Re: Toroid ballast design

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "Gerry  Reynolds" <gerryreynolds@xxxxxxxxxxxxx>

Hi DC,

I tried your design using a 0.08 inch gap, 280Vac, and 30amps on my 
program and it too came up with 110 turns.  However, the flux density 
was 24600 gauss.  I believe I physically have room for about 270 
turns in one layer on the core, but will require a larger gap (0.45 
inches).  This also will reduce the flux density to 11000 gauss.  I 
dont know where these cores saturate and if 24600 gauss is no 
problem, the fewer number of turns will mean less winding 
loss.  Maybe you could shed some lite on this.

Thankyou for your response

Gerry R.




> Original poster: "D.C. Cox" <resonance@xxxxxxxxxx>
>
>
> Epoxy both cores together and use them in parallel.  Wrap 120 turns 
> of #8 AWG around both cores.  Tap at 100 and 110 turns.
>
> Dr. Resonance
>
>
> > I just acquired two 120Vac 50A variacs that had a major diameter of 
> > 12 inches and an estimated cross section of 6 sq inches.  The 
> > windings were burnt so I took the core out and removed the windings 
> > and found the actual cross section of core was 3 sq inches.  I was 
> > originally planning on making two toroid ballasts and put them in 
> > series but now I'm thinking of combining the two cores and 
> > rewinding it as one core with 6 sq inches. I'm designing for 280Vac 
> > and using 8awg P180 insulation (180 degree C polyurythene) magnetic 
> > wire and using a 0.45 inch gap in the combined core.
> >
> > The design program I wrote suggest the following performance (I 
> > assumed relative permeability of 2100):
> >
> > 268turns   29.7mh   25amps rms   Bpeak=10114gauss   Winding loss 107watts
> > 245turns   24.8mh   30amps rms   Bpeak=11080gauss   Winding loss 141watts
> > 227turns   21.2mh   35amps rms   Bpeak=11968gauss   Winding loss 178watts
> >
> > The original 50amp variac had a tapered winding.  135 turns of 
> > single wound 11awg and 67 turns of double wound 11awg (2 strands of 
> > 11awg each in parallel).  The two 11awg strands in parallel is an 
> > equivalent of one 8awg which seems OK for 50 amps.
> >
> > Questions:
> >
> > 1. Since the variac was designed for 120Vac in and 0-120Vout,  how 
> > did it get by with a 3 sq inch core at 50amps.
> >
> > 2. With my ballast design using the 6 sq inch core, the peak flux 
> > density is ~12000 gauss for the 35 amp case.  What is the typical 
> > saturation level with these things realizing that the variac is a 
> > very old design using a silk insulation on the wire.  The core 
> > appears to be made with spiral band of material.
> >
> > Gerry R.
> >
> >
> > > Original poster: "D.C. Cox" <resonance@xxxxxxxxxx>
> > >
> > > They will saturate.  That's why when we used to use old variac cores
> > > we always gapped then at .080 inch.  A fabricated steel clamp held
> > > the variac from spring apart and a bandsaw did the work.  An 80 mil
> > > piece of plastic was epoxied in place to form the gap.  They didn't
> > > saturate.
> > >
> > > Now we use our own slug tuned cores with a small DC gearmotor to
> > > accomplish the same control and makes it a completely variable
> > > process.  Current is adjusted completely independent of voltage.
> > >
> > > Dr. Resonance