3 toroidal transformers in paralel ?

[hpa4]

Hello,
I'm currently building a 6 channel amplifier made out of 3 stereo amps in one box. Each amplifier needs a toroidal transformer to get +12/-12v.
I got 3 toroidal transformers (same model) and each amp works very well and does the job so far, it is now time to box everything!
I was wondering if it was possible to power all them 3 transformers from only one 220v power socket (basically connected to the wall, see the image). I don't see why it would be a problem, but as it is the first time I'm dealing with a "custom power supply" I prefer to be sure that my device won't be dangerous. The box is made out of metal and will be grounded by a wire connected to the G of the 220v power socket (i noticed industrial racks using this method with the green and yellow wire screwed in the box).

 

[Nigel Goodwin]

No problem at all, I presume they are of 'reasonably' low power - and not 2KW each or anything silly

 

[hpa4]

so my original assumption was correct, thanks a lot ! nothing silly, 24w per transfo

 

[Diver300]

With toroidal transformers in a metal box, make sure that the mounting bolt doesn't touch the case or lid at both ends. It's fine for the mounting bolt to connect at one end.

 

[hpa4]

Yes it will be mounted only at the bottom. I would be curious to have a deeper explanation of the phenomena (i've read it could melt or so) when it is connected bottom/top, if anyone have time to explain me I would appreciate very much !

Thanks anyway for being sure I'm staying safe

edit: I also have another question, if I wanna have a DC 5v in paralel as well, I can wire the circuit it to the 220v socket ? let's say to power a raspberry in the box for example (theorical here, the amp is full analog).

best

 

[Diver300]

It's quite easy to add turns to a toroidal transformer. This post https://www.electro-tech-online.com/threads/kef-psw2500-active-base-unit-repair.159244/#post-1380354 shows where I did that deliberately to give an additional output.

A 24 W transformer will give something like 0.1 - 0.2 V for a single turn.

If the bolt touches the metal case, it can form a secondary winding, with one turn, with very low resistance, so a lot of current can flow. I think that with a 24 W transformer wouldn't melt a bolt, but there would be a big increase in primary current and the transformer would get hot, and the voltage of the main output would be reduced.

 

[Diver300]

You can put as many things as you want in parallel, up to the rating of the 220 V socket. Having the transformers and a 5 V supply is electrically equivalent to plugging each one into a power strip.

Transformers and power supplies, especially toroidal transformers, can take a lot more current for a short turn-on surge. Also a 24 W transformer can take more than 24 W as input power, as it will be less than 100% efficient. However, three 24 W transformers and a small 5 V power supply will be taking far less than 1 A in total, so a 16 A plug will provide more than enough power.

 

[Nigel Goodwin]

It's quite easy to add turns to a toroidal transformer. This post https://www.electro-tech-online.com/threads/kef-psw2500-active-base-unit-repair.159244/#post-1380354 shows where I did that deliberately to give an additional output.

A 24 W transformer will give something like 0.1 - 0.2 V for a single turn.

If the bolt touches the metal case, it can form a secondary winding, with one turn, with very low resistance, so a lot of current can flow. I think that with a 24 W transformer wouldn't melt a bolt, but there would be a big increase in primary current and the transformer would get hot, and the voltage of the main output would be reduced.

And the transformer would burn up - a thick single shorted turn round a toroid will pass a LOT of current, and the transformer would seriously overheat and very quickly die. If there's a reasonably sized mains fuse in the circuit, that could well save it - but it depends on the rating of the fuse, and the actual overload current involved.

Obviously it's actually worse in this case, as he's paralleling three primaries - and 'presumably' with only a single fuse - in which case the transformer could take at least three times it's maximum rated current, and not blow the fuse.

If anyone has a sacrificial 24W toroid?, perhaps it would be interesting to measure the mains current with a large single turn round the core?.

 

[Diver300]

If anyone has a sacrificial 24W toroid?, perhaps it would be interesting to measure the mains current with a large single turn round the core?.

It doesn't need to be sacrificial. According to this:- https://www.farnell.com/datasheets/2712881.pdf a 30 VA transformer has about 5 W of copper loss. Assuming that half is in each winding, that would be 2.5 W in the primary. The primary current for 30 VA is about 150 mA, which would give about 110 Ohms to cause 5 W of heating. If the transformer is completely shorted, there would only be just over a couple of amps flowing, and 500 W of heating. I know that is 100 times the full load in the transformer, but it would survive a few seconds from a cold, long enough to take a current reading.

Also, you could just measure the primary resistance to get an idea of the worst case.

