That's it. In your case the 3 A is the secondary current per rail.
In other cases you have the power rating, e.g. 72 VA, divide that by the amount of secondaries which gives 36 VA per winding and divide that by the nominal output voltage, e.g. 12 V which also leads to 3 A.
For a transformer secondary you can use a slow-blow fuse with the nominal current rating, so T3A.
You usually need those graphs, if you want to protect semiconductors.
Some manufacturers make your life easier by stating an integral rating in A²s for fuses which you can also find e.g. in the datasheets of rectifiers. Then you only need to choose a fuse with a rating lower or equal to the rectifier's. Or a rectifier with a higher or equal rating to the fuse's.[/QUOTE]
I agree, but Pacific is trying to persuade us to do otherwise.
Let's hear what he has to say.
I have been working as an electrical engineer for several decades now and I have yet to see a fuse blow from fatigue, although I know the theory behind it. I have however several times seen old fuses carry much more than their nominal current for long times without blowing. Another fatigue mechanism?
But I will try an estimate. We have a 12 V 3 A transformer. A 25 V 22 mF capacitor has around 30 mOhms, the 72 VA transformer has probably 0,6 Ohms. 12 Veff corresponds to 16,92 Vpk (let’s ignore the reduction due to the rectifier), which gives a peak inrush current of ~27 A for less than 10 ms (first half cycle) which is 9 times the nominal fuse current. The datasheet from Eska for 5x20 mm T 1-3,15 A fuses specifies the pre-arcing time limit for 10 times the nominal fuse current to be 10-150 ms. So even the 22 mF capacitor will leave a 10 % safety margin for the worst case fuse, while the average fuse will withstand that current for much longer and will never blow due to aging.
In real life you will find that the voltage reduction due to the rectifier and the additional resistances from wires, traces and the fuses themselves will even reduce the actual inrush current to below the recommended 25% average temperature derating margin.
Shorts in the secondary circuit with low enough impedance will cause a primary fuse to blow. Some shorts with high impedance will not cause the primary fuse to blow. Overload will only trip the primary fuse if you use a soft-start arrangement that allows you to use a tightly fitted fuse rating.
You can only skip the secondary fuses when you can guarantee that the transformer cannot be overloaded. Commercial amp manufacturers do that by including current limiters.
Chip amps also have current limiters, so you could use a transformer with a current rating greater than or equal to the current limit threshold times the amp channels. E. g. use an LM3886 stereo amplifier. Each of the channels has a current limiter at 8 A, so you can skip the secondary fuse, if your transformer has a secondary current rating of 16 A. For the common 2x18 V transformer that means, you need a 576 VA transformer to skip secondary fuses for a 2x30 W into 8 Ohm amplifier.
Transformers with built-in overtemperature protection are a fine thing, but rarely used. You will rather find them in lighting transformers than in DIY audio.
naah there are no such things as guarantees concern overloading! No one can garantee a current limit circuit (chip amp or otherwise) will work under any fault or other conditions! Sorry Pacific that's s really a silly story about current limits and XFMR sizes, we are talking about PS faults after all here, right. The problem is not so much big transformers but with the small cheap ones ie the winding resistances are high enough not to blow the fuse ie smoldering shorts. The safety approval process, at least for UL listed products is to short the secondary winding on the transformer sample and wait till temps stabilize for the prescribed amount of time (hours) with no appearance of flames, molten metals etc. Think about all the wall worts out there!
All the more reason to use fuses for overload protection.
Depends on who "we" are. I was posting about transformer protection against shorts and overload, no matter where they come from.
Smoldering shorts won't happen with a right-sized overload protection by secondary fuses rated for the transformer's nominal current.
Most transformers will not work for several hours with their secondary shorted. One of the windings will assume the role of a fuse, melt and go open circuit.
Such an unreasonable procedure does also not sound at all like anything UL engineers would do, at least not the ones I have dealt with up to now. Or maybe you witnessed the test of a welding transformer?
yes most, but not all
A good outcome with a shorted secondary would be fusing open of the normally designed in primary safety fuse. or failing that, the transformer wire itself , hopefully quickly! There are UL tests that require this, and under any condition there cannot be combustion . Sometimes the shorts do not fuse open a transformer winding very fast at all, these are usually the very small ones. Even if you decide to use external secondary fuses you are not exempt from the testing. full stop AFAIK XFMRs wth built-in internal fuses w/ approvals are the exception.
I am still reading Pacific's replies.
He has not convinced me that inserting T rated fuses between the secondary and the smoothing capacitors offers any additional safety.
I still think that fuses in this location need to be so big that they will not rupture quickly enough when a fault does occur. In the meantime any one or more of the Primary fuse and the supply rails fuses and the mains fuse will have ruptured when the abuse and/or fault occurs.
He has not convinced me that inserting T rated fuses between the secondary and the smoothing capacitors offers any additional safety.
I still think that fuses in this location need to be so big that they will not rupture quickly enough when a fault does occur. In the meantime any one or more of the Primary fuse and the supply rails fuses and the mains fuse will have ruptured when the abuse and/or fault occurs.
I never claimed the opposite, but it is also true the other way round. Even if you have tested the transformer for UL listing, you are not exempt from installing adequate protection measures which include secondary fuses most of the time.
The fuses go between the secondary and the rectifier.
I am not trying to convince you of anything. I am describing the way transformer protection is taught at professional schools and the way it is done in industrial environments.
The following link is from a transformer manufacturer’s homepage http://www.block-trafo.de/assets/data/409/18608/short_circuit_protection_table_etc___0_48_mb_.pdf. Please note the hint on page 4 of the document:
If you leaf through the document, you will find that the primary protection leads to a much higher power rating than the secondary protection. That is because
- the primary protection must be big enough to take transformer losses into account.
