I understand David Berry recently provided plans for a simple inexpensive galvanic isolator in a short article in Practical Boat Owners magazine. I wonder if anyone can provide me with a copy of that article and the plans. Thanks.
Galvanic Isolator Plans
- Thread starter wufibugs
- Start date
- Status
In a word NO. Sorry to be unhelpful but:-
It is not without good reason that the professional galvanic isolators use high current diodes mounted on substantial heat sinks, the diodes being capable of continuously carrying the 'maximum prospective short circuit current'. This is why they cost as much as they do.
The risk with the DIY solution is that, in the event of a short between hot and earth, and there is no breaker or it does not interrupt the current, then your earth and the marina's earth could be live (hot).
Also there is no assurance that should the bridge blow it would fail to short circuit rather than open circuit which would again leave your boat earth hot.
Then there are the insurance aspects to be considered in the event of a fire or loss of life.
As a former regular reviewer of yachting equipment and writer of technical articles for PBO I intended to write to them warning readers of the possible dangers resulting from the use of the lower power standard rectifier bridges which the article recommends and the inadequate heat sinks.
Like everything else I never got round to it though.
It is not without good reason that the professional galvanic isolators use high current diodes mounted on substantial heat sinks, the diodes being capable of continuously carrying the 'maximum prospective short circuit current'. This is why they cost as much as they do.
The risk with the DIY solution is that, in the event of a short between hot and earth, and there is no breaker or it does not interrupt the current, then your earth and the marina's earth could be live (hot).
Also there is no assurance that should the bridge blow it would fail to short circuit rather than open circuit which would again leave your boat earth hot.
Then there are the insurance aspects to be considered in the event of a fire or loss of life.
As a former regular reviewer of yachting equipment and writer of technical articles for PBO I intended to write to them warning readers of the possible dangers resulting from the use of the lower power standard rectifier bridges which the article recommends and the inadequate heat sinks.
Like everything else I never got round to it though.
Don, did you happen to read the rather long thread that resulted from that article. There was a debate among those who know a great deal more than me about the contents of your post here. I was leaning, though not entirely persuaded, to consider fabricating an isolator not only because of the technical argument made by some posters there, but the contention of one that having opened an off the shelf product, he found it no more convincing or invinceable than the one that appeared in the PBO article.
Bologna!
Hate to be Johnny Raincloud Donalex but some of the scenarios you describe are pretty hard to imagine. A hot with out a breaker. I'm sure the shore power distro box has one. The only way you could short the hot and ground and not have a breaker is, well there is not a way that I can imagine. I suppose that IF the marina decided to take the risk it could install a dock with no breakers in the circuit. Oh, wait, the electric company would have a fit and not attach the power line.
So the diodes only have to handle the full current of the breaker (plus a safety factor of 2ish). You can buy these for $7-$12 each from digikey. You can even get them as anode post and cathode post so you have a slick professional looking unit.
As for the fire risk, heck most of us have already embraced that horror by doing our own AC wiring.
The folks that make sails want you to believe that only they can do that for you (except SailRite), the plumber wants you to believe that there is no way you could possibly tighten a fitting and get it right, the electrician wants you to never even open the breaker box to reset a blown CB. This is a recurring theme which all professional organizations have. That is the way they stay in business.
If you don't feel comfortable doing it then fine but don't try and tell me there is some mystical property to company X's diode from digikey and that the one I buy from them is somehow "unblessed" and simply can't work.
Hate to be Johnny Raincloud Donalex but some of the scenarios you describe are pretty hard to imagine. A hot with out a breaker. I'm sure the shore power distro box has one. The only way you could short the hot and ground and not have a breaker is, well there is not a way that I can imagine. I suppose that IF the marina decided to take the risk it could install a dock with no breakers in the circuit. Oh, wait, the electric company would have a fit and not attach the power line.
So the diodes only have to handle the full current of the breaker (plus a safety factor of 2ish). You can buy these for $7-$12 each from digikey. You can even get them as anode post and cathode post so you have a slick professional looking unit.
As for the fire risk, heck most of us have already embraced that horror by doing our own AC wiring.