 

[hpa4]

Thanks a lot for these explanations ! it starts to be more clear to me ! Ofc I can power a circuit until I reach the capacity of the power outlet, as coming from the same wall socket it's like everything is plugged together.

Silly questions then:
why don't we use plastic to mount the transformers in order to maintain attach them to the bottom or side of the enclosure, is it because metal is more robust ? Either, the top or bottom part of a device could be a wood or plastic panel for example ? (but then there is higher risks of fire I assume)

Indeed I didn't thought about installing a fuse, do you recommend doing it and if yes, where would be the optimal placement, pre-transformers I suppose ?

best

 

[Diver300]

There's very little voltage between the top of the bolt and the lid of the housing. It's only one turn, so just a small voltage, so there is no electric shock risk from the mounting bolt. It's very common to have exposed points - look at car batteries. Good designs have a gap to the lid so there is no problem. A plastic bolt would have to be much larger.

The reason that there is a warning is that there is nothing to warn that there is a danger. You could look at a commercial design with a toroidal transformer, like the one in the link that I posted earlier, and not realise that there must be a gap. There is no insulation other than the air, so nothing to see, and the voltage is so small that you can touch the bolt and you won't feel anything. Also it can be that the lid causes the short, so it could be a fault that goes away when the lid is removed. It is also difficult to protect against the over-current caused, because the extra current may not be very much, and the fuse has to take the turn-on surge of the transformer.

A fuse may help, but it is difficult to protect a transformer with a fuse. If you followed the calculation in my last post, there is less than 2 A at the input if the transformer is shorted out. So your fuse would have to be less than that to be of any use at all. There will be a turn-on surge when the transformer is turned on, which the fuse has to survive, and it would have to survive three transformers being turned on at the same time, but protect against one being overloaded. Transformers are extremely reliable if they are not overheated, so a fuse is protecting against something that will never happen. A fuse on the output of each transformer may be of more use. The ELCB that protect the socket will protect you better than an input fuse.

Many years ago, I saw the effects of poor fuse choice on a transformer. Some electrical equipment was supplied at 415 V and needed 110 V for a control circuit. A 415 to 110 V, 50 VA transformer was used, protected by 5 A fuses on input an output. The transformer, which was designed to supply about 0.5 A, supplied about 5 A when shorted, so the output fuse never blew, and the transformer burned out a few minutes later.

 

[hpa4]

My device will be 2u high, and the transformers are only 1u high, so the air gap of 1u should be enough here!

About the fuse, maybe for this circuit I will leave it like it is then, because it seems innefficient in my design. I'm planning to make an 8ch amp after this one, with an eq, and eventually only one transformer with same voltage but more current. Maybe a fuse will be interesting to integrate in this circuit, for the sake of learning/experimenting. I've read a bit about fuses since my last post, I read that it should be able to work up till 150-200% of the "normal" power.

Funny that I didn't remembered they existed, I have the feeling they are the big absent from basic electronics tutorials online, however it is always the first thing to check if equipment doesn't work.

Thanks a lot to both of you for your contributions here, it is a great help on this project.

Best

 

[Nigel Goodwin]

Thanks a lot for these explanations ! it starts to be more clear to me ! Ofc I can power a circuit until I reach the capacity of the power outlet, as coming from the same wall socket it's like everything is plugged together.

Silly questions then:
why don't we use plastic to mount the transformers in order to maintain attach them to the bottom or side of the enclosure, is it because metal is more robust ? Either, the top or bottom part of a device could be a wood or plastic panel for example ? (but then there is higher risks of fire I assume)

Indeed I didn't thought about installing a fuse, do you recommend doing it and if yes, where would be the optimal placement, pre-transformers I suppose ?

best

It's absolutely essential to fit a mains fuse, immediately where the mains enters the unit.

Even more so outside the UK, as most other countries don't have fused plugs - and even in the UK someone could replace the smaller plug fuse with a 13A one!.

 

[hpa4]

Hey Nigel, thanks for the answer, I spent a few day away from internet
So it is mandatory that I change my design, I get it!

I need a fuse in there. I can read online that I calculate by dividing the power used by the voltage of my circuit, I am not sure of the P for this circuit before the transformers tho. I know that 220v are going in, and each transformer is outputing 24V / 2A, from there not sure about the amperage

Also I'm not quite sure, if anything wrong happens, it' just my circuit getting damged, or my house going on fire? cause there should be a breaker at some point I guess. It's my first 220 circuit so I'm a bit anxious :/

 

[Diver300]

If you have a fuse that is too small it will blow. With the turn-on surge that you get with toroidal transformers, it will most likely blow as you turn on. If the fuse is only a little bit too small, it may be weakened by the turn-on surge and may blow when the transformers have been turned off and on a few times.

