DIY Galvanic isolator.

[Vara]

This months (Nov) PBO has an article by David Berry about making a DIY galvanic isolator, it looks to be well within my limited capabilities.

He is a bit optimistic about price as I reckon it will be about £20 not the "under £10" quoted.

Is there any compelling reason why I shouldn't proceed, or do the ones from chandlers contain a mystery ingredient worth the £50 difference.

 

[markdj]

I thought the article was a bit over complicated as it was using underpowered diodes in parallel to give the same protection as one larger one and this is bad practice just for the sake of saving money.

There was no mention of the rigorous testing that commercial galvanic isolators go through before they are approved as safe for use and I thought this was the most glaring omission.

Making one yourself with a poor heatsink as shown gives a false sense of security as it could be a fire hazard if supplied with current in an earth leakage situation, albeit for less than a second.

The standards of galvanic isolators is as follows:

"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."


Just my 2p

Mark
www.boatdoctorni.com

 

[oldvarnish]

Where is the sense in saving fifty quid to produce an inferior product which has been through no testing and has to conform to no specification, as the previous poster correctly points out?

Properly built, tested and installed, these are important safety items.

To make your own to save a few quid is to lose sight of the priorities.

 

[garvellachs]

I haven't read the article, but there is a similar idea here:
http://quinco.890m.com/Library/GalvanicIsolatorDIYAssy.pdf
There the diodes are not really in parallel - they are used to provide the necessary voltage drop in both directions of current? I have made one up using 50 amp diodes from China.

 

[Cantata]

Hear hear. This thing is there to save your life. Nuts to mess about making one to save a few quid.

 

[VicS]

My antivirus software stopped me opening the above link

But the idea of using two bridge rectifiers is well known

See http://www.yandina.com/acrobats/GIDIYAssy.pdf

and also http://www.yandina.com/galvanicIsolator.htm

This design and the one in PBO both exploit the bridge rectifier to use to diodes in series to obtain the required voltage drop and the two arms of the bridge in parallel to obtain the required current rating.

Bridge rectifiers can be obtained quite cheaply from Maplin Electronics


The danger of an improvised unit is that the diodes could fail when subject to a fault condition which will then leave the safety earth connection to the shore power installation open circuit. For this reason a GI should be tested after any fault that causes the mcb to trip

 

[misterg]

"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."

Looked at making one a while back, and came across something similar (I think it was the ABYC?? standard). I couldn't find any bridge rectifiers that got close to that for less than £100 each (150 amp + rating). The 'normal' 10 - 50 amp bridges will pop like fireworks if they get a full phase-earth fault across them, and probably do so in the first 10 or 20ms, probably before your RCD has had a chance to trip (40 - 200ms) and certainly before your MCB has tripped (400ms+). Upshot is that power stays on, and everything that should be 'earthed' is now sitting there at mains voltage.

I opted to unplug the shore power lead instead.

I thought that the PBO article was shoddy in not pointing out what these things *should* be built to withstand.

Andy

 

[wotayottie]

This months (Nov) PBO has an article by David Berry about making a DIY galvanic isolator, it looks to be well within my limited capabilities.

He is a bit optimistic about price as I reckon it will be about £20 not the "under £10" quoted.

Is there any compelling reason why I shouldn't proceed, or do the ones from chandlers contain a mystery ingredient worth the £50 difference.

Its your decision but I have gone ahead. I've no doubt that the end result is not quite as safe as an expensive commercial product but then I see the risk of failure as being pretty small since most mains electrical equipment is well insulated these days and indeed I use little of it aboard the boat. Certainly I see no sense in going out in rough weather in a small boat , itself a risky occupation, and then being totally risk averse about everything else.

Given that few of us have a limitless amount of money to spend on boating, the safety bit needs to be spent where there is the highest risk. How many people have proper modern gas lockers and up to date pipe for example? How many have decent quick acting fuses in their 12v systems? Liferafts? etc etc.

 

[seadog30]

DIY ?

The real point to consider is if you buy an expensive swindlery supplied unit if it kills you you have someone to sue

 

[markdj]

Thank goodness I'm not the only one who thinks some of these articles are just not up to scratch. There have been previous electrical pieces too which I personally could not contemplate to do to any person's boat.

I think if you want mains power plugged in all the time then a proper tested isolator is a small price to pay for protecting your anodes against stray currents. Having a unit on the boat that WILL burn or explode upon an earth leak is just simply careless at best and dangerous at worst.

For those who have such items installed, make sure they are mounted above the waterline and away from any flammable material, gas lockers, petrol cans etc.

I can't believe that some people are asking things like having in date gas pipe, fuses, liferafts !!!! this is basic, basic stuff.

 

[Vara]

I can't believe that some people are asking things like having in date gas pipe, fuses, liferafts !!!! this is basic, basic stuff.

My dear chap, it may be basic to you but the vagaries of galvanic protection are out of my fields of expertise.

 

[Vara]

Thanks everybody for all the information.

It would appear that a proper Isolator, tested and with a CE mark is the right way forward.

Which rather begs the question, why is PBO publishing a potentially(sic) misleading article promoting a possibly dangerous device.

Are these articles not Peer reviewed?

 

[bedouin]

Thanks everybody for all the information.

It would appear that a proper Isolator, tested and with a CE mark is the right way forward.

Which rather begs the question, why is PBO publishing a potentially(sic) misleading article promoting a possibly dangerous device.

Are these articles not Peer reviewed?

Or it rather begs the question why you trust the opinions on this board over that in the magazine.

