Thane Heins BI-TOROID TRANSFORMER

[lumen]

Lumen I think it's mainly because of the 3d aspect. A 2d model is already a big job on the cpu, so you can imagine the strain of a 3d model. What I find too bad is the lack of software that simulates an actual running transformer design where voltage induction is part of the simulation. That would save a lot of time.

Broli
I setup to center coil at 120v with 200 turns and .8 ohms resistance, and run the frames @ .001 second steps up to .019 seconds using a 60hz coil input . The animated gifs that I uploaded show the core "B" flux for one entire cycle.
It seems to work very well, providing there are no operator errors.
The entire calculation only takes about two minutes with a mesh size of .01"

[CRANKYpants]

GOOD LUCK! 
CHEERS
T

BTW - THIS IS WHAT WE DO HERE ON THE WEEKEND! http://www.youtube.com/watch?v=tfC5oYcknew

BITT COMPUTER MODELLING TEST ITEMS:

1. Make the Primary and (net) Secondary with the same wire gauge and the same number of turns so it will be a 1:1 transformer,
i.e. Primary = 100 turns, S1 = 50 turns, S2 = 50 turns.
2. TEST: S1 + S2 (net)NO LOAD voltage = primary INPUT voltage.
3. Primary flux should be evenly distributed through NO LOAD S1 and S2.
4. Note and record Primary Current and Power Factor NO LOAD baseline.
5. Place S1 and S2 ON 100 ohm LOAD and note if Primary Current or Power Factor changes.
6. SOFTWARE TEST:
Remove S2 from load and note S1 load voltage which must = 0 volts because S2’s flux path route now represents a lower reluctance route than S1’s ON LOAD high impedance route.
7. S2’s NO LOAD voltage must = Primary INPUT voltage because S2 is getting all the Primary flux.
8. REPEAT with 50 ohm, 25 ohm, 10 ohm, 1 ohm and shorted Secondaries.
9. If Baseline NO LOAD Current or PF change when placed ON LOAD increase Secondary Outer Core Area to reduce reluctance until there is NO CHANGE from NO LOAD Baseline to ON LOAD shorted Seconaries.
10. Replace Secondary Outer Core with HIGH PERFORMANCE Low Reluctance Permalloy of Superpermalloy etc. And note performance advantage (if any).

TEST NOTES: Test Date: __________________

1. Number of Turns: Primary = _________ turns, S1 = __________, S2 =__________ turns.
2. Primary Input Voltage: Primary = ______ Volts, S1 = ______, S2 = ______ Volts.
3. S1 NO LOAD Voltage = _____ V, S1 NO LOAD Voltage = _____ V,
Flux Distribution: ________________ (even - uneven).
4. NO LOAD Baseline: Primary Current = ______ Amps, Primary Power Factor = ______.
5. S1 & S2 ON LOAD (100 ohms):
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
6. S2 NO LOAD: S1 ON LOAD Voltage = ______ Volts. (Must = 0 Volts).
7. S2 NO LOAD Voltage: S2 = ______ Volts. (Should = Primary Input Voltage).
8. S1, S2 50 ohm load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
S1, S2 25 ohm load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
S1, S2 10 ohm load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
S1, S2 1 ohm load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
S1, S2 (shorted) load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%.
9. New Secondary Core Area Increase: _______ %.
S1, S2 (shorted) load:
Primary Current = ______ Amps, Current change % = ______%, Power Factor = ______,
PF change % = ______%. (NOTE; if primary current and PF do NOT change with a short-circuit load they won’t change with lesser loads either).
10. New Secondary Core Material: ___________. Primary Current = ______ Amps,
Current change % = ______%, Power Factor = ______, PF change % = ______%.

[ramset]

Thane
AHHH fun with the Kiddypooos!!
So Thats not to different  from my trip to the "Bronx" last week?

Thats Poindexter [MY lab assistant} he loves it when I forget to put my helmet back on!!
{little Rascal]

Test Data link

http://www.overunity.com/index.php?action=downloads

[lumen]

Broli,
I think you may be right about the calculation times. After checking and finding the mesh settings of .01" were not operational, I changed them to a more realistic setting of .2". Even at the new setting of .2" the calculations take about 20 minutes. The good thing is now the results are more of what one might expect to see.
As MR "T's" test procedures go, I only wish I was good enough to setup the program to perform those tests. This is some powerful program, but you need to write formulas to return the results in volts and current based off things like core energy or the core "B" field intensity.
So I think at this point I am still a bit away form any software solution until I can put all the pieces together.

The "B" field vector plots at least now, look real and small changes to the core have small effects also, so I think I'm on the right track anyhow.

[teslaalset]

Guys, just a quick update where I am with my FEMM simulations.

In my earlier simulations I applied "pure iron" from the default materials library.
However, that is solid iron. So, I made a change to use laminated "pure iron".

I found a nice reference model that can serve as example how to do the detailed simulations:
http://www.femm.info/wiki/MyTransformer .
This example uses a combination of FEMM and Octave scripting, which is new for me, but it looks usable to get all currents and voltages out of simulations. It's a lot of work though, so it will take some time.

Looking at this model I found that the BITT has a different basic model.
Mainly the M (couple factor, or mutual inductance) is the difficulty here.
In my view M is depending on the saturation state of Core1.
So, in the BITT, within a full AC sinus cycle, M will vary.
I will first think of a good first attempt and show it here later.
Maybe you have some ideas you could share here.

Below the simple model details of an ordinary transformer in the given example: