[sdiy] transformer question

[Jim Patchell]

   I have been using Torroids for about 15 years now.  While they are 
fairly stiff (voltage wise), they are not all that stiff.  They still 
suffer from the same I-R losses in the copper wires as any other 
transformer.  The only time I ever had any problems with a torroid based 
supply blowing fuses was a 650 Watt supply I did that put out +/- 55 
volts (unregulated) that used 30,000uF capacitors.  I solved the fuse 
blowing problem by putting something like a 10 ohm resistor in series 
with the primary winding with a relay across the resistor.  The relay 
coil voltage (110 volts) was tied across the +/- output.  When the 
voltage got up high enough, the relay trips, shorting out the 10 ohm 
resistor.

   Also, the Torroids I have been getting generally have seperate wires 
coming out for each winding.  This is kind of nice as I can make a true 
single point ground at the power supply.  With a center tapped 
transformer, this cannot be truely done....

   -Jim

Czech Martin wrote:

> Ok, torroids are nearly humless.
> They have low stray inductance. This means they represent a 
> very stiff voltage source, e.g. if you short them
> you'll be sorry. They also tend to blow fuses when turning on.
> For rectifiers with caps this means huge
> peak currents, a lot of noise in the mains supply.
> At least I think so. Ok, we can handle that and use a 
> inductor coil to smooth that current. ANother component.
> 
> ????
> 
> Wouldn't it make sense to use a transformer with a large stray
> inductance in the first place. Ok, it will make stray fields,
> but the turn on current and rectifier peak current will
> behave. OTOH it has a soft characteristic, which means that
> you need more voltage margin, leading to more regulator
> drop/heat.
> 
> ??
> 
> I'm not so sure that a torroid core is always a good choice
> for a regulated dc supply.
> 
> Sure , it will make a difference where the inductance is placed,
> before or behind the rectifier. Certainly not the same.
> Have to think about that again...
> 
> 
> m.c.
> 
> 
>