Exploring the Inductive Resistor Heater

[TinselKoala]

Oh, and Gmeast: If you want your circuit to self-oscillate, that is easy. Just throw out all your nice construction techniques, and do what Rosemary Ainslie did: Use lots of excess wiring, especially on the Gate wiring, and use loose connections like clipleads. I know you don't care about my experience... even though it's a lot greater than yours... but that's exactly how I got Tar Baby to self-oscillate, after a proper tight layout would not. I looked at Ainslie's photographs and added the "rat's nest" of wires. Then the circuit self-oscillated without difficulty.
The magic oscillations you are trying for are nothing more than feedback.

[lanenal]

Did I misunderstand, or did you? This is what Gmeast is claiming and what you apparently are defending:

TK, please read my post carefully. I didn't say that gate current can only get to 100nA. What I was saying is that that 100nA could be referring to the leakage current. Is that clear?


BTW, Greg already measured the a loose upper bound to the possible energy injection through the gate signalling. Even if he was wrong on the gate current, it does not invalidate his result.

[poynt99]

Good question, and my answer was:


The current depends on the frequency, max voltage difference, the current limiting resistor, and the capacity of the hidden cap.

Well, read my two questions carefully; you didn't actually answer the two questions that I asked.

[TinselKoala]

TK, please read my post carefully. I didn't say that gate current can only get to 100nA. What I was saying is that that 100nA could be referring to the leakage current. Is that clear?


BTW, Greg already measured the a loose upper bound to the possible energy injection through the gate signalling. Even if he was wrong on the gate current, it does not invalidate his result.

This is a direct quote from YOU. Please read your post carefully, including the part that I have highlighted in RED:

This circuit should be the easiest to tune as well (Reason: when the pulse is at 6V, all transistors should be open switch, so logically, there will never be battery shorts). The pulse voltage source V5 can be replaced by a 555 astable circuit. V6 can be obtained by using an equal voltage divider (two 5K resisters, for example) sharing the same power as the 555 circuit. All the JBT transistors can be replaced with equiv. MOSFETs. I used voltage dependent current source and a resistor to mimick the behavior of a transformer. You can use 4 transformers instead in your implementation (MOSFETs are probably better than JBTs, I am affraid, as MOSFETs have no gate current). I used no diodes for switching purpose in this circuit. LTspice simulation is performed and the plot of the current over the load R3 is given below (I omitted the two capacitors at each end of the load R3 for the sake of simulation, in your implementation, you should include them).

http://www.overunity.com/6793/simplest-solid-state-tesla-switch/msg156706/#msg156706

You here state that Mosfets HAVE NO GATE CURRENT. That is what the words "MOSFETs have no gate current" which you wrote in the post above mean in English. IS THAT CLEAR?

Now, once again, please explain how that can be true, at the same time that I can light up a light bulb on current passing through the Gate-Source and Gate-Drain capacitances of a mosfet.

The 100 nA cited in the mosfet data sheet as leakage current is leakage current, nobody has ever said that it wasn't. The issue is that both YOU (as shown above in the direct quote from YOU) and Gmeast both seem to think that is the only current that can pass between the gate and the other pins of the mosfet. Now, you are waffling about, since you have been presented with incontrovertible proof that your claims are false, now you are trying to pretend that you didn't make them. But as the DIRECT QUOTATION above proves.... you did indeed claim that "MOSFETs have no gate current." Is that clear?

What did you mean by that statement?  Why does the lightbulb probe light up in my video? Is there current flowing through my mosfet's gate-drain and gate-source capacitances, or not? Is this the current required to charge the gate and switch the mosfet, or is it in excess of that? 

Even if he was wrong on the gate current, it does not invalidate his result.

Of course it does. Crippling the gate driver by using the large inline resistor helps his cause by limiting the current here, but he is still wrong not to include the contribution of the driver to the circuit's overall power dissipation.

[lanenal]

This is a direct quote from YOU. Please read your post carefully, including the part that I have highlighted in RED:http://www.overunity.com/6793/simplest-solid-state-tesla-switch/msg156706/#msg156706

You here state that Mosfets HAVE NO GATE CURRENT. That is what the words "MOSFETs have no gate current" which you wrote in the post above mean in English. IS THAT CLEAR?

Now, once again, please explain how that can be true, at the same time that I can light up a light bulb on current passing through the Gate-Source and Gate-Drain capacitances of a mosfet.

The 100 nA cited in the mosfet data sheet as leakage current is leakage current, nobody has ever said that it wasn't. The issue is that both YOU (as shown above in the direct quote from YOU) and Gmeast both seem to think that is the only current that can pass between the gate and the other pins of the mosfet. Now, you are waffling about, since you have been presented with incontrovertible proof that your claims are false, now you are trying to pretend that you didn't make them. But as the DIRECT QUOTATION above proves.... you did indeed claim that "MOSFETs have no gate current." Is that clear?

What did you mean by that statement?  Why does the lightbulb probe light up in my video? Is there current flowing through my mosfet's gate-drain and gate-source capacitances, or not? Is this the current required to charge the gate and switch the mosfet, or is it in excess of that? 
Of course it does. Crippling the gate driver by using the large inline resistor helps his cause by limiting the current here, but he is still wrong not to include the contribution of the driver to the circuit's overall power dissipation.

TK, what a great find -- I wonder what tool you have used, looks like you've got CIA team behind you .


In that post, I was talking about the time average current (cause that's what matters in that case) while ignoring the leakage current, in that case, it is obvious that the time averaged gate current is zero.


lanenal