Toroidal Transformer Inrush

[Nigel Goodwin]

Ummm have you actually read his site ?

I see nothing of commercial gain for him and the majority of it is tutorials, examples and free software to help out beginners.

Yes, nothing of commercial gain at all, and not even a hit counter - on any of my sites, not just the one in question.

The reason for it been in my signature is that the tutorials were written basically to answer a great many of the PIC related questions on here - and I got tired of continually retyping the answers.

As for actual 'commercial' links in posts or signatures, I only allow them for proven valuable members - anyone posting here just for advertising purposes is deleted and banned.

 

[q5101997]

So, the magnetic field builds up, inducing a back EMF, and restricting current flow. Saturation occurs and the back EMF can't prevent an increase (surge) in current.

 

[q5101997]

Yes, nothing of commercial gain at all, and not even a hit counter - on any of my sites, not just the one in question.

The reason for it been in my signature is that the tutorials were written basically to answer a great many of the PIC related questions on here - and I got tired of continually retyping the answers.

As for actual 'commercial' links in posts or signatures, I only allow them for proven valuable members - anyone posting here just for advertising purposes is deleted and banned.

Point taken.

 

[chrisjpitt]

It's certainly not complex, just a simple timer - a 555 would be an obvious choice.

In any case, that's how it's done and it's commonplace on anything with a large toroid.

The only problem is - you end up having to build a small mains power supply, just to power the transformer anti-surge timer circuit! You could use the rising voltage on the secondary side to turn on the anti-surge resistor shorting relay - but this kind of simple circuit has the drawback that it does not reset if the mains is briefly turned off and back on again.

Another option is to use an off-the-shelf industrial control, mains powered on-delay relay – possibly something like this from RS Components? (Stock No: 300-5856)

This solution would be a bit more expensive than the parts to make delay circuit (and you would not get the experience of building yet another 555 project), but it would save a lot of time?

 

[Hero999]

The impedance of an inductor is ZL = 2(pi)fL where:
f = frequency = 50Hz (in Australia).
L = inductance in Henries.
Being toroidal, the inductance of the transformer might be 500mH, so:
ZL = 2(pi)x50x500m ≈ 157Ω. A bit higher than 5Ω, don't you think?

Your calculations may be correct but your reasoning is flawed.

The surge current has nothing to do with the inductance and everything to do with the DC resistance.

If you measure it with a DVM you'll get the DC resistance.

The inductance determines the magnetising current which will be very small. I would also expect the inductance to be much higher than 500mH, it'll be more like 10H.

the "inrush" current has nothing to do with the winding resistance. the transformer inductance prevents current from flowing when there is no load on the output.

the inrush current is actually a function of the load capacitor charging. there are a few ways to take care of it: severely overrating the fuse to the point of it being useless under some circumstances, adding an inrush current limiter, or ensuring that the power is turned on when the line voltage is at 0V.

Actually a toroidal transformer will have a high inrush even with no rectifier and capacitor connected.

The inrush current occurs because the transformer is not an ideal inductor. When the power is first applied the core isn't magnetised so the inductance of the coil is the same as though the core isn't there. The inductance will only be a couple of mH so the instantaneous current will be huge. If a huge superconducting core-less toroidal were used the inrush current would be very low because it would be a near-ideal inductor.

I am surprised it is 5 ohms, at that level I would suspect it would be lower still.

It depends on the primary voltage rating, if it's rated for 120VAC then I agree, 5R sounds high but not at 230VAC. A 50Hz transformer will also have a slightly higher DC resistance than a 60Hz transformer.

With a 5R DC resistance the effective promary voltage drop at full load will only be 3.4%, if the same is true on the secondary the regulation would be 6.8% which sounds reasonable for a 300VA transformer.

 

[The Electrician]

When the power is first applied the core isn't magnetised so the inductance of the coil is the same as though the core isn't there. The inductance will only be a couple of mH so the instantaneous current will be huge. If a huge superconducting core-less toroidal were used the inrush current would be very low because it would be a near-ideal inductor.

This is not correct.

Transformer cores are made of a material which doesn't retain much flux when the exciting current is absent. So, when voltage is first applied, the flux is nearly zero (that is, the core isn't magnetized), but this doesn't mean that the inductance is low. The inductance is proportional to the slope of the B-H curve, and when the core isn't saturated, that slope is not (nearly) horizontal, as it is when the core is saturated. This means that the inductance of the primary is high when the core isn't magnetized, not the other way around. Saturation is a result of the core becoming as magnetized as it can. It's as a result of saturation that the inductance decreases precipitously; that the inductance of the coil becomes nearly the same as if the core weren't there.

You can see that this is true by examining the first scope photo in my earlier post. The grid voltage is applied just before 2 cm. The current pulse doesn't even begin until about 5 mS later, just before 3 cm. This is because it takes that long for the flux to reach saturation.

If your explanation were correct, the current surge would have occurred immediately, at the 2 cm mark.

 

[The Electrician]

The earlier scope captures I posted don't show unequivocally when the grid voltage was applied to the primary of the toroidal transformer. I've captured a couple more, with the green trace showing the voltage across the primary, instead of the unswitched grid voltage.

I've left the secondary rectifier+caps connected.

Tcur3 shows the current in the primary when the grid voltage is applied at the peak of the sine wave. There's a small current surge of maybe 10 amps due to the charging of the caps, but there is no large surge due to saturation of the transformer.

Tcur4 shows the current in the primary when the grid voltage is applied at the zero crossing (actually, just slightly before) of the sine wave. You can see the nearly 180 amp surge which doesn't even begin until 5 mS after the voltage is first applied. This surge current is so large that it distorts the voltage applied to the primary by causing a substantial voltage drop in the wiring resistance. The DC resistance of the primary in this particular toroidal transformer is about .5 ohms; that, plus the resistance of the house wiring, etc., is what limits the current to 180 amps.

 

[Nigel Goodwin]

The only problem is - you end up having to build a small mains power supply, just to power the transformer anti-surge timer circuit!

It's normally just powered from the secondary side of the transformer you're feeding, there's enough voltage from the transformer even with the resistor feeding it (you design it that way).

It's not complicated, and it's essential for a large transformer - most large PA amps used to use them.