Magnetic Implosion Transformer Replication

[David G Dawson]

Coils & Transformers

Coils & Transformers:
Mark,
Thanks.
Are you intending a build as your excellent input here makes me feel you will be?
A Flyback 'C' or 'U' Core Ferrite would make a good option for the one displayed.
I heard Graham saying 'with a gap' and immediately saw a Flyback.
Looking at this with Vacuum Tubes in mind as I am distinctly not a fan of solid state.
A good description of why the paper thin gap here on Page 7:

Coils and transformers

Smokey

[David G Dawson]

Barium Magnets

Barium Magnets:
Mark, again you might be the person with an answer to what I seek in Barium/Ferrite Magnets.
I have made a stop production date of using Barium somewhere towards the end of the1960s.
Do you know of any method by which a Magnet containing Barium can be identified from a later production type?
Brian Prater designed a 'Magnet Tickler' which I built but saw that as a detector of Floyd Sweet's magnet 'conditioning' for his VTA but that is the only tool I might have to supply an answer but no idea of what to look for.
Both the VTA and MRA appear to have caused the cessation of Barium use to the general population - my observation.
Corporations appear to be able to order whatever they want but NOT you and I.
Thanks.

Smokey

[Spokane1]

Availibility of Barium Ferrite Magnets

Barium Magnets:
Mark, again you might be the person with an answer to what I seek in Barium/Ferrite Magnets.
I have made a stop production date of using Barium somewhere towards the end of the1960s.
Do you know of any method by which a Magnet containing Barium can be identified from a later production type?
Brian Prater designed a 'Magnet Tickler' which I built but saw that as a detector of Floyd Sweet's magnet 'conditioning' for his VTA but that is the only tool I might have to supply an answer but no idea of what to look for.
Both the VTA and MRA appear to have caused the cessation of Barium use to the general population - my observation.
Corporations appear to be able to order whatever they want but NOT you and I.
Thanks.

Smokey

Dear Smokey,

Barium Magnets were banned by OSHA because they claimed that some workers were suffering ill health effects from the dust that is created during their manufacture. Was there another reason? Who knows but I give it a 50/50chance that they were withheld because of some other black box technology that we don't even know about and they wanted to insure that nobody could stumble across it. I really doubt that the cloistered inventions you mentioned would present enough of a threat in themselves to cause this to happen. I think that something much bigger was being protected.

Companies who want Barium Magnets can manufacture them in house with a variance permit from OSHA. It takes about $50K- 100K in equipment. The raw materials are available and the manufacture technology is not that difficult. It involves grinding, sifting, baking, and machining. Most companies wouldn't bother. There are custom fab plants in China that, for a fee, will make batches of these kinds of magnets. I hear that some of them are of very high quality, but again you are talking about a $50K up front order.

As far a most industrial applications go Neo's have supplanted ferrite magnets so there is really not much demand for them except in low cost markets.

There is a place in Seattle that is set up to do arc spectral analysis of ferrite powders. I believe it is about $250 and they will do 10 samples at a time. Other than that there is no way to tell a Barium Ferrite from a Strontium Ferrite.

Spokane1
Mark McKay

[Spokane1]

Fly-Back Cores

Coils & Transformers:
Mark,
Thanks.
Are you intending a build as your excellent input here makes me feel you will be?
A Flyback 'C' or 'U' Core Ferrite would make a good option for the one displayed.
I heard Graham saying 'with a gap' and immediately saw a Flyback.
Looking at this with Vacuum Tubes in mind as I am distinctly not a fan of solid state.
A good description of why the paper thin gap here on Page 7:

Coils and transformers

Smokey

Dear Smokey,

Common CRT Fly-back transformers (which are no longer common) are to small for this apparatus for now. We need a core that is 4" wide and has 1" square arms. Until we are sure we can downsize the central core it is best to reproduce Grahams equipment as close as we can. Maybe we might be able to change the frequency in the future and use a smaller core. By then there will by no fly-back transformers available.

