Another way - a bit more complicated - to reduce heating significantly is to
feed DC not via the center tap of the output XFM but via separate DC chokes.
This would eliminate the DC bias of the core(s).
>From: [email protected]
>[mailto:[email protected]] On Behalf Of Markus Vester
>Sent: Friday, March 24, 2017 8:48 PM
>To: [email protected]
>Subject: Re: LF: Stacking toroids
>Yes Chris, this will work. Doubling the core cross section
>will halve the flux density, and reduce the per-core
>dissipation by at least a factor of four. The improvement may
>be more if the flux was so high that losses were dominated by
>hysteresis (up to a factor of 8).
>73, Markus (DF6NM)
>Von: Chris Wilson <[email protected]>
>An: rsgb_lf_group <[email protected]>
>Verschickt: Fr, 24. Mrz 2017 20:15
>Betreff: LF: Stacking toroids
>Hello LF people,
>If I have a ferrite toroidal output transformer for a push
>pull Class D 136kHz amp, that gets warmish after a WSPR15
>session, what would happen if I wound the same number of
>primary and secondary turns on two cores stacked together,
>all else remaining the same? Would it run cooler, but
>retain its transformation ratio, or would other things
>occur to make it not that simple? Thanks.
>Chris 2E0ILY mailto:[email protected] <mailto:[email protected]?>