The folks that make sails want you to believe that only they can do that for you (except SailRite), the plumber wants you to believe that there is no way you could possibly tighten a fitting and get it right, the electrician wants you to never even open the breaker box to reset a blown CB. This is a recurring theme which all professional organizations have. That is the way they stay in business.
If you don't feel comfortable doing it then fine but don't try and tell me there is some mystical property to company X's diode from digikey and that the one I buy from them is somehow "unblessed" and simply can't work.
PBO is probably my favourite boat magazine (because of all the stuff I still need to do/learn), but I haven't yet picked up the November issue; it gets to the colonies rather late...
Richard - did you mean this PBO forum thread?
Richard - did you mean this PBO forum thread?
That is it, Kenn. The post by Halycon says that he looked into an isolator that cost in the neighborhood of $150 us, and found little to differ from what could be made at home. Gospel? No. But the weight of the authority in that thread seemed to be that a properly constructed home grown GI could work well and safely.
Bill
Bill,
I note your views and respect them - BUT; I connect my portable generator to the shore power socket for on board distribution of AC and my H376 came with only an overcurrent breaker but no GFI. These breakers are too slow to save a life anyway. Many boats have no on board breaker at all.
Also, as I remarked, there is no guarantee of a GFI breaker operating - they have been known to fail and dock outlets be incorrectly wired! I have experienced the latter on three occasions.
A 30 amp cord running 60 amps as you suggest means this current passes through two diodes in series for each half cycle. Total heat generated at Vf per diode is approx 140 watts. Max junction temp of diodes - say 180°C max.
Thermal resistance of complete device needs to be 160/140 = 1.1°C/watt of which the rectifier itself will use 0.5°C/watt leaving only 0.6°C/watt for the heatsink.
Please see heatsink 100mm x 100mm x 32mm on http://uk.rs-online.com/web/search/searchBrowseAction.html?method=retrieveTfg&Ne=4294922322&Ntt=heat+sink&Nr=AND%28avl%3auk%2csearchDiscon_uk%3aN%29&Ntk=I18NAll&Ntx=mode%2bmatchallpartial&N=4294957144+4294654333&Ns=stockPolicy_uk|1||new_uk|1&Nty=1&binCount=495&multiselectParam=4294957144&selectSubRange=High%20Power%20Extruded%20Heatsinks%20and%20Fan/Heatsink%20Assemblies#breadCrumb
See second item down No 158-578. This is about the size of the heat sink on proprietary units.
I would, and no doubt, you would build your own and make sure it was conservatively specified, correctly wired and thus safe. To recommend others to build their own based on the PBO article is, in my view, irresponsible. (I can send it to you privately).
But the real crunch is the insurance problem in cases of accident.
Bill,
I note your views and respect them - BUT; I connect my portable generator to the shore power socket for on board distribution of AC and my H376 came with only an overcurrent breaker but no GFI. These breakers are too slow to save a life anyway. Many boats have no on board breaker at all.
Also, as I remarked, there is no guarantee of a GFI breaker operating - they have been known to fail and dock outlets be incorrectly wired! I have experienced the latter on three occasions.
A 30 amp cord running 60 amps as you suggest means this current passes through two diodes in series for each half cycle. Total heat generated at Vf per diode is approx 140 watts. Max junction temp of diodes - say 180°C max.
Thermal resistance of complete device needs to be 160/140 = 1.1°C/watt of which the rectifier itself will use 0.5°C/watt leaving only 0.6°C/watt for the heatsink.
Please see heatsink 100mm x 100mm x 32mm on http://uk.rs-online.com/web/search/searchBrowseAction.html?method=retrieveTfg&Ne=4294922322&Ntt=heat+sink&Nr=AND%28avl%3auk%2csearchDiscon_uk%3aN%29&Ntk=I18NAll&Ntx=mode%2bmatchallpartial&N=4294957144+4294654333&Ns=stockPolicy_uk|1||new_uk|1&Nty=1&binCount=495&multiselectParam=4294957144&selectSubRange=High%20Power%20Extruded%20Heatsinks%20and%20Fan/Heatsink%20Assemblies#breadCrumb
See second item down No 158-578. This is about the size of the heat sink on proprietary units.
