Toroid question

[Dr.EM]

I have seen this comment quite a few times now and still don't really understand it:

"N.B. Under no circumstances should both ends of the fixing bolt be allowed to simultaneously come in contact with metal chassis or framework so that an electrical path is formed through the screw in the centre of the transformer via the external framework. This would result in a short, and cause irreparable damage."

Source:
**broken link removed**

It seems to imply that each end of the bolt cannot join to a metal chassis electrically or a short will occur. However, the bolt itself is isolated electrically from the transformer so how is it that there could be any danger in this? Is a current induced through the bolt?

 

[Nigel Goodwin]

It places a sorted turn through the core, just like placing a wire loop through the core.

Follow this process:

Bottom of bolt, connected to metal chassis bottom.
Metal side, connected to metal bottom.
Metal top, connected to metal side.
Top of bolt, connected to metal top.

This gives a complete shorted turn around the core - SERIOUSLY bad news, don't do it!.

 

[Dr.EM]

Ok, that makes sense, thanks. The metal chassis can still be earthed though can't it? It will have to be really.

I don't have much clearance at the top at all (1mm really) so it will have to be insulation tape, although the casing is anodised and doesn't seem to conduct. Either way, the short would just blow the fuse (not that I intend for this to happen of course)?

 

[Nigel Goodwin]

Ok, that makes sense, thanks. The metal chassis can still be earthed though can't it? It will have to be really.

I don't have much clearance at the top at all (1mm really) so it will have to be insulation tape, although the casing is anodised and doesn't seem to conduct. Either way, the short would just blow the fuse (not that I intend for this to happen of course)?

It would 'probably' blow the fuse, but might damage the transformer before it does! - it's something to be VERY careful with.

As an alternative?, have you considered 'cable tying' it down?.

 

[zevon8]

On bigger torroids you may end up with a really hot bolt. I seen a good example of this where somebody made a bass guitar amp using one of those FET amp modules and a big torroid power supply. They used two " U -bolts" to hold the torroid down to the rack chassis. Bad idea. Lotsa smoke.

Nigel has a really good idea, get 3 of those heavy duty cable ties, lash it down at 120 degree intervals.

 

[eblc1388]

One would requires 6 additional holes on the chassis to apply the three cable ties, and this cannot be considered to be a long term satisfactory mounting solution. Personally I don't trust the use of sticky pads as cable tie anchor for a heavy transformer.

One can also buy special kits designed to mount the toroid, which use a circular metal plate with a "dimple/depression" in the center to receive the mounting nut so the whole mounting adds only the thickness of the circular metal plate to the original height of the torroid and it can also easily be insulated from the top chassis than using a flat plate and a protruding screw and nut.

 

[oldtimer]

Just use a plastic (nylon) nut and bolt or if the centre hole is large use one but pour epoxy resin around it .

 

[Dr.EM]

I originally wanted a mounting where it was all internal and no fixing could be seen on the outside (something like epoxy) but have given up on that idea and have altered the orientation of the unit so the bolt head is underneath.

The mounting eblc1388 describes is supplied with the transformer and is probably what i'll use. I just need to make sure the other end of the bolt doesn't touch the case (electrically at least). The voltage induced would be tiny wouldn't it, but has the potential for very high currents. A bit of plastic insulation stuck to the top plate of the case and insulation tape on the toroid should keep things safe.

I did consider a nylon bolt yeah, and may still use one if I can find small quantity (100 x M6 50mmm bolt + nut would be pricey). The transformer is 120VA btw.

 

[Oznog]

Why choose a toroidal transformer BTW?

 

[Dr.EM]

Compactness (especially in height) was nessecary. I will post pics once the unit is done

 

[JimB]

I have seen some toroidal mains transformers where the centre is filled with epoxy, and set into the epoxy is a threaded brass bush which will accept a mounting screw.

JimB

 

[Dr.EM]

This range of much smaller ones seems to be:

**broken link removed**

At the moment i'm a little concerned at the output voltage of the transformer. I am reading 16.7v from the 15v unit (unloaded). With the quoted 10% regulation, 16.5 should be the max. I rectified to DC and applied a load to simulate the quiescent current of the IC and read 22v, the absolute max to be applied to the chip (TDA2030AV). I was using a half wave rectifier though and a bridge will be used in the unit, so I can expect to loose a little bit more in that. It is a bit close though.

Actually just realised i'm powering 2 ICs for stereo, so I should loose more voltage there. It is possible my multimeter isn't all that accurate too (seeing as how the range changes at 20v, i'm reading right at the bottom of the 200v range).

 

[oldtimer]

You could try a resistor and Zener and crowbar the output.

 

[Dr.EM]

I was thinking of using a silicon diode in the supply of each amplifier (4 in total). I know this doesn't protect as fully, but i'm hoping the output will never exceeed yesterdays reading by much:

**broken link removed**

 

[Nigel Goodwin]

I would just try it 'as is', presumably what you're looking for is the maximum output from the TDA2030's? - to do that you need maximum supply voltage. As soon as you start drawing power the supply will fall a fair amount, so the voltage won't be a problem then!.

You should bear in mind that the voltage limitation probably refers to high levels of output anyway, and not under no load conditions.

Imagine the output transistors are only rated at 20V Vce, and your supply is 22V off load - the two transistors in series will stand 40V (20V each), but under full drive each transistor will alternately get the full supply voltage (22V). But under these conditions the supply voltage will fall, and it will be under 20V - so no problem. Bit of a simple example?, but the principle holds!.

 

[audioguru]

The TDA2030A needs a dual polarity supply. Don't you have a center-tapped 30VAC transformer?
Doesn't your mains jump up in voltage when the air conditioner or stove turn off?
The absolute max supply voltage is 44V without any regard for loading. I wouldn't exceed it. When the supply voltage is about 71% of max then the amplifier's power is down only 3dB (half power) which isn't much of a loudness drop but the reliability is improved a lot.

 

[Dr.EM]

Yep, I am using a dual polarity 15+15VAC transformer. It's just that is gives quite a high voltage output (16.7VAC per tap apparently). Most things in the house were off at the time so mabye that is partly why. I think that I will use the 3A diodes in all the rails and possibly have a fixed mimimum load (power resistor) just to make sure. Should stick around 21v then whilst quiescent.

I guess it depends on why the max voltage is specified, if it's because the semiconductor (any of them) breaks down over that voltage then it must not be exceeded. If it is because overheating is possible then I wouldn't need to worry.

 

[audioguru]

The power dissipated by the IC causes it to get hot, not the supply voltage.
If it gets too hot, it automatically shuts down. If its supply voltage is too high, it gets permanently zapped.

 

[Nigel Goodwin]

Doesn't your mains jump up in voltage when the air conditioner or stove turn off?

He's in the UK, mains voltages generally don't flucuate here, we have a sensible voltage and good electrical systems

The use of ring mains in a house mean there is good current capacity in the circuits, and each item is individually fused via a fused plug. Larger items (such as emmersion heaters, showers, and cookers) are wired directly to the fusebox on their own individual cables (heavier ones), and have their own fuse in the fuse box.

Certainly to make the lights flicker you have to do something SERIOUSLY!! bad, and probably NOT on one of the ring mains - hitting underground cables with a JCB usually does it though?.

Air con is also VERY rare in the UK.

 

[Hero999]

we have a sensible voltage and good electrical systems

The Americans do too, large appliances like cookers and air conditioners are normally run from 240V AC.