If you have fuse that is too big, and nothing else goes wrong, there is no consequence.

If you have a fuse that is too big, and a transformer shorts out, then too much current will flow, wires will get hot, and the weakest link in the supply will blow. That might be the cable from the plug, or the circuit breaker feeding the plug, but without detailed knowledge of the wiring the the building, it's difficult to say.

The transformers are 24 W each, I think you said. If that's right, the total power for three of them is 72 W, which is around 0.33 A at 220 V. However, transformers aren't perfectly efficient, which means that they take more current than that. There is also magnetisation current that they take even with no load, so you need a larger fuse, maybe 0.5 A or 1 A.

The large current that a fuse has to survive is the turn-on surge, which would be around 2 A per transformer for up to 1/4 of a mains cycle, or 5 ms at 50 Hz. A slow-blow fuse of about 1 A would easily handle the 6 A for 5 ms (https://www.farnell.com/datasheets/2349120.pdf)

You should be aware that the fuse is not always going to protect the transformer if it is overloaded. If a single transformer is shorted, it will only take 2 A. A fuse rated at 1 A can take 20 minutes to blow at 2 A, so the transformer could be damaged before the fuse blows.

 

[hpa4]

Hm okay, so in this case having a pretransformers fuse seems useless for the thing I wanna prevent, and reveals a bad design from my part. I might just do better on the next device then. If theoretically I don't care that my amp gets damaged, there is no problem leaving the unit as it is, without a fuse, according to you?
Nigel's post suggested that every device should have a fuse for safety purpose.

 

[Nigel Goodwin]

Hm okay, so in this case having a pretransformers fuse seems useless for the thing I wanna prevent, and reveals a bad design from my part. I might just do better on the next device then. If theoretically I don't care that my amp gets damaged, there is no problem leaving the unit as it is, without a fuse, according to you?
Nigel's post suggested that every device should have a fuse for safety purpose.

It's folly NOT to have a fuse! - and I don't see where Diver300 suggested otherwise?.

While it might not blow quickly if the transformer, or rectifiers, go S/C depending on the fuse value - there's at least a chance of it blowing - where with no fuse at all there's obviously no chance whatsoever.

Diver300's misconception is that the fuse is there to protect the 'transformer' - this is completely untrue. The fuse is there to protect YOU the user, and any other people in the vicinity. In the case of a serious overload, such as a S/C rectifier or S/C transformer, it will take massively higher current and probably eventually burst into flames. Before this happens the mains fuse should blow - so the transformer has gone, but you're still alive, and your house hasn't burnt down.

THAT'S WHY YOU FIT A SUITABLE MAINS FUSE!!

 

[hpa4]

OK so we are talking about the safety fuse here (main fuse, correct?), to protect the user and not necessarily the machine, for this purpose I should use 1A fuse so it won't blow when I start the device (0.3 to 0.5A when I turn the device on).

I don't really care if something happens to the transformer or amp circuit and damage that, it would be silly but nothing bad/dangerous, this needs to be corrected with another design. I care about the safety of the house/people, so thanks Nigel I got my answer, and will remember to integrate that to my next designs as a necessary first component on the live wire.

 

[Diver300]

I agree that a fuse should be fitted, to protect the wiring, if a transformer fails.

I didn't say that the fuse was there to protect the transformer. I said that it was likely to fail to protect the transformer. A better protection for the transformers would be an additional fuse in series with the output of each transformer.

On one application where it was possible to overload a transformer, I put a thermal switch glued to the casing. That was easier with a bobbin type transformer which has exposed core and a flat surface to glue the switch to.

Many commercial power supplies have thermal cut outs or thermal switches on the transformers.

It is also worth remembering that a fuse can only protect what is downstream of it. If you have a mains lead, and a metal box with a fuse and a transformer, any fault in the mains lead, where the mains lead enters the box, or before the fuse will not be protected. It will usually be the case that a fault will be a dead short, resulting in the circuit protection for the socket tripping quickly.

Using something like this:-https://au.element14.com/schurter/6200-4330/iec-c14-inlet-10a-250vac-qc-panel/dp/2888474 is quite a good way of having mains come into the box. The are versions with switches as well.

 

[Diver300]

It's interesting to note that plug top adaptors like these:- https://uk.rs-online.com/web/p/ac-dc-adapters/4006535/ are still available.

The only protection is a thermal fuse. Upstream of that, the usual circuit breaker would be 32 A in the UK.

Of course, there really isn't much chance of a short between the plug and the transformer, as there is only a couple of centimeters of wire at most, inside a welded plastic housing.