Pierre's 170W in 1600W out Looped Very impressive Build continued & moderated

[partzman]

Hi seaad,


1) This is indeed a bit strange, if the supercaps are low in voltage, I would expect a large inrush current, but maybe the supercap boards have some current limiting.

Well, I see that the videos are once again removed so I can't refer to the first video when Pierre shows the charging of the supercaps from 4.64v to 10.4v over a ~32 second period.  It could be seen that the current drawn on the primary during this time was ~.5a rms if the meter used can be trusted.  From this one can compute the overall total impedance and resistance in the charging path including the 4 ohm resistor.  The primary to secondary leakage inductance, diode drops, and the pri and sec dcrs must be included in these calcs.  With best guesses on the transformer parameters, this should be capable of being simulated.

2) This would mean this 25.1V is the max voltage that the transformer can put out after rectification and the caps are charged up to that so current draw goes to zero. This would be the peak voltage of the secondary.

Yes this is correct!  This should give us the ability to arrive at the approximate open circuit turns ratio of the transformer assuming there will be some esr in the suppercaps.

3) I believe Pierre states in his first video that the transformer is home made. Mains is 170Vpp, secondary is 25Vpp, so winding ratio is 6.8, or 1.5A at primary becomes 10A at secondary which would cause 40V voltage drop over resistor... so indeed this doesn't make any sense... Only thing I can think of is that the primary current meter is not reading correctly.

The transformer is not home made but is more than likely modified.  The transformer is made by Signal and is model HPI-27-1072.  The label could be seen in the first video.  This series is available thru Digi-Key in the US and sells for ~$525 each!  With the core laminations being tig welded, the core was not disassembled so any mods would only be able to be done to the secondary. 

The data sheet for this series from Signal does not include the inductance of the primaries nor the magnetization current of the primaries under no load however, the idle current shown by the meter used by Pierre toggles between 0.00 and 0.10 amps which would/should be the magnetization current.  Having designed many such power transformers for commercial and industrial use, this magnetization current is unusually low IMO.  In my LtSpice sims of this transformer, a single primary at 114v rms with 1 henry inductance has an idle current of ~0.30 amps rms.  So, Pierre must be using the two primaries in series to achieve the low idle or magnetization current. 

At this point in time, I've not been successful in simulating Pierre's power transformer circuitry supplying the supercaps under all the conditions such as open circuit charging of the SC's, output no-load idle voltage of ~25vdc, and loaded (stator running with no rotor) output voltage of ~20.5vdc with an input of 113v rms at 2.0a rms!

Regards,
Pm

PmgR

[konehead]

Hi Ole and everyone
That all about circuits you put up about using mosfets to switch (or block specifically I think in that link)
There should also be a diode (very fast good amps good voltage) from the source to the drain....otherwise it only will switch "one mosfet at a time"
Look att this diagram for bidirectional mosfets

https://sites.google.com/site/alternativeworldenergy/shorting-coils-circuits

[pmgr]

Well, I see that the videos are once again removed so I can't refer to the first video when Pierre shows the charging of the supercaps from 4.64v to 10.4v over a ~32 second period.  It could be seen that the current drawn on the primary during this time was ~.5a rms if the meter used can be trusted.  From this one can compute the overall total impedance and resistance in the charging path including the 4 ohm resistor.  The primary to secondary leakage inductance, diode drops, and the pri and sec dcrs must be included in these calcs.  With best guesses on the transformer parameters, this should be capable of being simulated.

Yes this is correct!  This should give us the ability to arrive at the approximate open circuit turns ratio of the transformer assuming there will be some esr in the suppercaps.

The transformer is not home made but is more than likely modified.  The transformer is made by Signal and is model HPI-27-1072.  The label could be seen in the first video.  This series is available thru Digi-Key in the US and sells for ~$525 each!  With the core laminations being tig welded, the core was not disassembled so any mods would only be able to be done to the secondary. 

The data sheet for this series from Signal does not include the inductance of the primaries nor the magnetization current of the primaries under no load however, the idle current shown by the meter used by Pierre toggles between 0.00 and 0.10 amps which would/should be the magnetization current.  Having designed many such power transformers for commercial and industrial use, this magnetization current is unusually low IMO.  In my LtSpice sims of this transformer, a single primary at 114v rms with 1 henry inductance has an idle current of ~0.30 amps rms.  So, Pierre must be using the two primaries in series to achieve the low idle or magnetization current. 

