Kapanadze Cousin - DALLY FREE ENERGY

[itsu]

Remember that solid ferromagnetic Iron resonates at 45.5MHz due to the huge internal 33T magnetic field between its domains, so any sub-harmonic of that (i.e. 22.75MHz, 15.2MHz, 11.375MHz, 9.1MHz, etc...) should be paid special attention to.
Of course the Yoke core probably is not composed of iron powder but of iron oxides with nickel, zinc or manganese, so these frequencies can be shifted.
The TL494 is not capable of oscillating in the MHz range.
...but MOSFETs could ring there as well as LC parasitics.

To find out whether the Yoke's core is responsible for some of these peaks, wave a ceramic magnet in the vicinity of the yoke and observe if any peaks shift their frequency. 

Warning: Unlike ceramic, NdFeB magnets are conductive so they can affect resonances by eddy currents developing inside them, which is undesirable.

Ok, i have some ceramic magnets i can wave over the yoke.   
Perhaps i could kind of zoom in with the probe of the spectrum analyzer to see where it comes from (yoke or MOSFETs).

I am planning on using a T520 core instead of a yoke with your loss-less clamp design to see if this will be any more stable (or is this NOT what we want    )

Thanks,  Itsu

[itsu]

Hi Itsu. That's interesting about the 14.3 MHz when out of resonance, and seeing much more high frequency content
when you set the yoke secondary to its resonance frequency. As Verpies suggested, that 14.3 MHz peak may be due to
harmonics generated form the TL494 and the driver FET's switching, and the other high frequency content that appears at
resonance may also be generated by transistor switching at higher current draw, but on the other hand, driving that yoke core
hard from the PWM at the resonance frequency and also the high current flowing in the three turn secondary winding (or whatever
turn count you are using) at resonance must put some major stresses in the ferrite. Whether that is any factor in the high frequency
content you are seeing at resonance or not is hard to say, since switching circuits typically generate a lot of harmonic content, but it
seems worthy of paying at least some attention to. Verpies idea of using the ceramic magnet might prove interesting as well.

Hi Void,

yes that solid 14.3Mhz signal when not in resonance was surprising to me to, i will try to zoom in on that situation trying to see where it comes from.

Itsu

[verpies]

I am planning on using a T520 core instead of a yoke with your loss-less clamp design to see if this will be any more stable (or is this NOT what we want    )

On the EEVblog.com, we strive to stabilize and eliminate any weirdness.
However on this forum, we nurture weirdness, especially if it is unexplained 

[verpies]

Yes that solid 14.3Mhz signal when not in resonance was surprising to me to, i will try to zoom in on that situation trying to see where it comes from.

If that frequency shifts in presence of a magnet and this shift comes from ferrite's internals, then the shift will be pretty small because the 0.3T field from the ceramic magnet cannot compete well with the 33T internal field between the core's domains. 

0.3T/33T = 0.009 which would mean only ~1% change in frequency.
So a high zoom would be necessary to notice it.


P.S.
That 33T internal field is documented in pure iron.  In calcinated iron oxides, the inter-domain field is undocumented and can be different.

[itsu]

I have 16 of them, but i understand that stacking more then 4 will not further increase their force.

I will give it a try.

Itsu