Calculating fuse for power transformer
- Thread starter warpie
- Start date
Help Support GroupDIY Audio Forum:
ruffrecords
Well-known member
Probably too low. At power on there is likely to be an inrush current as the caps charge up in the power supply driven by this transformer which could well blow a 375mA fuse. I think I would probably go for a 1 amp slow blow. The principle function of the fuse is to stop the device bursting into flames when a fault occurs. Several times the normal secondary current would need to flow for this to occur.
This is a very good question as it is a topic that is rarely discussed.
Cheers
Ian
This is a very good question as it is a topic that is rarely discussed.
Cheers
Ian
warpie
Well-known member
- Joined
- Feb 7, 2009
- Messages
- 1,583
warpie said:
This is confusing the hell out of me, so I hope you dont mind me jumping on so I can learn something/get this cleared up.
Is the question about fusing the transformer itself as opposed to the circuit? And if so, why? All things being equal, going by circuit current draw would blow the fuse before a fuse rated to protect the transformer, so unless were talking different secondaries where current can differ on the taps, that would be redundant - or?
Gustav
The transformer it self could fail, if it fails short for a bad insulation, which could happen for something as simple as bad varnish, windings move around, the insulation fails, you get a short circuit. If that's the case you need protection so you don't get flames output, then you need, always, a fuse as the first point in your mains connection, preferably in the live side so you don't get live chasis in the case of a really bad fail.
Then if you want, you could protect the circuit itself and then you make the fuse fail before the circuit breaks, so you need to replace fuses more often but not transistors...
JS
Then if you want, you could protect the circuit itself and then you make the fuse fail before the circuit breaks, so you need to replace fuses more often but not transistors...
JS
I was answering gustav q about fuses on the pri or sec. You could calculate your inrush current based on the capacitance in your power supply, the first caps after the diode bridge are the mandatory, slow start up or not the capacitance on the other side of the regulator should be lower and always will end with some effective resistance, so for a rough number that should work. I don't know if that's useful since that would be true for the first charge for the caps, and then lowered down, so if using a slow blow you can almost ignore that, as long as it's not turning your slow blow in a fast blow...
There are several calculators out there for power supplies but I don't remember having a fuse recommendation... In any case, you know how much current the circuit will sink in a normal operation, make it a bit bigger than that as starter, then look for one bigger than that which doesn't blow at start up, and you are good to go. You may blow a few trying out but once you did a few you'll start to know how to estimate them. Capacitance is a big factor, since determines the inrush current, so knowing the VA rating of a transformer is not enough to take account of that, since transformers are sloooow blow, they can take the inrush current with no problem (usually) and then you are limited by faster things in the circuit (the rectifier for example, should always be quite bigger than the nominal current of the PS)
JS
There are several calculators out there for power supplies but I don't remember having a fuse recommendation... In any case, you know how much current the circuit will sink in a normal operation, make it a bit bigger than that as starter, then look for one bigger than that which doesn't blow at start up, and you are good to go. You may blow a few trying out but once you did a few you'll start to know how to estimate them. Capacitance is a big factor, since determines the inrush current, so knowing the VA rating of a transformer is not enough to take account of that, since transformers are sloooow blow, they can take the inrush current with no problem (usually) and then you are limited by faster things in the circuit (the rectifier for example, should always be quite bigger than the nominal current of the PS)
JS
ruffrecords
Well-known member
I found this quite informative:
http://electrical-engineering-portal.com/overcurrent-protection-transformer-nec-450-3#1
It seems that one switch on, the core of the transformer briefly saturates, so that there is an inrush current that peaks at about 25 times the rated current. That's why you need a slow blow fuse. Thereafter the recommendation is the the fuse be rated at about 110 to 125% of the nominal rating which is close to what you did.
The final point it makes is that secondary fusing is to be considered completely separately from primary fusing.
Cheers
Ian
http://electrical-engineering-portal.com/overcurrent-protection-transformer-nec-450-3#1
It seems that one switch on, the core of the transformer briefly saturates, so that there is an inrush current that peaks at about 25 times the rated current. That's why you need a slow blow fuse. Thereafter the recommendation is the the fuse be rated at about 110 to 125% of the nominal rating which is close to what you did.
The final point it makes is that secondary fusing is to be considered completely separately from primary fusing.
Cheers
Ian
Harpo
Well-known member
This was my wording for a 30VA (not 50VA) toroidal transformer to a member from this forum. Might as well be useful for you ...
>> dunno what fuses your kit contains, so just running some numbers for the (mains voltage and type of transformer depending) blowing fuses issues from the xxx thread.
To my knowledge you are selling the 30VA Avel Lindberg transformer from Y23 range (84.8% efficiency).
Including a +/-10% mains variation and about room temperature inside the box
for 115VAC mains/60Hz my calc comes up with 0.37481A rating needed.
Next avail. standard parts rating with slow blow (/t) fusing characteristic would be 375mA or 400mA.
For european 230VAC mains/50Hz my calc comes up with 0.18741A rating needed.
Next avail. standard parts rating with slow blow (/t) fusing characteristic would be 200mA for this type mains transformer.
Min. mains side fuse rating (in front of the mains transformer) =
[Transformer VA] / [AC mains voltage] * ( 1 / ( 1 - 2 * (1 - [transformer efficiency %] / 100 )))
IE 30/115*(1/(1-2*(1-84.8/100)))=0.34781259A according to IEC standard for your Avel Lindberg type
(increase fuse rating by factor 1.33 if fuse is designed according to UL standard for continuously operating at 75% of rated current of the fuse).
Rough approximation is transformer VA / VAC mains * [1.3 ... 1.6], depending on efficiency.
Lesser VA transformer come with lower efficiency and vice versa. A transformer without core and copper losses is still not invented.
The time lag/slow blow fusing characteristic for a 400mA/t non resettable fuse from ESKA or Shurter has a melting integral A²s of 0.4 for the inrush current peak for the reservoir caps to initially charge to normal operating current.
These up to 6.3A fuses have a sustained dissipation of max. 1.6W (look up your fuse datasheet), so the fuseholders power acceptance value should exceed the fuse dissipation value or it might melt/burn.
Fuses are rated at 23°C room temperature. With increasing ambient temperature the derating factor is about 10% per each 20°C temperature increase, so the 375mA rated fuse @ 23°C would need an increase to 375mA*1.2=450mA @ 63°C. Just look up the derating curve for your type of fuse.
>> dunno what fuses your kit contains, so just running some numbers for the (mains voltage and type of transformer depending) blowing fuses issues from the xxx thread.
To my knowledge you are selling the 30VA Avel Lindberg transformer from Y23 range (84.8% efficiency).
Including a +/-10% mains variation and about room temperature inside the box
for 115VAC mains/60Hz my calc comes up with 0.37481A rating needed.
Next avail. standard parts rating with slow blow (/t) fusing characteristic would be 375mA or 400mA.
For european 230VAC mains/50Hz my calc comes up with 0.18741A rating needed.
Next avail. standard parts rating with slow blow (/t) fusing characteristic would be 200mA for this type mains transformer.
Min. mains side fuse rating (in front of the mains transformer) =
[Transformer VA] / [AC mains voltage] * ( 1 / ( 1 - 2 * (1 - [transformer efficiency %] / 100 )))
IE 30/115*(1/(1-2*(1-84.8/100)))=0.34781259A according to IEC standard for your Avel Lindberg type
(increase fuse rating by factor 1.33 if fuse is designed according to UL standard for continuously operating at 75% of rated current of the fuse).
Rough approximation is transformer VA / VAC mains * [1.3 ... 1.6], depending on efficiency.
Lesser VA transformer come with lower efficiency and vice versa. A transformer without core and copper losses is still not invented.
The time lag/slow blow fusing characteristic for a 400mA/t non resettable fuse from ESKA or Shurter has a melting integral A²s of 0.4 for the inrush current peak for the reservoir caps to initially charge to normal operating current.
These up to 6.3A fuses have a sustained dissipation of max. 1.6W (look up your fuse datasheet), so the fuseholders power acceptance value should exceed the fuse dissipation value or it might melt/burn.
Fuses are rated at 23°C room temperature. With increasing ambient temperature the derating factor is about 10% per each 20°C temperature increase, so the 375mA rated fuse @ 23°C would need an increase to 375mA*1.2=450mA @ 63°C. Just look up the derating curve for your type of fuse.
There are many factors involved. As mentioned elsewhere there is an inrush current at turn-on due to the magnetizing of the core. The inrush current depends on the timing. Counter-intuitively, it is maximum when turn-on coincides with zero-crossing. And the inrush current due to charging the capacitors depends on the actual value of these capacitors, not on the steady-state current. For a given rated current, one designer may decide to put oversec'd capacitors, which will result in a larger inrush.warpie said:Say that I use a 50VA toroidal transformer (efficiency>93%). Is it correct to assume that I need a 315mA fuse for 230VAC?
(50VA/ 230) X 1.5 = 326mA
thanks
Heaters and lightbulbs are another non-linear load that create a large inrush. The cold resistance is about 10 times smaller than the resistance at nominal temperature.
There is no way to calculate a reliable fuse value. basic calculations give a ballpark figure, but experimentation is required.
I strongly recommend the use of CTN resistors; they allow installing fuses of a more sensible value.
quick and easy way-short the sec leads, hook ammeter to pri, turn on xfmr for 5.29 seconds, read ammeter, pick fuse,
no harm will be done to the xfmr in 5.29 seconds,
also, sometimes toroid transformers have built in thermal fuses, so you got that working for you also,
no harm will be done to the xfmr in 5.29 seconds,
also, sometimes toroid transformers have built in thermal fuses, so you got that working for you also,
The ESSENTIAL thing is to protect the line-cord. It is outside the case where a fire is very bad. If you do not have UK-type fused-plugs, the next best thing is a fuse right where the cord comes into the box. This won't protect against a cord-pinch short outside the box (the UK plug would), but does protect against nearly any in-box short.
For all US-type line-cords, 10 Amp fuse will let-go long before the cord burns, and is more than enough for 99% of DIY boxes. I've seen lighter wire on 230V cords, so I'd guess 5 Amps would be safe for those.
Transformers do not smoke fast, but when they do, it is very stinky. And occasionally they can burn. If inside the box or metal shell, this should not burn your house down, but you might rather avoid it. The short-term worst-case current in a transformer may be 10X the rated load. So your "50VA" might suck 500VA for CJ's 5.14159 seconds without self-harm.
"Most" faults will not be 50.1VA, they will be 500VA. A shorted rectifier or main filter cap is a dead-short on the PT, current would be "infinity" except PT self-resistance limits it to maybe 10X normal.
Another complexication: fuse will usually carry 125% of rating forever, but will fail "for no reason" if cycled on/off hundreds of times because of thermal expansion and metal fatigue. Even if you find a fuse which carries your starting surge dozens of times, it may still "wear out" hundreds of gigs down the road.
Slow-Blow is mostly correct for line-cord and transformer protection. You want to pass several times nominal for a few seconds, but cut-out before fire might start.
I might figure double 0.217A and round it to 0.5A. If you only get a choice 0.3A and 0.6A, maybe flip a coin. Or think about if you are running ALL of those 50VA into BIG caps, or if you only bought 50VA because a 25VA was a hair shy for the job. However 0.63A at 230V is "only" 144 Watts. Since current is likely either >0.21A (max happy) or 2A (dead-short PT), a 0.6A should cover most accidents.
For all US-type line-cords, 10 Amp fuse will let-go long before the cord burns, and is more than enough for 99% of DIY boxes. I've seen lighter wire on 230V cords, so I'd guess 5 Amps would be safe for those.
Transformers do not smoke fast, but when they do, it is very stinky. And occasionally they can burn. If inside the box or metal shell, this should not burn your house down, but you might rather avoid it. The short-term worst-case current in a transformer may be 10X the rated load. So your "50VA" might suck 500VA for CJ's 5.14159 seconds without self-harm.
"Most" faults will not be 50.1VA, they will be 500VA. A shorted rectifier or main filter cap is a dead-short on the PT, current would be "infinity" except PT self-resistance limits it to maybe 10X normal.
Another complexication: fuse will usually carry 125% of rating forever, but will fail "for no reason" if cycled on/off hundreds of times because of thermal expansion and metal fatigue. Even if you find a fuse which carries your starting surge dozens of times, it may still "wear out" hundreds of gigs down the road.
Slow-Blow is mostly correct for line-cord and transformer protection. You want to pass several times nominal for a few seconds, but cut-out before fire might start.
I might figure double 0.217A and round it to 0.5A. If you only get a choice 0.3A and 0.6A, maybe flip a coin. Or think about if you are running ALL of those 50VA into BIG caps, or if you only bought 50VA because a 25VA was a hair shy for the job. However 0.63A at 230V is "only" 144 Watts. Since current is likely either >0.21A (max happy) or 2A (dead-short PT), a 0.6A should cover most accidents.
Yup, PRR is correct that fuses are used mostly to prevent fires. That said many transformers have thermal fuses built in so the transformer will open and shut down before catching fire. Sometimes UL testing will insist on adding extra fuses inside after the transformer secondary, if their testing reveals internal components that can seriously overheat and cause fire without opening the mains fuse. Often fuses are added just to reduce the cost of repairs should common faults occur (like power amp output devices that like to fail shorted). I've owned hifi amps with fuses in series with the output lead to protect speakers/amp from playing too loud too long (not all that hifi).
JR
JR
Points 1, 2, and 3 are absolutely vital in any fuse selection.
Point 4 comes up when working in higher voltages, especially when breaking line power. The peak current in a "dead short" can be very high, hundreds of amps from a wall-outlet. While breaking, an arc forms which could be 100V at 100A. Which is 10,000 Watts. 10 times the power in a toaster, in 1/1000th of the space. It gets very violent in there. The fuse must not go-off like a bomb or spew molten metal. Use your 120/240V-rated fuses for such work.
The "voltage" spec is about blow-up violence. On the secondary of most power transformers the full line Power is not possible. That is why "250V" fuses tend to do OK in the 500V lead of Marshall tube amplifiers. The maximum short-circuit current is just a few Amps, the fuse body is long enough to stand-off 500v after blowing, so they tend to blow without drama.
The other specs: if you have Government or Military specs to meet, you really need to pay an Expert for a documented design.
Point 4 comes up when working in higher voltages, especially when breaking line power. The peak current in a "dead short" can be very high, hundreds of amps from a wall-outlet. While breaking, an arc forms which could be 100V at 100A. Which is 10,000 Watts. 10 times the power in a toaster, in 1/1000th of the space. It gets very violent in there. The fuse must not go-off like a bomb or spew molten metal. Use your 120/240V-rated fuses for such work.
The "voltage" spec is about blow-up violence. On the secondary of most power transformers the full line Power is not possible. That is why "250V" fuses tend to do OK in the 500V lead of Marshall tube amplifiers. The maximum short-circuit current is just a few Amps, the fuse body is long enough to stand-off 500v after blowing, so they tend to blow without drama.
The other specs: if you have Government or Military specs to meet, you really need to pay an Expert for a documented design.