At this point in time, I've not been successful in simulating Pierre's power transformer circuitry supplying the supercaps under all the conditions such as open circuit charging of the SC's, output no-load idle voltage of ~25vdc, and loaded (stator running with no rotor) output voltage of ~20.5vdc with an input of 113v rms at 2.0a rms!

Regards,
Pm

Partzman, I see only two suitable ways to use this transformer in his setup (without cutting into the windings):

1. Use two primaries in series and use two secondaries in parallel: primary voltage 230V, secondary voltage 115V, so 2:1 turn ratio. Then input 115V on primary, that would give 55V rms on secondary with 24A rating.


2. Use two primaries and one secondary in series: input voltage would be 345V, output voltage 115V, so 3:1 turn ratio. Then input 115V on primary, that would give 38V on secondary, which is close to what Pierre states as "36V home made transformer".

Let's consider case 2. above: Having 1.5A on primary would then translate to 4.5A on secondary and the voltage drop over the resistor would be  18V drop, or about 20V left for the capacitor.
This looks like it is in the ball-park.

Would this configuration make sense for the magnetization current? What resistances and inductance values are you assuming for the windings?

PmgR

[seaad]

Hi  pmgr, all

26 V, 30 A

Maybe someone can answer where to that thin red wire goes , starting from first filter cap?

AND this question:   Check in film at 5min. 06sec. Arduino have just Start the program, The supercaps  holds  just below 25Volt and decreasing its  voltage slowly down to about 20 V,  while (1.5) -- 2 Amps is Now feeding the transformer!! Why 2 Amps when the charging of the super caps only took 0.5 Amps??
If there is a built in current limiter for charging?

[partzman]

Partzman, I see only two suitable ways to use this transformer in his setup (without cutting into the windings):

1. Use two primaries in series and use two secondaries in parallel: primary voltage 230V, secondary voltage 115V, so 2:1 turn ratio. Then input 115V on primary, that would give 55V rms on secondary with 24A rating.

With this turns ratio, the ideal OCV at the SC output would be ~76v dc assuming a relatively high esr of the SC's. 

2. Use two primaries and one secondary in series: input voltage would be 345V, output voltage 115V, so 3:1 turn ratio. Then input 115V on primary, that would give 38V on secondary, which is close to what Pierre states as "36V home made transformer".

With this configuration, the ideal OCV would be ~52vdc.  Both configs produce too high an OCV as compared to Pierre's tests.

Let's consider case 2. above: Having 1.5A on primary would then translate to 4.5A on secondary and the voltage drop over the resistor would be  18V drop, or about 20V left for the capacitor.
This looks like it is in the ball-park.

Would this configuration make sense for the magnetization current? What resistances and inductance values are you assuming for the windings?

To meet Pierre's apparent idle current of ~ .100A rms would require a primary inductance of ~3H which could be possible with 1 primary on the core size of the Signal transformer.  This would depend on the core permeability and turns of course.  In my simulation I used a 1H primary and 27mH on the secondary with 3 ohm and 1 ohm dcr respectively and a coupling K = .95.  The primary dcr is maybe too low and should probably be in the 5-10 ohm range but the outcome of the loaded tests were so far off that I didn't change that parameter.

My reason for analyzing the power transformer and SC output was to determine with reasonable accuracy the load presented to the SC pack by the running stator without the rotor in place.  IOW, if the electromagnetic fields in the stator windings are held relatively constant, this means the winding currents would be nearly constant and generally limited by the sum total of the stator winding's dcrs.  This is where my sim analysis really fails!  To reproduce Pierre's results with the stator load, the sim needs ~10-20 ohms to even get close to his dc output voltage and current levels but, the input current draw is too low at ~500ma!   IMO, Pierre's stator should represent a 0.5-1 ohm load if connected as he has revealed and if these values are used, the SC voltage drops far below the ~19.5v dc level.  Perhaps I'm missing something here so I'm certainly open for correction as this is a huge difference.

Regards,
Pm

PmgR