Hello Alan,
Hi Alan, I cannot thank you enough for the reply on the "AMPS" forum
to my questions. I did use 78 material for the output transformers of
the two amps I built to W1VD's specs, and guessed he may not have
updated his pages to reflect this better material for use in the
combiner, but i thought i wouldn't be too clever and change anything.
a subsequent correspondence with Jay, W1VD showed he intended the
combiner to be used with 2 of his 500W amps, so I am pushing the
limits presumably with whatever material. I see Fair-Rite lit a 98
material which is supposedly better again than 78, but getting info on
what sized toroids are available in 98 is proving difficult. Also,
Fair-Rite USA charge insane postage to the UK making small orders very
costly.
Even with the combiner gate and drain waveforms seem good, with a
slightly more spiky edge to the drain ones when combining, so i guess
the saturation level hasn't been exceeded, but material choice is
poor, and maybe turns counts are sub optimal for this power level.
I will try two approaches, a copy of my existing combiner transformer
using spare 78 compound FT-240 sized toroids I have to hand, and
another with the tripled turns count Manfred kindly suggested. The
problem with LF for a 2E0 is the guys running "down there" are all
very experienced and their replies are often a bit over my head,
especially as I foolishly chose to lark about in maths classes, much
to my regret when I took up amateur radio at over 60 years of age.
My old teacher's words never rang so true, "Wilson, one day you'll wish
you had paid more attention, as maths is often needed to do things
that interest you, rather than just at a purely academic level that
you seem to find so unimportant and boring". Damn, he was dead
right.... :(
Much appreciated, and thank you again Alan!! 2E0ILY
> Chris,
> Nobody is answering so I'll have a go. If I muck up the arithmetic
> others will be quick to cut me off at the knees :-)
> I make the max flux density around 160mT (1600 Gauss).
> If you can't be bothered to do the calculation from scratch there's an
> online calculator here
> http://www.daycounter.com/Calculators/Max-Flux-Density-Calculator.phtml
> The data sheet for type 77 material says it has 300mW/cc loss at 100KHz,
> the upper freq limit recommended by Fair-rite
> http://www.fair-rite.com/77-material-data-sheet/
> It's probably around 400mW/cc or even more at 136Kz. Your cores have a
> mag volume of 22.8cc according to the data sheet
> http://www.fair-rite.com/product/toroids-5978003801/
> so 68.4cc for your stack of three. That makes 27.4W dissipation for
> brick on the key WSPR operation. I'm not really surprised the cores get
> hot. Fair-rite want you to use type 78 material, which has 1/3 the loss
> at these frequencies. This was the experience of the guy who's combiner
> design you copied
> http://www.w1vd.com/137-500-KWTX.html
> where he says
> "At the kilowatt level issues developed with the output transformer. The
> FT-240-77 cores that worked well in the 500-watt deck began to show
> signs of 'stress' while testing the prototype 137 kHz kilowatt deck. The
> core ran noticeably warm to the touch, transition ringing on the drain
> waveform was difficult to tame and efficiency was less than expected.
> Additional core losses at the increased power level and excessive
> leakage reactance were to blame. After a study of the ferrite
> literature, the best solution appeared to be a switch to 78 ferrite
> material which was designed specifically for the 100 - 200 kHz frequency
> range. As luck would have it, 78 cores were available in FT-240 size and
> readily available through several distributors! A few days later
> FT-240-78 cores were being tested in the prototype 137 kHz deck. The
> cores just warm to the touch, ringing was much easier to tame and
> efficiency was back in the mid 90% range".
> To answer your specific question, a scope on the drain will probably
> show flat topping if the cores goes into saturation as its inductance
> collapses. But in your case it seems unlikely you are saturating the
> core, it's just that the AC loss is too high.
> Mouser and RS in the UK stock the type-78 part: Fair Rite 5978003801. RS
> is cheaper.
> Hope that helps,
> Alan G3XAQ
--
Best regards,
Chris mailto:chris@chriswilson.tv
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|