This is a multi-part message in MIME format.

------=_NextPart_000_0139_01C2A46B.595083E0
Content-Type: text/plain;
        charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

> > If you want to buy a core aim for a core rated to work at between 20 and
> > 100kHz (the usual frequency range for what you are trying to build -
this
> > is what the material determines) and meant for dc/dc converter use

> But I've used a 1uF capacitor in RM's circuit - which blinks at about 1-5
> Hz.  This is my frequency isn't it?

If you are referring to my* 2 transistor LED flasher using a PNP and an NPN
circuit transistor (diagram attached) then it is a special case. In MOST
simple flasher circuits the inductor plays a major part in the determination
of the repetition frequency. However, in this circuit the time constant of
the RC network (t=RC) is the main factor in the repetition period. BUT the
inductor chaging takes place over a much smaller perriod controlled by how
long Q2 is held on while C discharges through its base-emitter junction.
This is typically much smaller than the Rc time constant. The charge period
"T" is related to the frequency that Peter is talking about by approximately
f=1/(2T).

In this circuit the main aims are

1.     that the LED flashes brightly enough but
2.     doesn't saturate and waste power unproductively in resistive
dissipation.

Because of the poorly defined drive characteristics when Q1 is turned on (a
capacitor being discharged directly into a transistor's base) it is hard to
perform a useful theoretical analysis of the on time.

One way to check for saturation in practice is to place a SMALL resistor
between the inductor and ground and observe the waveform across it. The
voltage will ramp upwards and then drop backwards with an approximately
triangular waveform. If the top end of the triangle grows a large vertical
spike which is a significant proportion of the height of the triangle then
the core is saturating. With this simple circuit the solution to saturation
is
to use a smaller capacitor or a larger inductance inductor or an inductor of
the same inductance with a higher current rating.

An easier practical solution (or, at least, one which doesn't require an
oscilloscope) is to try different inductors and choose one which gives
lowest average current draw for a given LED brightness.

Another method which gives you SOME idea of peak inductor current, provided
you know it's inductance is -

Measure mean current drain of inverter I. For a very occasional flash use a
LAAARGE capacitor and measure current drain uphill from it.
Measure voltage of supply V.
Power in is about I x V

Power delivered to LED is 0.5 x L x Ilpeak^2 x F
F = flashes per second (may be < 1)
So VI = 0.5LIpk^2F
Ipk = sqrt(2VI/FL)

If Ipeak given by this formula exceeds or is near the current rating of the
inductor you may have saturation occurring.

Using an oscilloscope is the easy answer !



        Russell McMahon

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.



------=_NextPart_000_0139_01C2A46B.595083E0
Content-Type: image/gif;
        name="led1cell2.gif"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
        filename="led1cell2.gif"
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------=_NextPart_000_0139_01C2A46B.595083E0--