- the primary protection must be big enough to take reactance into account.
- the primary protection must be big enough to withstand the inrush current.
Secondary fuses avoid those three pitfalls.
You can always test it.
How can rail fuses according to your layout blow when a short or overload occurs in the power supply?
They cannot, because the fault is upstream and they can only react to downstream faults.
How can mains fuses blow before primary fuses?
How can primary fuses blow before secondary fuses?
The same answer applies to both questions. You ignored selective fuse coordination in your design.
that's news to me I have never seen that in consumer or industrial gear. Sounds like some weird requirement for instrumentation not UL. cite please. I stopped working in power in 1986, About a year prior I had developed a custom SMPS for Tektronics. I had to work with UL to get it listed, the UL engineer was greener than me, so he had us going off on tangents from time to time, but luckily had help from a Tek senior fellow who pretty much wrote the book on safety, step in when needed. I don't think much has changed since peeking in from time to time.
Your answer to AT sounds like your are making a case against primary protection more than a case for secondary.
but I'll see or add to his answer later.
Last edited:
post630.
What is "selective fuse coordination"? Did you mean discrimination?
What? I did not ignore it. I couldn't have, if I don't know what you are referring to.
The rest of your post is logically flawed. I cannot draw any sensible conclusions/rules from the contradictory statements in there.
What is "selective fuse coordination"? Did you mean discrimination?
What? I did not ignore it. I couldn't have, if I don't know what you are referring to.
The rest of your post is logically flawed. I cannot draw any sensible conclusions/rules from the contradictory statements in there.
I have been working with industrial gear for several decades now and I always see it done that way. Circuit-breaker before the transformer for lead protection (short-circuit protection only) and another circuit-breaker or a fuse right behind the transformer for its own protection against shorts and overload.
Both are required.
Yes, seems that I found the American translation instead of the British one.
Circuit-breakers are available with nominal current ratings from 0,1 A upward, maybe even smaller. No kVA there. Of course it makes sense to replace the circuit-breaker with a fuse for your amp power supply for reasons of physical size.
You can see the products I am talking about in the link in post #630. They range from 20 VA upward.
Have a look beside my post. That is the German flag you see there. I am not working according to UL, but according to IEC, EN and DIN regulations.
And your assessment is wrong. The way I describe transformer protection is the quasi standard in Germany. So no advantage for competitors.
There is no regulation that obliges you to work that way. If you deal with this stuff on a regular basis you know that regulations tell you what you need to achieve, not how to achieve it. So if you can proove that a different arrangement is equally safe, nobody will forbid you to use it. You will however have to expend more effort to achieve that, which makes it unattractive for most companies.
Take another look into the link in post #630. On page 6 there is an unregulated power supply (GNC). The datasheet recommends a circuit-breaker for the primary which you can of course again replace with a fuse. In the right two columns it lists the fuses/circuit-breakers that are installed inside the power-supply. Lamentably they don't provide a schematic, but they do install the fuse/circuit-breakers between transformer and rectifier. That company has a far too big output to be wasting money on unnecessary fuses. Which leads me to the conclusion that they also found this arrangement to be the most economical way to comply with all applicable safety regulations.
OK didn't mean to offend, it was sexist too. lol replace your "UL guy" with "your representative for the intended markets safety organization". or VDE gal
So I still think after reviewing the sales PDF that additional secondary over load protection goes above and beyond what is sensible and practical for consumer and light industrial equipment and you don't cause it's in "the code". BTW what does this do to wall warts in Germany?
Last edited:
Most wall warts use transformers with such a high stray inductance that they are virtually short-circuit-proof. You wouldn't want such a transformer in your amp, though, because high stray inductance means high regulation, too.
Most wall-warts are cheap, so you cannot rely on them being sufficiently well designed. Therefore a built-in temperature fuse is compulsory in them, at least in IEC member countries. If you are lucky it is of the self-resetting type, otherwise each overload or short renders the wall-wart useless.
So, how do you go about transformer protection when the inrush current makes it impossible to right-size the primary fuse for overload protection?
Most wall-warts are cheap, so you cannot rely on them being sufficiently well designed. Therefore a built-in temperature fuse is compulsory in them, at least in IEC member countries. If you are lucky it is of the self-resetting type, otherwise each overload or short renders the wall-wart useless.
So, how do you go about transformer protection when the inrush current makes it impossible to right-size the primary fuse for overload protection?
In-rush is a transient condition, so most SB fuses, breakers, or XFMR heating doesnt respond until an average level over their time constant is exceeded. if the surge levels under the time curve exceed some primary protection or for that matter any other component in the path , like switches , rectifiers, etc. THEN Adding slow start circuitry is generally a good design practice this is rarely a safety concern but more of a reliability issue.
In otherwords design the primary current overload to the XFMR max ratings, not to a transient event. Turn on surges are mainly because of low Req not Leq. in larger > 500-600 VA EI cores and less 30% larger toroids.
I have a 140r soft start feeding the 240Vac primary of a 160VA toroid transformer.
An F1A fuse feeding both the soft start and a direct feed to a 3VA EI has not failed yet due to nuisance blowing (fatigue).
If I had some T500mA and T800mA fuses, I would be reporting on those as well.
The secondary fuses (4off), after the smoothing caps (+-20mF), feeding two dual polarity amplifiers are F1.6A, and they too have not blown.
An F1A fuse feeding both the soft start and a direct feed to a 3VA EI has not failed yet due to nuisance blowing (fatigue).
If I had some T500mA and T800mA fuses, I would be reporting on those as well.
The secondary fuses (4off), after the smoothing caps (+-20mF), feeding two dual polarity amplifiers are F1.6A, and they too have not blown.