Fuse size

[Dave Plowman (News)]

Any tables around for selecting fuse sizes?

Specifically, on the mains side of a 30vA 15v toroidal transformer. Presumably anti-surge, but any other fiddle factor required?

[tabbypurr]

the fiddle factor is huge due to inrush. Try 2A. Secondary fusing or thermal fusing can protect against overheating, a primary fuse can't.

NT

[Dave Plowman (News)]

Right. Which makes sense.

It's just that I'd come across two devices with similar power supplies where the mains fuse was very different. But the secondary ones the same.

[tabbypurr]

A fuse characteristic graph will show what size is needed to cope with the inrush. But the inrush may be unknown to the designer, and the graph not co nsulted. And even if it is it wouldn't give the right answer, as fuses run anywhere near their i squared t limit repeatedly fail in time. So ultimatel y falling back on eperience that says '2A doesn't blow, 1A eventually does' is good enough, and is likely to be what one engineer went with.

NT

[mick]

30VA is 130mA at mains voltage (if the transformer is fully loaded), but you have 2 inrushes to think about. The first is the magnetising current of the transformer, the second is the initial charging current of any reservoir capacitor that you might be using if the transformer is being used for a DC power supply.

I'd generally use a x2 safety factor for something like this, making the fuse about 250mA. It would almost certainly need to be a slow-blow, type T probably.

You can get more precise, but you need to do a lot more work. :)

[tabbypurr]

A factor of 2 is way inadequate for a toroidal.

NT

[mick]

Not with a type T fuse. That's the sort of job they are designed for - time delayed blowing. At 2x rated current you're in the realms of 10s to blow. At 10x it'll still be around 200ms.

In this particular case, 30VA is a very small toroid so it's inrush won't be all that big anyway. We don't know what the secondary load is, so the whole thing is guesswork anyway. :)

[tabbypurr]

Toroids have major inrush current, especially when the core was left magnetised the wrong way at previous switch off. IME time delay fuses don't even have enough delay for EI transformers, they still need to be oversized.

Partly

NT

[Dave Plowman (News)]

How about 30vA? ;-) 1 amp per secondary winding?

[tabbypurr]

It's the startup profile of the load that matters

NT

[Dave Plowman (News)]

Yes - but mick said we don't know the secondary load. I do - or rather the maximum.

[tabbypurr]

You'd need to know the transformer's regulation to work that out. Obviously it's several times the 30VA rating.

NT

[mick]

Hehe... It's only 1A per secondary if the load is purely a resistor. A

30VA toroid is perfectly capable of giving several times that for a short time without any damage. That's a good thing when you consider that we tend to rectify the AC and feed a big electrolytic from the half-cycles to get smooth DC. The initial charging current of that capacitor is reflected back to the transformer primary and forms part of the inrush, and hence affects the size of fuse that you need.

Of course, in your case there may be no secondary load when the transformer is switched on. We don't know that. :)

Toroids are a sod to protect with a fuse. Small ones have a pretty low ratio of full-load to overload primary current and sometimes you can't easily find a fuse that is within the right range. They either blow before full load is reached or they don't protect the transformer against overheating. The smaller the toroid gets the worse the problem is.

It's not really safe to overrate a primary fuse, even marginally. It stops giving the protection needed from overload situations - which need not be dead shorts on the secondary. That's why type T and TT (even slower than T) fuses are used - they will almost always blow at somewhere between 1.25x and 2x their rating, but they will take their time about it. The transformer may get hot, but eventually the fuse will blow. Just using a higher rated fuse is like replacing it with connecting wire. It does nothing. The transformer has a maximum primary current that can flow, based on it's inductance. (That may actually be less than the wire size will allow!) If the fuse can pass more than that for any length of time then the transformer will burn out under overload conditions and you may as well not have bothered with the fuse.

Big toroids can have such a huge inrush that they can take out what looks like a well over-rated fuse or breaker. They do usually have a much better primary current ratio though.

[tabbypurr]

There are a few approaches. What mains fuse would you use on a 12VA transformer?

NT

[The Natural Philosopher]

Insufficient data.

Need to know the primary inductance and the gauge of wire on the primary.

Good rule of thumb is 'two gauges smaller than the primary wire'

I'd probably guess at 500mA though = 10 x 12VA roughly.

[mick]

A 500mA type T fuse won't blow at 500mA, it'll probably be between 625mA and 1A before it takes under 20mins to blow. 12VA is about 50mA at 230V. I'd be very surprised if the primary could carry 1A for over an hour without melting (that would mean winding it in something like 20swg wire, which is pretty thick for a primary winding and may not even fit into a

12VA toroid). It's not the outside of a toroid you need to worry about, that cools nicely while the primary underneath it gets hotter and hotter. I'd be surprised if the primary could handle much over 200mA for any length of time. Are you willing to test that? ;)

Once again, there is load info missing. :) I'd start with a 100mAT or

125mAT for a 12VA.

[tabbypurr]

those fuse at what, if we say upto 250mA in 20 minutes that's way beyond th e thermal ability of a 12VA transformer to handle, it's anything upto 60w. So those don't offer thermal protection, that has to be done another way. I normally do it with a secondary fuse. Once you have a secondary fuse you c an then safely use a primary fuse that's much less fragile than 100mA, and is thus reliable.

NT

[mick]

A 100mAT should start to blow at between 125mA and 200mA (this is variable between manufacturers & types of fuse - it's not gospel). Note that fuses have an inverse time though, so the higher the current the faster they blow, but the rate of increase also increases. What you say is true, a slight overload probably wouldn't be seen and the transformer might fry, whereas a bigger overload would take the fuse first. A lot depends on the transformer and load characteristics.

Secondary fusing is very useful, but it's main use is to protect the transformer against short-circuits or heavy secondary loads such as a shorted diode in a bridge rectifier. You still can't overrate the primary fuse as you don't know what sort of current your secondary fuse will have to take. It will certainly have to withstand charging currents of any reservoir capacitors but it must be rated to protect the rectifier too. For that reason it's not usually a type T, as rectifiers can blow faster than fuses! That makes it an oversized, but fast, fuse. You also have to be careful of secondary fuses where there are multiple secondaries as it's very likely that you won't get any primary protection if only one secondary is overloaded.

For small to medium transformers the very best protection is a thermistor, preferably buried in the windings, in series with the primary. It's far better than any fuse.

[tabbypurr]

200mA is upto 48w in a little 12VA transformer.

yup

I can't make sense of that.

it can't. Trying to protect silicon with a fuse of a few amps is futile.

exactly. It's not type T on cost grounds, pure & simple. It doesn't need to be a T. Making it time delay doesn't give any added protection to any part of the system, and it can work without being time delay.

on the contrary the overloaded section will fuse and protect the transforme r. And when the overload is small, it has far less effect on total transfor mer dissipation than with a single secondary transformer.

Oh yes. But it costs more.

Trying to protect small mains transformers with fusing is like having one h and tied behind your back. There is no fuse scheme that confers full protec tion against damage. But that isn't needed since transformers are very reli able & robust, and only fire protection is needed. That fusing can protect against.

NT

[Terry Casey]

I've never forgotten my brother's reaction many years ago when he accidentally connected a 60mA 32mm fuse across the PSU we used for repairing car radios - i.e.: 12 volts.

It didn't blow - just glowed orange like a festoon bulb!

(No doubt the wire would have melted sooner or later if he'd left it connected long enough ...)