Partnered Output Coils - Free Energy

[tinman]

With a FET that has a small gate charge that is true.  The FG has a 50 Ohm output impedance that forms a pole with the gate capacitance.  Overdriving does speed up the turn-off.  Tinman's scope may be too slow to see the difference.

I have the 50 ohm impedance switched off.
I will lower the gate resistor value today,and raise the offset on the SG signal.

[tinman]

The duration and flatness of the observed -DC offset during the secondary off period sure looks more like it should be the "return to ground" point as opposed to a steady offset.  I would question everything, including the scope, prior to fully accepting that amount of offset as being "real" (as in the load actually being driven during that period).  Perhaps scoping a CSR inline with the bulb would be telling.

This we can sort out quite easy
And a CSR is no problem at all.

[tinman]

The two test requested by TK.
So im seeing close to a 2 volt drop across the mosfet's,but i think that the drop may be across the DMM amp meter,and not the mosfets. I will run test again without the amp meter,as it only has a 10amp rating.

[MarkE]

So a little design tip is that you want the diode anode to connect as close as possible to the MOSFET drain, and at the diode cathode you want to add a capacitor connected to the MOSFET source, all with minimum area loop from the MOSFET drain through the diode and capacitor back to the MOSFET source.  The connections off to the coil aren't that important other than from a radio noise standpoint, which you can suppress by tightly twisting them together.  Then you want to put the scope grounds at the MOSFET source.  Then you want the signal generator black lead to connect as close to the MOSFET source as possible.  If you take all of those measures, then your falling edge waveform should clean right up.

[picowatt]

Tinman,

Have you been connecting one of the scope probe grounds to the drain of the MOSFET in all of your tests (or just the recent ones)? 


As for verifying the flat line -DC offset observed at the inner secondary during the off period (what you highlighted in red in my previous post), if you would indulge me perform the following test: 

The following assumes you have verified that there are no significant leakage paths between the various windings of the XFMR under test (i.e., they have been verified with an ohmmeter).  In particular, ensure there is no significant leakage path between the external primary and the inner secondary utilized during the following test.

It also assumes that you have verified that the scope channels, with the vertical sensitivity set to 5V/div (I believe that is what you have been using) produce a zero volt display when the probe tips are shorted to their respective grounds (i.e., no weird offset observed in the scope channels). 

It also assumes that the scope channels are set to DC coupling, and that you have verified that DC coupling is indeed functioning properly by connecting a small battery across each channels probe tip and ground for 30 seconds or so to ensure that the scope maintains a steady indication of the DC voltage of the battery used.

(Your observed "effect" is weird enough to warrant confirming all of the above...)

Disconnect the inner secondary(s) from the drain of the MOSFET, only the low side of the primary should be connected to the drain.  Use the BATT- as the common for the FG ground.  Connect one scope channel's probe tip to the MOSFET drain and that probe's ground to BATT-.  Connect the other scope channel's probe tip to one side of the secondary under test and that probe's ground to the other side of that secondary.

Set up your drive signals as you have been doing so that you are confident your FG settings are switching as in previous tests and that you are producing the flat line -DC off set on the secondary during the off period. 

Once you are satisfied the circuit is performing similarly to previous tests (i.e., demonstrating the flat line -DC offset at the secondary during the off period), disconnect ALL scope probes AND scope grounds from the circuit, including your external trigger connections.

With the inner secondary being tested still completely isolated from all other circuit connections, connect a lamp across that inner secondary.  Attach one channel of of the scope across the lamp, connecting the probe tip and probe ground directly at the two lamp connections.  You can just trigger the scope from that probe's input channel.  Other than the lamp and the singular scope probe/gnd, there should be no other connections to the secondary.  As well, the scope should also have no other connections to the circuit, just the one probe/gnd across the lamp connected to the secondary.

If you continue to see that flat line -DC offset on the secondary during the off period, connect the scope probe (still connected as above) to the other scope input and with that scope channel set similarly to the first channel used, note if there is any observed difference in the waveform.

(Alternately, you could just connect a second probe across the lamp (paralleling the first probe/gnd) and connect the second probe to the other channel of the scope so that you can look at both probes/channels at the same time to observe any difference).

Please post your observations. 

Thanks,
PW