galvanic isolator

[andrew215]

Does anyone have a schematic, with values, of a galvanic isolator. I have a basic schematic, but I don't have the diode values. it just shows the set up for the diodes, but no other components or anyvalues.
thank you in advance.

 

[ChasB]

Uncanny. I was just looking at one on mailspeedmarine at about £80 for 30kW. How much do you anticipate a home-made one costing?

 

[andrew215]

I've got some diodes from my junk box, 4 of them (I don't have the nomenclature or the inverse voltage rating handy, but they look pretty "robust") but to answer your question, I should imagine under 10 lbs, assuming not to much else is in it. from looking at the pictures, it just looks like diodes in a heat sheild of some sort. (going by the description of the way it works, and apparently the diodes have to be rated for 135 percent of the voltage in case of a breaker failure. ) it says it only passes a minimum of voltage and only is used on the ground wire between the shore power and the ground connection on the boat. (the schematic I have is two diodes in series biased one way and two diodes biased the other way, with the ends tied together, like a rectifier circuit. )I'm sorry I don't have the url for the diagram I do have, which I got off the internet. I have been searching for a more detailed one and thought I would try this forum. (they cost here in the US from 90.00 US to over 200. based on amperage.

 

[andrew215]

btw, it sounds like ur looking at an isolation transformer, which are more expensive than the isolator, but supposed to be much better..

 

[VicS]

This site describes how to make one but no diagram and no component values. So was hardly worth the bother of posting!

You have a ciruit diagram so you know that the AC connections are not used (cut off) and that the resultant rectifiers with just positive and negative terminals left are connected in parallel but with their polarities reversed.

In choosing bridge rectifiers to use the important thing is that they must have a forward current rating in excess of the maximum fault current that your supply could provide. That is more than the shoreside fuse or overload trip (MCB). Maplin list them up to 35 amps. Your shore power lead should be fused within that limit if the MCB setting is higher.

Sorry a bit sketchy. If you are not happy designing and making one then you should buy a commercial one and not compromise your safety.

Unless you have a circuit diagram for something more sophisticated you must realise that this idea of using two rectifiers only is very basic. It does not comply with currect recommendations in the US for example.

I'd be interested in any more sophisticated circuit you have if you would care to post it.

William_H will be along soon no doubt!


I should add that using bridge rectifiers effectively doubles the current rating but that only 4 individual diodes could be be used. Two pairs, each pair being two diodes wired in series, that are then wired in parallel with the polarities reversed. The reason for using diodes in series is to increase the volage that is blocked to around 1.3 instead of half that.

 

[rogerthebodger]

VicS is correct the important thing is the forward current rating of the diodes.

If you only have a current limit breaker the current rating must be at least twice that current.

If you have a RCD (earth leakage trip) the diodes can be very small 1 amp or so (wait to be shot down)

I have designed and made lots of eletronic kit for my yacht but I purchased one from there people

I waited until the price was right and bid. They come up all the time as he manufactures.

 

[andrew215]

Thanks for the info. will keep looking.

 

[VicS]

[ QUOTE ]
If you have a RCD (earth leakage trip) the diodes can be very small 1 amp or so (wait to be shot down)

[/ QUOTE ] In theroy that is true but RCDs can fail and they do not, at least in my experience, fail safe. They just dont trip when they should. They are after all dependent an an electronic circuit. That, I think, is why GIs have the high current rating that they do.

If you designed a GI to just withstand the trip current (30mA perhaps) they could be very small devices utilising 4 of the smallest rectifier diodes. In reality they are hefty gadgets with ratings of tens of amps.

Shot down?

 

[Frontier]

I bought one like the one in Rogershaws reply. I think I read they need to be able to handle 70 amps. Hence the big heatsink. Its not worth making one really,. I suppose you could make a heatsink and get some high current diodes for a lot less, but then there is your time, and it may no look as professional as the manufactured one. (well not if I made it /forums/images/graemlins/grin.gif )

You dont really want your earth wire to break in the event of failure so you need high current diodes.

 

[rogerthebodger]

Yes Vic

I just wished to point this out.

In the real world when pluging in to an unknown shore supply I would not rely on the supply having a RCD or even a circuit breaker and if so what the current rating is so the safest route INHO is to have youe own RCD and circuit breaker / fuse and have the galvanic isolator like I purchased is cheep enough to go for one that will work safley in most shore power instulsations.

If you search on the US ebay you will find lots more different units

One of the problems is ensuring the unit is not open circuit and this can be done with a simple neon between live and earth as was discussed in a previous thread.