I would, and no doubt, you would build your own and make sure it was conservatively specified, correctly wired and thus safe. To recommend others to build their own based on the PBO article is, in my view, irresponsible. (I can send it to you privately).
But the real crunch is the insurance problem in cases of accident.
Donalex, if "Many boats have no on board breaker at all.", then those owners should be addressing that problem, and the isolator addition becomes secondary importance. Current ABYC standards call for AC breakers within 10 feet of the shore power inlet.
Last edited:
How Much for DYI Isolator?
This one seems reasonable enough if it's any good:
http://www.defender.com/product.jsp?path=-1|328|1212796&id=605562
This one seems reasonable enough if it's any good:
http://www.defender.com/product.jsp?path=-1|328|1212796&id=605562
Ron makes an interesting point. It does appear that few boats, particularly under 40' are manufactured with isolators. Certainly my Hunter was not. Since it appears that a good isolator is, conservatively, $125, and few people are going to skinny down into the hold where the shore power cord enters the boat to put one in, we are probably looking at another $125 to install, the minimum for this project is $250. For those of us in the under 45' class, that's far from chump change. That being so, am I correct in thinking that a "potentially" underrated [and this is obviously open to considerable debate] isolator is far better for everyone at the dock, than none at all? Put another way, is a potentially false sense of security worse than no security at all?
The one Rick linked certainly seems reasonable enough particularly when compared with the cost of the components for a diy equivalent. I seem to recall mine taking less than an hour to install at a cost of less than $100, the result of which were zincs which appeared almost new after one year when previously I had gone through 2 in a 6 month season and this excluded the protection it affords to my no so cheap Maxprop.
I haven't read the specs for the PBO isolator but it isn't rocket science to acquire and assemble parts equivalent to that of a good quality commercially available one.
In the context of Wufibugs question, even poorly constructed isolators made from inferior components offer some protection - the issue is how much risk is one willing to take.
I haven't read the specs for the PBO isolator but it isn't rocket science to acquire and assemble parts equivalent to that of a good quality commercially available one.
In the context of Wufibugs question, even poorly constructed isolators made from inferior components offer some protection - the issue is how much risk is one willing to take.
In looking quite a number of isolators under $200, most seem to use the case as the heat sink. In that regard, I would be more inclined to construct an isolator using a dead inverter case than what was, I believe, suggested in the PBO article. Again too, the major beef with the PBO article seems to be the capacity of the diodes. If true, why not take the PBO design and simply advocate for different diodes? I would certainly love to hear a specific recommendation for what specs one would prefer. Don seems to know what he is talking about, but I am a little concerned that we can't really be sure what has been installed by a manufacturer. Since that information is normally not provided, and I could not interpret it if it were, on what basis would I have significantly greater confidence in an off-the-shelf unit than one I built based on an "overdesign" using the factors suggested by Don and Bill? I think someone also mentioned insurance and recourse. Surely we are not thinking that either I or the insurance company would take on the isolator manufacturer, hiring the experts necessary to reconstruct the event with the level of courtroom confidence that points to an underdesigned isolator as the cause of the fire that consumed my boat? Similarly, I doubt they would undertake the cost or expense of arguing with me about the competence of the isolator I built. I know this invites a multitude of fascinating anecdotal horror stories about our relations with claims adjusters. I am basing my opinion on the cost/benefit analysis that most insurance companies employ is claims assessments.
Well said Donalex
It would be irresponsible to have a boat with no AC breaker near the shore power inlet. Also irresponsible to try and build something you have now clew about the working of. Also also irresponsible to listen to some one who knows what they are doing and blindly go off and assume that since he did it such and such a way that that will work for you. If you don't know what you are doing then ask someone or do your homework.....
The issue here is the diodes shorting open and leaving you with no ground circuit. Not to get to geeky but if you put a buzzer in parallel with the unit and make sure the resistance of the buzzer is more than the diodes but less than it takes to electrocute yourself you would have an audible indication that the diodes where bad. Course you would have to make current flow in the ground circuit....... there are lots of ways to give indication that the ground circuit is open or closed. Basic electricity! That is what drives up the cost on the units. The base unit is pretty cheap but the indicator circuit and fail safeing the unit are pretty expensive.
