Romero's experiments and OU principles

[totoalas]

Hi Romero,

Of the different cores you have used, which give you the best results?

Thanks!

Hi Romero  additional question

Besides the cores
Winding in Neogen  bifilar   or just bifilar  by Konehead in RV   mode
Which is more  efficient  with N S configuration??
thanks again
totoalas

[bolt]

Cores depend on the frequency you wish to operate. Ferrite is good from hundreds of Khz to several Mhz so it not much use at lower frequencies as the magnetic retention on domain flips is too short.

Iron Dust is best for lots of  Khz  thus is use in switch mode PSU's typically running from 25khz to 90Khz after that the iron dust actually can not respond fast enough for domain flipping. A small 1" core can handle as much as 1000 watts without saturating.

Standard transformer especially large ones are excellent for low frequency work. A 1000KVA utility transformer is >98% efficient.  You will see excellent results working with large transformers over 500VA and will work very nice from 20Hz to 450 Hz is about the upper range.  Use variable pulse width, frequency and amplitude drivers to make KVARS's

You will really struggle working with tiny little cores solid state systems when the switching events have to take place at very HF creates all kinds of problems and tuning becomes a real bitch.

[teslaalset]

Iron Dust is best for lots of  Khz  thus is use in switch mode PSU's typically running from 25khz to 90Khz after that the iron dust actually can not respond fast enough for domain flipping. A small 1" core can handle as much as 1000 watts without saturating.

For some real data:
http://www.micrometals.com/materials_index.html
Iron powder suitable from several KHz up to 1 Ghz.

You're probably mixing up with permeability differences.

[romerouk]

Hi Romero  additional question

Besides the cores
Winding in Neogen  bifilar   or just bifilar  by Konehead in RV   mode
Which is more  efficient  with N S configuration??
thanks again
totoalas

Winding in Tesla-Neogen bifilar is really the best for generator and for the driver coils but not easy to make the coils like that. I have made 2 coils just to test  them but I spent almost 4 hours to do them and wasted a lot of selfbonding wire.
I am in the process of making many coils directly wound on the irondust core. I will post pictures with them next week. In NN config I find iron dust better than ferrite but in NS config the difference is not that high but still irondust is better.
Today I have just finished milling a rotor made from nylon66 and I am preparing another Acrylic rotor for a different config.
No more work today!

[bolt]

For some real data:
http://www.micrometals.com/materials_index.html
Iron powder suitable from several KHz up to 1 Ghz.

You're probably mixing up with permeability differences.

No im generalising without going into exotic iron mix like carbonyl iron with ceramics can run to a Ghz. But in practical terms anything over a couple of megs  is already heading towards air cores. As Ismael uses Tesla bifilar Air-Cores for his MEG.

Wiki

"Iron Powdered cores made of hydrogen reduced iron have higher permeability but lower Q. They are used mostly for electromagnetic interference filters and low-frequency chokes, mainly in switched-mode power supplies.
[edit]Ferrite
Main article: Ferrite (magnet)
Ferrite ceramics are used for high-frequency applications. The ferrite materials can be engineered with a wide range of parameters. As ceramics, they are essentially insulators, which prevents eddy currents, although losses such as hysteresis losses can still occur.

So unless you have an incredibly fast rotor its probably going to work better  with iron powder while the old AM LW MW radios used ferrite rods, slugs and cores for these frequencies and certainly not iron powder.

Mu-metal is pretty powerful stuff and as romero said VERY expensive. Don Smith used a lot of Mu-metal as well as another type which is around 10 times the cost of Mu-metal i cant think of the name right now,  Mu-metal has  permeabilities of 80,000â€"100,000. There is a trade off for materials there is no point making a 15 watt OU device that will take 200 years to pay for itself:)