Many thanks to Andy, Warren and Mal for their helpful info. It
certainly looks like I have a shorted turn.
I will start by trying to dismantle the existing pot core, but I am
not optimistic.
Warren's inductance measurements are very useful as I presume I just
need to replicate those figures with my new core.
I have a couple of 3C85 toroids which could usefully be pressed into
action.
But if anyone has a spare transformer (or even a Decca chassis minus
the amps) it would save me some work.
And I am still baffled by the missing 3.3n cap shown on the circuit
but nowhere to be seen in practice.
Thanks again.
Mike
====
> Any large modern SMPSU core will cope with 1kW of sinusoidal 137kHz
>
> Just make sure B is kept below ~0.1 Tesla using the equation to
> determine teh minumum number of turns
> V(rms) = 4.44.F.N.A.B
>
> F Hz, N turns A core x-sectional area in m^2, B Tesla.
> (the factor 4.44 comes from SQRT(2) * PI )
>
> In my 137kHz 700 Watt Tx, I used an ETD49 core with cross sectional
> area of 200mm^2 made from 3C85 material designed for Bmax = 0.05T.
> That runs cold. Thes days, even better ferrites are availble, like
> 3C90. These are less lossy still, but Bmax is no more. TV LOPTs use
> generally older types of ferrite, even older than 3C8, so its worth
> going for a modern SMPSU core rather than cutting corners dismantling
> old TVs. Although as Mal has clearly found, using LOPTs beyond their
> design frequency, provided B is kept low by using enough turns, even
> they will run OK. Just don't use too many turns that copper losses
> begin to rear their head.
>
> Andy
> www.g4jnt.com
>
>
> On 4 April 2011 19:29, mal hamilton <[email protected]> wrote: >
> Output transformer info. > Faulty output transformer. > Strip old
> transformer windings and rewind as follows. > Prim 3 x 6 turns Sec 31
> turns > or > use large LOPT 3C material with same amount of turns as
> above. > > Wire guage guide Prim 3mm litz per winding and Sec 1 or 2
> mm enamel > > > de G3KEV >
>
|
