Graham Gunderson's Energy conference presentation Most impressive and mysterious

[Reiyuki]

  Anyway, since we're starting to evolve into proper analysis of this, I might as well post the rest of my photos/notes from the presentation.  Enjoy :
https://drive.google.com/drive/folders/0B0_1dJJ_ezEmSzZMRmxNSFkwZmc

Well, based on what we see in that PDF things are very murky and unclear.  Why don't you ask him for a complete schematic and bill of materials for the device, and also ask him for a complete test setup description including a test setup schematic with clearly indicated test points and test waveforms and test data?

  I answered that in the previous post.  Job and family come first, and I realize the hundreds of hours that went into building and tuning the actual devices, plus the hours on top of that I spent gathering firsthand information to share with the rest of you.  I wouldn't blame him for taking a couple months off after all of this.
  Besides that, I'd like to consider him a mentor and guide rather than a babysitter.  Surely, with our collective intellect we can design an equivalent timing circuit and synchronous rectifier without his help?

Don't get thrown off by the "unavailable cores" business or the business about "multiple permeabilities" in a magnetic circuit.  That is not "secret sauce" and a suitable substitute magnetic core implementation is easily doable.

  Agreed, it's probably not the 'secret sauce', at least by itself, because it's been thoroughly tried before.   But his setup as a whole has a unique configuration that I have not seen tested anywhere else around here, and since there is much detail already present, I thought it an interesting direction to pursue.

The virtual self-flagellation when it comes to these claims where you agree to suffer without proper information needs to come to a positive resolution.  I don't think the majority of people interested in this stuff are mental masochists.

  Run through the new material above and tell me what you think we might need next.  I think we have enough information to start exploring in this direction.  And since it's relatively unexplored, there should be much to learn

-rei

[gotoluc]

It would be interesting to know the actual value of the Purple trace. Its baseline is offset below the display area. Does the Measurement of 4.07 Vrms represent the actual value or just what is shown on the screen, from the bottom graticule marker to the trace itself? The question mark indicates unreliable measurement, doesn't it?

Nice work, as usual, anyhow.

Thanks TK


ch 3 (Purple trace) should be okay as its vertical position is set to -4.00. Its side arrow shows off screen (down arrow) after you pass -3.96


Luc

[DreamThinkBuild]

Hi Reiyuki,

Thank you for all the information, based on your pictures I made a interpretive drawing of the switching. The multi-material may be a way to force it to take the path of least resistance so to speak.

The only thing that is really throwing me off is the top bias magnets are they also N<->N (S<->S) pointing outwards or N<->S?

The magnets on the outside where they are parallel to the main coil are also of interest, in the top view picture. The joined south poles could form a split in the center of the main coil making north dominate until the drive coil forces it out of balance.

Again these are only my interpretations with the current data which could be wrong.

[Spokane1]

Dear DreamWorker (or something like that)

All the nice work you did was probably based upon the diagram that Graham sketched for Reiyuki at the conference.

That arrangement didn't make sense to me either, so I asked Graham about it. He said the he was attempting to show Reiyuki some other construction feature about the bias magnets - I couldn't follow what that was, however Graham did assure me that all of the PM's were aligned in the same direction to form a continuous magnetic loop.

The little white dots on the end of the PM's are casting notes to establish which end of the magnet was at the top of the mold. Apparently there is a significant variation of intensity from one end to the other. Graham was using Barium Ferrite magnets that he had in his personal inventory. The dots don't indicate polarity.

Speaking of the bias magnets; if you look at the photos that Reiyuki has contributed you will notice a spacer between the PM's and the surface of the central transformer core. This is a 5/16" or 3/8" machined piece of acrylic. Graham describes how he spent a fair amount of time "tuning the bias field" by using different thickness (and materials) for spacers. Changing this spacer also requires a slight change to the operating frequency. Graham has a lot of experience using PM's to modify the core magnetic dynamics. He was doing this kind of technique 10 years ago. On his work bench he had about 4 different thickness of spacers. The thinner ones (down to 1/6") were of a white material not unlike polyethylene.

Speaking of core magnetic dynamics. Graham says that when the machine is running the magnetics in the lower "C" core is not the same as that which is flowing in the upper "C" core. He said this is not normal in a closed magnetic core even with different permittivity's.

The lower "C" core is a common 100mm (4") HC90 with 1" square legs (or arms). The core is composed of two "C" sections with 0.005" of Mylar sheeting (one on each leg) for a total gap of 0.010".

Maybe you can salvage your simulation and add the impact of the spacers.

Also, those 18 bias magnets on the top of the apparatus are also responsible for a significant amount of performance improvement. I certainly don't see how being that far away. Graham said that he could hold a few magnets in his hand at a distance of about 24" from the operating device, rotate them and watch his output lamp vary in intensity. That is some sensitive device. This only shows how narrow the window of operation might be.

Spokane1

[MileHigh]

The moment you see actual permanent magnets incorporated into some kind of magnetic circuit that is effectively functioning as a transformer then warning bells should be sounding.  Except for a very few specialized applications, adding permanent bar magnets to magnetic circuits to "bias" them is usually nothing more than electronics quackery.  The simple test is to run the setup with the permanent magnets in place, and then run the setup without the permanent magnets in place.  The reasonable expectation is that the circuit will perform better and have more power headroom without the permanent magnets.

I anybody builds this (Do we know what the configuration actually is?) then I strongly recommend testing it with and without the permanent magnets.

Also note that depending on the magnet, you may be demagnetizing your permanent magnet when you place it in some sort of AC transformer setup.