Hello LF'ers,

Saturday, March 25, 2017

  Thanks for all the replies on my heating toroid question, as best as
  i can tell it's not the core per se that's heating, more the 4 turn
  primary winding, made of pure copper, multi strand speaker "figure
  of eight" cable of about 2.5 mm core diameter. It never gets over 65
  to 70 degrees Celcius, in a warm room and then only on WSPR15. I
  have added a small computer processor fan above it and it stays at
  ambient then. I think I am worrying needlessly. the amp draws about
  20 amps from the PSU. Apart from that I have not had one of my FET
  explosions in weeks and it works 100% reliably now.

  To W1VD, thanks for the update re PCB's i would like to buy a few
  more should you get another batch done, so please bear me in mind.

  The cause of my FET failures was either my doubler, or pre-amp, or
  more likely my TS-590's peculiar waveforms at the end of a WSPR
  transmission. i now drive the amp with a simple Hans Summer QRP Labs
  U3S which is PC independent, and shows a clean stop waveform at the
  end of WSPR and OPERA TX periods. I could have had excess agin in
  the pre amp before the doubler, or oscillation, but as the U3S frees
  the TS-590 and PC from TX duties I am happy to just use the U3S for
  now, with a TCXO installed. The FET's, on an uncooled medium size
  heat sink stay around ambient even after long TX sessions, they got
  quite hot when driven by TS-590 at 1mW / pre-amp / frequency doubler
  combination. Appreciate all the helpful responses!


> Gluing two together improves heat dissipation and no change in ratio is
> required. So far I have found no requirement to alter primary turns 
> either until at least 10 turns are reached.
> On 24/03/2017 19:13, Chris Wilson wrote:
>> 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.
>>





-- 
Best regards,
 Chris                            mailto:[email protected]