past why GIs are installed as they are.
If they were fitted not in the incoming earth conductor but in the bonding connection to the DC negative etc there would be less likelihood of them being subjected to these big fault currents. Even if they were and they failed they would not disable the safety earth connection, merely the bonding to the DC negative .. that in itself something that many people wont accept anyway. I think the diagram below illustrates what I am asking about. Fig 1 shows a normal installation, Fig 2 show the GI just in the bonding connection to the DC negative etc
Can you explain what the objection to Fig 2 is .
I assume the objection in Fig2 would be the lack of protection from stray AC currents from the shore supply or from an onboard leaking appliance.
VicS
Well-known member
Sorry I dont follow. Any currents going to earth from the AC installation would still trip the RCD in the consumer unit.
If they were leaks to the DC system, since the DC system should still be connected to the Ships ground and then via the GI to the shorepower earth the RCD should trip.
lw395
Well-known member
A properly installed isolation transformer is the dog's danglies at solving all these problems. You do need to make arrangements for an earth connection if you're using the installation with the boat ashore.
Shame they're so big, heavy and expensive, though
(Not sure where you're proposing to put fuses, or what they'll achieve over and above the circuit protection that should already be in place.)
Andy
You need to remove that connection when afloat. Which is where the GI came in.
Personally I would not be surprised if 1.2 V of ground difference was sometimes wholly inadequate anyway.
Ashore there is a case for locally bonding the incoming earth to a stake in the ground, if you are at the end of a long cable. There have been cases of people finding their ali ladder is not equipotent with the guardrails as they board.
Fusing is tricky when you don't trust the shoreside or earth.
But a fuse will always fail open. The danger is if you fuse what is supposed to be incoming live, and live is really on another line.
Maybe you should fuse all three incoming lines and crowbar them to boat earth in the event of a fault. (is that the sound of the IET coming to get me?)
If you do not need mains appliances onboard, apart from battery chargers, would a double insulated battery charger, L+N isolated from boat by a proper margin, be a good solution?
And would it meet the regs?