From: "james moritz" <j.r.moritz@herts.ac.uk>
To: rsgb_lf_group@blacksheep.org
Date: Mon, 4 Apr 2005 19:33:57 +0100
Message-ID: <000001c53944$dd1f69f0$e6a4c593@RD40002>
MIME-Version: 1.0
Importance: Normal
In-Reply-To: <326392110@web.de>
Subject: LF: RE: RE:  TX Cores, Gap from DK 8 ND
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Dear LF Group,

Hans-Albrecht makes an interesting point here - certainly, texts on
switch-mode power supplies often refer to problems with push-pull converters
caused by an imbalance of flux in the core, either due to the core itself,
or some asymmetry of the circuit, so it is surprising it does not seem to be
an issue with the push-pull class D amplifiers. 

I think that there is a self-balancing action that results from feeding the
DC input through a high-impedance choke - if the core is closer to
saturation with the current flowing in one direction in the primary, it will
effectively have somewhat lower impedance than when the current is flowing
in the other direction. Since the current is nearly fixed over the duration
of one cycle, there will be less voltage drop across the winding during the
half cycle where the core is closer to saturation and the winding has lower
impedance, and so more flux will be generated in the other half cycle,
tending to pull the average flux in the core towards zero and restore
balance.

As an experiment, I applied a DC bias current through a 3mH air cored choke
(or loading coil, as some people call it...) into the 50 ohm secondary
winding of a class D P-P current switching output stage, and monitored the
voltage waveform at the primary centre tap. With zero DC bias, the waveform
is approximately a "full-wave rectified sine", but as the DC bias increases,
the voltage peaks become asymmetrical, with the peaks of the higher-flux
half of the cycle slightly smaller than the lower-flux ones. Once the DC
bias is removed, the original waveform is restored.

This apparent balancing would seem to be an argument in favour of
current-switching rather than voltage-switching in the transformer-coupled
push-pull type of circuit. 

Cheers, Jim Moritz
73 de M0BMU

-----Original Message-----
From: owner-rsgb_lf_group@blacksheep.org
[mailto:owner-rsgb_lf_group@blacksheep.org] On Behalf Of Hans-Albrecht Haffa
Sent: 01 April 2005 17:30
To: rsgb_lf_group@blacksheep.org
Subject: LF: RE: TX Cores, Gap from DK 8 ND

Dear Lowfers, 
any ferrite core should have a gap. This is necessary to reduce the remanent
flux from about 180 mT to as low as 30 mT. It may be just the residuous gap
of lapped cores, preferrable is to put a layer of capton tape on one outer
leg of an EE-Core (EDT59 and the like). Lowering remanent flux will rise
flux swing at operation. Apparently it is not a problem at 136kHz, as core
saturation is unlikely at this frequency.  Saturation will occure at
frequencies below 50 kHz during startup of push-pull switching circuits
(symmetrical half bridge or full bridge), designed by beginners. A gap as
low as 0.05 mm (one layer self adhesive film tape or a bit of varnish) will
hardly  lower permeability, so from urel=2700 down to urel=1500. If You can,
use this effect. A toroid core is often second choice in power electronics.
Two paralleled ETD cores have greater surface than two stacked toroids. With
the ETDs copper volume is doubled too. not so with the stacked toroids. So
if You can buy ETD cores , do it. They are available at Conrad.com or
conrad.de.

55, Hans-Albrecht
DK 8 ND


-----Ursprüngliche Nachricht (HTML) wurde als Anlage beigefügt.-----
Von:      rsgb_lf_group@blacksheep.org
Gesendet: 01.04.05 11:11:24
An:       "Hans-Albrecht Haffa" <dk8nd@web.de>
Betreff:  Re: LF: RE: TX Cores


__________________________________________________________
Mit WEB.DE FreePhone mit hoechster Qualitaet ab 0 Ct./Min.
weltweit telefonieren! http://freephone.web.de/?mc=021201