Dr. Tesla was able to advance this technology by using the properties of sparks as his switching system. It is only though the recent development of SiC Power MOSFETS that the switching speeds have come close to the operation of a spark switch at 10 ns rise times.

I'm sure that vacuum tubes can be used to do this kind of work as well, but you have to admit that one would have to have one heck of a COP to overcome the filament heat losses in the switching elements, not to mention the need for isolated filament transformers.

The 0.010" gap helps reduce the coupling k factor between the primary and secondary. Apparently each magnetic circuit need some degree of freedom for this process to work and a high coupling factor does not help. Apparently this optimum gap was determined experimentally.

Spokane1

[aniccame]

Diagram

Thanks for this circuit Mark. I do have a question.

It's my understanding that the SiC MOSFETs require a 20v/-5 volt gate drive. How is that accomplished in the schematic given?

Also, I may have missed it, but what function does the toroid located at the bottom serve?

Thanks,

Alan

[Spokane1]

Synchroinous Diode Details

Thanks for this circuit Mark. I do have a question.

It's my understanding that the SiC MOSFETs require a 20v/-5 volt gate drive. How is that accomplished in the schematic given?

Also, I may have missed it, but what function does the toroid located at the bottom serve?

Thanks,

Alan

Dear Alan,

Graham actually has a dual channel isolated DC to DC switch mode power supply on the Synchronous diode circuit board, that is the purpose of the ferrite core in the center of the board. There are six windings on that ring. I believe that two of them are primaries and the rest are secondary's. Graham told me that the six photo isolators are Zener controlled and provide regulation feedback to gate the master oscillator. I believe that the TS556 timer, way back on the logic board, is the actual Power supply master clock. His circuit generates three voltages +12, +5, and -5 Volts.

Keep in mind that there are a lot of components on the back side of that board that are not shown in the photos we have.

My next drafting project is going to be a block diagram of what I think is on that board. That should help explain what is going on.

Really there are several ways to do this. As far as I understand it the control of the synchronous diode is simple. Both FETS are closed for 99% of the time. then they are opened between 5 and 1000 nS at a precise time in the systems timing, one pulse per cycle of operation.

While the FETS are closed there are two current loops feeding the big storage capacitor. the opposing currents cancel each other out so the net charge on the capacitor is whatever has been rectified into it in that brief moment.

Spokane1

[aniccame]

PC

Very interesting. I will wait to see what your block diagram looks like.

I am tempted to use my basic IRF mosfets to do a more simple replication just to play with it.

I also don't know what cores would be ideal but I have some smaller ones to play around with.

Your work is very important to making it viable though because SiC MOSFETs and Litz wire are going to be necessary.

[Spokane1]

Very interesting. I will wait to see what your block diagram looks like.

I am tempted to use my basic IRF mosfets to do a more simple replication just to play with it.

I also don't know what cores would be ideal but I have some smaller ones to play around with.

Your work is very important to making it viable though because SiC MOSFETs and Litz wire are going to be necessary.

Dear Alan,

You are correct on the drive voltage for SiC devices. Graham specifically used +12 and +10 Volts probably for example rather than actual values. Everyone is going to have to adjust their power supplies to the devices they can afford.

Attached is my take of the Synchronous Diode Schematic. I'm sure improvements will be needed.

Spokane1

[RAMSET]

A contribution for consideration

An old acquaintance of Grahams has offered a video to view

Verpies
Quote
...and please send Graham the link to this video:
https://www.youtube.com/watch?v=av53ZMeWQ-k
------------

respectfully
Chet K

[Spokane1]

MOSFET Gate Leakage

An old acquaintance of Grahams has offered a video to view

Verpies
Quote
...and please send Graham the link to this video:
https://www.youtube.com/watch?v=av53ZMeWQ-k
------------

respectfully
Chet K

This 8 minute video is well worth the time. It is done by Mr. "Itsu". You can tell by the quality of his equipment that he is well funded for this sort of work. He is demonstrating the magnitude of MOSFET gate leakage to the Drain and Source with a number of power MOSFET's at a number of frequencies.

