To: rsgb_lf_group@blacksheep.org References: <000c01c55a6e$8d82f740$37d0fc3e@l8p8y6> Date: Wed, 18 May 2005 02:05:24 +0100 MIME-Version: 1.0 X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2615.200 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200 X-SPF-Result: relay.thorcom.net: 217.12.12.204 is neither permitted nor denied by domain of btopenworld.com Subject: LF: Re: CORES Content-Type: text/html; charset=windows-1252 X-SA-Exim-Scanned: Yes Sender: owner-rsgb_lf_group@blacksheep.org Precedence: bulk Reply-To: rsgb_lf_group@blacksheep.org X-Listname: rsgb_lf_group X-SA-Exim-Rcpt-To: rs_out_1@blacksheep.org X-SA-Exim-Scanned: No; SAEximRunCond expanded to false X-Spam-Filtered: by PlusNet SpamCORE (v4.00s) Content-transfer-encoding: 8bit

Dear Mal, LF group,

----- Original Message -----

From: hamilton mal

To: rsgb_lf_group@blacksheep.org

Sent: Tuesday, May 17, 2005 12:24 AM

Subject: LF: CORES

I have been winding some transformers recently and have been getting some odd results.

All cores have a prim of 9 turns and 27 turns sec.

1. 3C85 58 mm diam Prim inductance 167uh Sec 744 uh

2. Same material abt same size but E core Prim 292 uh Sec 2810 uh no gap.

3. Material unknown out of a smps similar size Prim 21 uh Sec 178 uh no gap

4. TV lopt core again abt same size Prim 79 uh Sec 865 uh no gap

The figures for the primary inductance seem quite reasonable, except no. 3 has rather low AL value for a ferrite transformer core, if that is what it is - perhaps it is iron dust? But with 3:1 turns ratio, the secondary inductance should be close to 9 times the primary - definitely not so in the case of 1, so something odd there. As to whether that is a good number of turns, it depends what DC voltage and power output level you are looking to operate at.

The Prim wire is litz 2mm and sec 1mm copper enam wire in all cases.

1 and 2 seem to have a big difference considering the same material, 3 must have a low u.

Has anyone else tried comparisons of cores.

If cores are selected at random and not checked there would be a big variation in results even using the same turns ratio.

Selecting components at random will usually give random results ... but provided the transformer has enough inductance, the right turns ratio, and does not saturate or overheat, a very wide range of constructions will actually work. These cores are probably capable of handling between several hundred watts to a couple of kilowatts with optimum winding design, but even with non-optimum design they will still work OK but be less efficient and more likely to overheat.

The above are intended as output matching transformers in the FET amps like the YXM and MRF class D/E design.

I normally use 1 in my amps but have tried 3 and in spite of the big difference in values it does seem to work and I have not noticed any difference in RF output, which seems strange.

73 de Mal/G3KEV

If you could take a selection of ideal, loss-free, inductors, capacitors, switching devices and transformers, connect them together in a circuit, then apply DC and 137kHz drive at one end and take output at the other end via a 137kHz low-pass or band-pass filter, you would have a 100% efficient switching mode PA stage. It would not matter what the component values were, or even how they were connected together, since the only way energy could leave the system would be as a 137kHz sine wave. Of course the real world is not quite like this; many possible combinations of components would result in impractically large or small voltages and currents, and real components do have losses, get hot, go bang etc. But even so, provided the circuit is halfway sensible and the components have reasonably low loss at the voltage and current levels they operate at, the majority of the DC input power will end up as RF power in the load, simply because there is nowhere else for it to go. This makes switching PA stages very tolerant of variations in some components. It is quite easy to end up with a PA circuit that "works" in so far as it generates substantial RF output, but on connecting an oscilloscope you find the most peculiar waveforms, with more harmonics than fundamental and wild voltage and current spikes. But because the output usually has a quite good low-pass filter, a nice sine wave is delivered to the load. However, a nasty waveform usually means at least some of the components are under a lot of stress, so it is always worth checking the PA has reasonable-looking waveforms, even if it otherwise seems to be working.

Cheers, Jim Moritz

73 de M0BMU