It would be irresponsible to have a boat with no AC breaker near the shore power inlet. Also irresponsible to try and build something you have now clew about the working of. Also also irresponsible to listen to some one who knows what they are doing and blindly go off and assume that since he did it such and such a way that that will work for you. If you don't know what you are doing then ask someone or do your homework.....
The issue here is the diodes shorting open and leaving you with no ground circuit. Not to get to geeky but if you put a buzzer in parallel with the unit and make sure the resistance of the buzzer is more than the diodes but less than it takes to electrocute yourself you would have an audible indication that the diodes where bad. Course you would have to make current flow in the ground circuit....... there are lots of ways to give indication that the ground circuit is open or closed. Basic electricity! That is what drives up the cost on the units. The base unit is pretty cheap but the indicator circuit and fail safeing the unit are pretty expensive.
I thought that was what I was doing with this thread, asking people who do know what they are doing. I assumed that that the publishers of PBO engaged pretty knowlegeable people to write their articles, and would have published a caveat if the feedback indicated the article was incorrect, as somehere pointed out about the outpute wire. Rather than just building it based on the article, I did do some homework: I read through the critique on this and other forums. If your suggestion is that "homework" means learning as much as about this subject as either you or Mr. Perry possess, I'm sorry but that is improbable. I don't buy a set of plans designed by an architect, and then become as knowledgeable as he in order ensure that his design is correct. One goes through this world with a certain level of trust. To some degree that level is matter of judgment; in another sense the line between evaluation and blind faith is always somewhat arbitrary. Some people seem to think Perry is both incompetent and irresponsible. Others disagree. On some level, Bill, your response is a critique of DIY in general, the essence of which is non experts attempting to do what experts have been trained to do. It can be electrical, it can be laying a floor, or anything else I suppose, short of brain surgery. At this point I am having still trying to evaluate whether his design is defective, and whether it is necessarily true that a commercial product will be so far superior to his design to justify the very substantial additional cost. Please don't misinterpret this as being unappreciative of your comments. On the contrary.
Totally agree wufibugs
IF I purchased a set of plans (from a reputable person) I'd expect them to be correct. However, all I'm seeing from this thread is that the plans are "too dangerous." Now if someone would post a link where we could have a look at them and check out the person who designed them.....
IF I purchased a set of plans (from a reputable person) I'd expect them to be correct. However, all I'm seeing from this thread is that the plans are "too dangerous." Now if someone would post a link where we could have a look at them and check out the person who designed them.....
Oh dear there's more
I went on PBO's own site where one can buy a copy of the November issue of the Dave Berry article for £6.95. Then I looked at their own forum on YBW.com where there is a massive thread on this subject with several links to other DIY isolators. See:- http://www.ybw.com/forums/showthread.php?p=2272795
One posting by 'Halcyon' points to the ABYC specification, updated July 2008 which now calls for a monitoring device to alarm if the Galvanic Isolator blows or otherwise goes open circuit.
The link to ABYC spec A-28 is below and is a good read for the techies amongst us.
New installations without this should attract adverse comments from a surveyor and consideration must be given to potential sale of the boat with a DIY isolator.
Others should just keep clear and do the safe thing.
One interesting issue is the location of the isolator and the result of an open circuit isolator with an on board ground fault, which may be the result of the same incident.
http://www.nmma.net/Lib/docs/nmma/cert/techupdates/A-28_Galvanic_Isolators_Standard.pdf
I went on PBO's own site where one can buy a copy of the November issue of the Dave Berry article for £6.95. Then I looked at their own forum on YBW.com where there is a massive thread on this subject with several links to other DIY isolators. See:- http://www.ybw.com/forums/showthread.php?p=2272795
One posting by 'Halcyon' points to the ABYC specification, updated July 2008 which now calls for a monitoring device to alarm if the Galvanic Isolator blows or otherwise goes open circuit.