A number of the apparently-well-informed comments on this thread are just plain wrong

 

[duncanmack]

The real point to consider is if you buy an expensive swindlery supplied unit if it kills you you have someone to sue

I like that!

It is really quite apt. Did you mean it?

 

[Vara]

Or it rather begs the question why you trust the opinions on this board over that in the magazine.

A number of the apparently-well-informed comments on this thread are just plain wrong

Yes but which ones ?

Now, if you know the correct answer, or even know that a correct answer doesn't exist, don't tease, supply some references.

 

[wotayottie]

I can't believe that some people are asking things like having in date gas pipe, fuses, liferafts !!!! this is basic, basic stuff.

Dont really understand your point. There are many risks on boat a boat and some are greater than others. Given that you dont have a bottomless pocket to buy every safety feature known to commerce, you have to make choices. I've seen three boats go up in flames with gas explosions, two hit by lightning, and one with a serious fire thanks to an engine moving on its mounts and shorting out unfused 12v battery cables - not to mention boats with keels falling off and sinking (not Bavarias), masts falling down etc.

Yet none of the four boats I've bought personally had up to date gas equipment on board , none had CO detectors, none had escape hatches to rear cabins, three didnt have currently compliant gas lockers and one didnt even have a bilge pump. Against this sort of background, it seems to me that the risk from this project exists but is small and extra money saved could be spent elsewhere.

In the end it's a personal decision.

 

[bedouin]

Yes but which ones ?

Now, if you know the correct answer, or even know that a correct answer doesn't exist, don't tease, supply some references.

I know enough to know that some of the others are talking nonsense, but not enough to give an authoritative statement myself; as this is an important safety matter I would rather keep quiet than add to the misinformation that is flying around here.

For example an RCD MUST trip within 40ms (not 40-200ms as one poster says) and most will trip in 10ms or less if a significant earth current is applied. Similarly provided that there is an RCD fitted the isolator would provide zero fire risk.

Provided that an RCD is fitted then I cannot see anyway a device such as that described would pose an additional risk; but I am not an expert and I would not want anyone to make any safety critical decisions based on my opinion. However I would tend to trust the advice given on the yandina site - they seem to know what they are talking about.

 

[misterg]

For example an RCD MUST trip within 40ms (not 40-200ms as one poster says) and most will trip in 10ms or less if a significant earth current is applied. Similarly provided that there is an RCD fitted the isolator would provide zero fire risk.

Explanation for my quoting 40-200ms:

I think the IEE 17th edn. specifies that the RCD must trip in and <200ms with a leakage equal to it's trip rating and <40ms with current which is 5x it's trip rating. (It *might* only only be 30ma or less RCDs that are tested a 5x - I don't know for sure). Obviously it is the shorter time that's relevant here, my apologies for the confusion.

If 10ms is typical, then I'm not surprised, but it doesn't change my mind.

Not particularly a reply to Bedouin, but further explanation of my comments above:

(happy to be corrected)

Bear in mind that under fault conditions, your 'box' will effectively be connected directly between 'live' and 'earth'. It needs to be able to survive this for several mains cycles until either the RCD trips, or (if this is faulty / absent) the MCB trips.

Personally, my concern is that unless adequately rated, the diodes will be destroyed virtually instantly in the event of a heavy fault (on the first 1/2 cycle of the supply following the fault - i.e. within 10ms) leaving the 'earth' lead open circuit. I don't know, but have my doubts about an RCD tripping in those circumstances, and if it did trip, the most likely human reaction is to reset it, and see what happens. I doubt that a 'circuit breaker' would trip.

Part of the spec on the rectifiers is the maximum 1/2 cycle current that they can take. While this might not be directly applicable, I take it as a good indicator. From my (limited) research, cheap-ish, commonly available bridges from RS, etc. have a max. 1/2 cycle current of ~150 amps. A solid 'live' to 'earth' fault could generate peak current ~1000 amps. That discrepancy was a bit too much for me to reconcile. RS do list bridges with 1/2 cycle currents of ~800A, but they cost as much as / more than a GI. (There's also an I^2 t rating which might be relevant - I don't know).

Since it was no big deal to me to unplug the shore power lead I didn't take it any further.

Again personally: I don't think these pose a fire-risk as such, provided the gubbins are inside a metal box to contain the sparks and smoke if they blow up.

To a large extent, we're discussing how many angels can fit on the head of a pin - The situations where the 'earth' cable actually does *anything* are rare, and the situations where a home made GI might cause a problem are even rarer.

You pays your money and takes your chances, but at least be aware of the trade-offs.

Andy

 

[john_morris_uk]

Hear hear. This thing is there to save your life. Nuts to mess about making one to save a few quid.

With respect, 'these things' are there to prevent stray galvanic currents.

There are safety implications if they fail or don't work properly, but the prime use of them is to reduce stray galvanic currents. They are not built to 'save life'.

Rant over

 

[bedouin]

With the sort of currents that would blow the diodes, both MCBs and RCDs will typically blow in about 10ms (say greater thn 5 times rated load). Longer times are related to current draws nearer to the rating of the device (say twice) but those sort of currents are not going to blow the diodes.

In practice even a dead short is not going to cause 800A to flow - there are just too much resistance for that to happen and things would be going pop all over the place; in a typical marina installation I would have thought the highest current is more likely to be in the order of 100A-200A, possibly much less. Earth connections are not designed to carry much current.

Provided that an RCD is fitted and working, I think the chance of the diodes being blown is vanishingly small; and even if they do then it still would not risk life or fire as the RCD would still protect the installation. If that is a worry then the circuit could easily be amended with a resistor/diode combination in parallel with the bridge rectifiers.