I believe that the purpose of this demonstration is to show that there is energy leakage between a gate driver system and what ever load is connected to the drain/source. This has a noticeable impact to the proper measurement of low COP OU systems, in that these losses need to be accounted for or measured in order to get an accurate performance evaluation of the device under test.

What wasn't shown here, but maybe next time, is the magnitude of the losses involved. I believe that LED's operate at a forward voltage of around 3.3 volts while they can consume anywhere between 10 and 100 mA. That could be a power loss between 33 and 330 milliwatts per switching element.

This is good technical information that researchers in this field should be aware of.

Spokane1

[RAMSET]

A leaky observation ? ....perhaps not !!

Forward from Smudge [Cyril Smith]
Quote
From GG's demonstration of a sawtooth flux waveform I have taken the primary current waveform then deduced the secondary current needed to create that sawtooth, assuming normal transformer action where the flux comes from the difference between pri and sec ampere turns. Done crudely from my sketched waveforms. And I get a secondary current very much like that shown by GG. So I deduce that there is no significant flux leakage and my original estimate of the primary flux waveshape is wrong.

Smudge
----------------------------------------------

respectfully
Chet K

[Spokane1]

Gentlemen,

Folks disregard this drawing - it is missing some components - I shall repost an updated version below.

In my simulations of the H-Bridge circuit I have found that two of the MOSFETS are not active therefore making them unneeded. Upon further examination I have found that the entire H-Bridge can be replaced with one switch and a diode. The input to the primary of the Gunderson Transformer is the same in that it produces the same voltage and current wave forms with far fewer components.

k4zep made additional simplifications by proposing that the switching elements be placed on the low side of the circuit to better reference driver connections. This circuit variation also simulates the same as a full four element H-Bridge.

Graham said 7/18/2016 that he was planning to redo his H-Bridge with a two switch design. He also said that providing schematics to his demonstration device would be pointless since he was planning to re-do all the circuits. Both statements seem to be correct.

Anyway, here is the proposed simplified circuit. The elimination of three switching elements should reduce the gate leakage issue by 50%.

Does anyone see any issues with this design and why it might not work as good as a full H-Bridge in this application?

Oops the polarity on the 220 VDC power supply is backwards.

Spokane1

[Spokane1]

Gentlemen,

Here is my take on the operation of the Synchronous Diode.

Comments Welcomed!

Spokane1

[Spokane1]

Dear All,

Here is the updated schematic, Lt Spice schematic and a timing chart for the primary circuit for your use and pleasure.

Spokane1

[Spokane1]

12 Volt Operation

Dear All,

This weekends work showed a little promise. I was able to get the junk box mockup circuit to lock into the proper timing sequence to display the discontinuous sine wave. I just have the primary circuit operational. This evening I shall add the synchronous diode circuit and start to get a look at the backend operation. For what its worth this circuit shows:

1. This circuit will function at 12 VDC (using the single switch approach)

2. Operation can take place as low as 3.2 kHz

3. A classical laminated iron core will support the novel oscillation pattern.

None of this means that we are in the OU window of operation, but it is nice to see some hardware implementation of the circuit under discussion. Using the component parameters I have in the actual circuit the simulation shows that I might be able to harvest up 6.5 Watts at a COP of .863. I'm sure the real world performance will be far less than this.

The advantage of the 12 V approach is that all the excitation energy comes from one source. All the gate losses, skew losses, and magnetic probe interference issues can be evaluated. If this circuit can perform as a self runner then this is the direction to go.

The simulation also shows that the circuit performance improves if I were to use a smaller charging inductance choke. The junk box device (from a salvaged X-Ray transformer) has a 1 KHz inductance of 11.5 mH. The simulation shows that I need a 2 mH unit there. Grahams circuit apparently used a .0324 mH custom device.

If any one is interested in a schematic for this temporary evaluation circuit let me know. I'm sure there will be several variations employed as the circuit technique improves.

Spokane1