The link to ABYC spec A-28 is below and is a good read for the techies amongst us.
New installations without this should attract adverse comments from a surveyor and consideration must be given to potential sale of the boat with a DIY isolator.
Others should just keep clear and do the safe thing.
One interesting issue is the location of the isolator and the result of an open circuit isolator with an on board ground fault, which may be the result of the same incident.
http://www.nmma.net/Lib/docs/nmma/cert/techupdates/A-28_Galvanic_Isolators_Standard.pdf
Last edited by a moderator:
Ill add my .002 on this..
Heat sinks are generally designed around steady state conditions where thermal equilibrium is achieved (ie, thermal "impedance" equations). But for short time frames – like several cycles of AC, this model is extremely conservative. In this case where a fuse should trip fairly fast (i.e., several cycles of AC or 30 msec), you need to use an adiabatic model ( Power*time equals M*C*delta Temp). Given the short time frame, the allowable power levels for a given junction temp are a huge amount larger than for the steady state equilibrium condition. As an example, a chunk of aluminum 6 cubic cm takes 5300 watts to heat up 125C in 33msec (2 AC cycles in the US). Steady state (and with a heat sink), the input power for the same temp rise would be a huge amount lower. Thermal run away.. not likely. The forward voltage in a diode drops with temp – the power dissipated drops as temp rises.
The safety of an earth ground needs to have a breaker in order to work. In general, the fuse must blow before the galvanic isolator does for the safety aspect to work. If for some reason there is not a breaker (Yikes..), I think I would want the galvanic isolator to blow up. Consider that if the isolator has 60 amps flowing through it, somewhere else in the circuit is dissipating 60amp *117 volts = over 7000 watts (not exactly.. but hopefully it shows the point). This other place in the circuit that is dissipating the 7000 watts is likely the wires and instead of getting shocked, you are possibly going to have a fire. My preference would be getting shocked over the fire as this sort of shock you have a good chance of just swearing about. Once shocked, you know there is something wrong.
In the five years or so I’ve been following sailboat electrical accidents, the only one I’ve heard about was recently described here where a kid was swimming in a marina near an outboard with AC on it. In this case, it is a GFCI which would have prevented this accident from happening. A normal fuse would have done nothing. Note that if the slip power would have had GFCI, the earth green wire would NOT have even needed to be hooked up to prevent the accident. I.e., someone could have just used a two wire extension cord and for the particular case, the GFCI would have tripped. It would not have mattered if the galvanic isolator was fried or not. GFCI works by a current imbalance between hot and neutral – and the imbalance current does not need to flow back through the green wire. The path through water back to the marina breaker box works just fine for tripping GFCI.
I didn’t see the plans for the diode isolator.. but if the diodes are rated for some high currents, the device probably works fine. If the fuse trips in the time frame its supposed to, I dont think a heat sink would be needed. But.. if the fuse trips slow (somewhat of a fault condition), then the heat sink might be necessary.. If I sold the boat, Id definitely remove my kludge..
If there is an accident on the boat such as a deadly shock or a fire, do insurance providers hold the boat owner responible for doing something "not to standard"? I think a lot of boats have wiring done by the owners and maybe they have a lot more to worry about than a garage made isolator.
Heat sinks are generally designed around steady state conditions where thermal equilibrium is achieved (ie, thermal "impedance" equations). But for short time frames – like several cycles of AC, this model is extremely conservative. In this case where a fuse should trip fairly fast (i.e., several cycles of AC or 30 msec), you need to use an adiabatic model ( Power*time equals M*C*delta Temp). Given the short time frame, the allowable power levels for a given junction temp are a huge amount larger than for the steady state equilibrium condition. As an example, a chunk of aluminum 6 cubic cm takes 5300 watts to heat up 125C in 33msec (2 AC cycles in the US). Steady state (and with a heat sink), the input power for the same temp rise would be a huge amount lower. Thermal run away.. not likely. The forward voltage in a diode drops with temp – the power dissipated drops as temp rises.
The safety of an earth ground needs to have a breaker in order to work. In general, the fuse must blow before the galvanic isolator does for the safety aspect to work. If for some reason there is not a breaker (Yikes..), I think I would want the galvanic isolator to blow up. Consider that if the isolator has 60 amps flowing through it, somewhere else in the circuit is dissipating 60amp *117 volts = over 7000 watts (not exactly.. but hopefully it shows the point). This other place in the circuit that is dissipating the 7000 watts is likely the wires and instead of getting shocked, you are possibly going to have a fire. My preference would be getting shocked over the fire as this sort of shock you have a good chance of just swearing about. Once shocked, you know there is something wrong.
In the five years or so I’ve been following sailboat electrical accidents, the only one I’ve heard about was recently described here where a kid was swimming in a marina near an outboard with AC on it. In this case, it is a GFCI which would have prevented this accident from happening. A normal fuse would have done nothing. Note that if the slip power would have had GFCI, the earth green wire would NOT have even needed to be hooked up to prevent the accident. I.e., someone could have just used a two wire extension cord and for the particular case, the GFCI would have tripped. It would not have mattered if the galvanic isolator was fried or not. GFCI works by a current imbalance between hot and neutral – and the imbalance current does not need to flow back through the green wire. The path through water back to the marina breaker box works just fine for tripping GFCI.
I didn’t see the plans for the diode isolator.. but if the diodes are rated for some high currents, the device probably works fine. If the fuse trips in the time frame its supposed to, I dont think a heat sink would be needed. But.. if the fuse trips slow (somewhat of a fault condition), then the heat sink might be necessary.. If I sold the boat, Id definitely remove my kludge..
If there is an accident on the boat such as a deadly shock or a fire, do insurance providers hold the boat owner responible for doing something "not to standard"? I think a lot of boats have wiring done by the owners and maybe they have a lot more to worry about than a garage made isolator.
Yikes indeed. Let us hereby agree that no-one should have AC wiring in their boat without also having breakers or fuses rated properly for their wiring, so that you have independent (hopefully tested) protection and don't have to take your chances with the shorepower breaker.
In other words, if your boat doesn't have proper protection on the shorepower circuit, you have something to do BEFORE you even consider galvanic isolation.
alternate solutions
Well I read through the other sites posts. Thanks Donalex.
The original poster decided to just unplug from shore power as a solution. I'm not ready to go that route yet but it did give me a couple of ideas.
First I agree that that you can build an isolator for yourself that is better than the commercial units.
Second; why not just put a monitor circuit in the ground line that gives an indication of when there is DC current flowing in the ground wire. Just like a reverse polarity indicator to tell you there is a problem.
Third; use your inverter for all AC and only connect the shore power to the battery charger. Marine batt chargers are isolation types and do not have a electrical connection to the boat wiring.
Well I read through the other sites posts. Thanks Donalex.
The original poster decided to just unplug from shore power as a solution. I'm not ready to go that route yet but it did give me a couple of ideas.
First I agree that that you can build an isolator for yourself that is better than the commercial units.
Second; why not just put a monitor circuit in the ground line that gives an indication of when there is DC current flowing in the ground wire. Just like a reverse polarity indicator to tell you there is a problem.
Third; use your inverter for all AC and only connect the shore power to the battery charger. Marine batt chargers are isolation types and do not have a electrical connection to the boat wiring.
I’m not exactly sure why the earth ground is not at "ground" potential for some marinas. However.. one possibility is that boats on the marina ground distribution are using the earth ground wire to carry current (i.e., incorrectly wired). When the ground distribution wire is carrying current, because the wire has some resistance, it develops a voltage with a DC component. When you use this ground on your boat, you get the electrolysis problem and then need the isolator. Doesn’t matter if your ground is wired perfectly with no current on the ground.
At a marina with GFCI outlets, if your boat is incorrectly putting current on the ground line, the GFCI plug will trip. The GFCI forces everyone (who hasn’t bypassed the plug) to wire their boats AC correctly. According to this theory - which could be wrong - marinas with GFCI outlets should have much less of a ground induced electrolysis problem because no one is able to put any current on the ground - so it acually will be at ground potential.
- Status