another small breakthrough on our NERD technology.

[TinselKoala]

@MH: OK, I understand now and will try it after a while.

.99: In the video I realize that I'm oscillating the Q1 mosfet because I've got the 555 pin 3 going to the FG + location on the circuit, but now I've fixed that, simply by routing the Pin 3 to the FG - location on the circuit (duh). Using a variable power supply at 10 volts input to the 555 circuit, I have perfectly stable oscillations on the Q2 mosfet drains (and everywhere else of course) and a reading on the inline meter of about 320 mA..... and the load is warming nicely. (9 volts from the 9v battery wasn't quite enough to get stable in the q2 osc mode.) I found that the 555 gets hot and glitchy so I put a heat sink on it and now it is perfectly stable, has been running the Tar Baby and heating the load with Q2 oscs only for an hour or so, load is up to 104 F.

Rosemary, you are the very definition, the perfect textbook example, of pathological wilful ignorance. Watch my videos or don't.... fall more and more behind the discussion and progress..... lose track of your mendacities while trying to find any reason not to test...

Meanwhile, real work and progress continues, not with your "cooperation" but IN SPITE OF IT.

And as to my contact info: I am not trying to publish a scientific paper reporting an experiment that claims a miracle in peer reviewed journals. YOU ARE. The referees of journal articles are ALWAYS anonymous.

And as to my raw data: every bit of it is publicly available, posted on the internet, to those with the wits to look at it. Yours? Not.

[picowatt]

I ran up one of my simulations of the RAT circuit with two MOSFETs, Q1 and Q2, and you were pretty close PW, the average battery/FG current is about 175mA in my sim running in negative bias mode (Q1 inactive, Q2 actively oscillating).

The average battery power is about -2W, and average FG power about -1W, for a total of 3W to the circuit. I have a 4 Ohm FG resistor, not 50 Ohm. The oscillations don't appear correct if I use 50 Ohms.

It's interesting to note, that in my sim, the VGS does not seem to exceed -4V, even if I set the FG to -9V. Something is limiting it to about -4V.

Hi all,

Lots to do today so I have to keep it short.

.99

In the common gate configuration, the Vgs will always be at the turn on voltage of the device for the current being drawn.  The device will, in effect, self-regulate (pretty fair current requlator, bipolars are a bit better in my opinon, more predictable Vbe).  If the gate is grounded, and a fixed negative voltage is applied to the source via a resistor, varying Rsource will vary Ids with little change in Vgs.  Also, if Rsource is constant and the negative voltage to Rsource is varied, the Ids can again be varied as Vgs remains constant.  At all times, Vgs will be relatively constant.  At the edge of the device turn on (i.e., at very low Ids) things are very non-linear and Vgs will vary a bit as Ids is increased, but once you get close to ohmic, Vgs should be fairly constant (tempco's and the like not withstanding) and vary only a bit. 

The Vdrop across Rsource divided by Rsource is a good way to measure/calculate Ibias.

We know in Rosemary's gen, Rsource (actually Rgen) is 50R.  But, we do not know the Vdrop across Rgen as we do not have the open circuit voltage of the FG so Ibias can only be guessed at based on text comments regarding FG settings.  If you look at all the scope captures, the negative voltage depicted as being applied to the source of Q2 during osc is always a relatively constant DC value (one has to optically integrate the HF noise on the source drive), as it is doing precisely what you are seeing the device do, i.e., trying to maintain a constant Vgs.

TK,

Thanks for doing that experiment last night!  Do you ever sleep?

Rather than use a 555, I would try a variation of .99's burst osc circuit.  Pull out Q1 for now.  Put a 50R resistor between the source of Q2 and the CSR.  Use your F43 as a pwr supply, tie the signal common to the CSR, or to BAT- for now.  Decouple across the FG at the board with some caps such as an electrolytic and a paralleled ceramic, values not critical.  Then run a few inches of wire from the gate of Q2 to the signal side of the two parallel caps decoupling the FG (clip lead?).  Measure VDC across the 50R at Q2's source and slowly adjust the FG towards a positive voltage until the Vdrop on the 50R says you are at 100ma.  Vgen should now be at 5 volts+Vgs.  As .99 has done in his sims, you may have to play with the length of wire (inductance) going to the gate (or coil it up) to get an osc.  Alternately, a bit more wire between the Q2 source and the 50R may be needed to get the osc (emulates FG lead inductance in original setup).  .99 uses a diode from the Q2 gate to the Q2 source to emulate the body diode in Q1.  It should not be needed for osc but it does shape the waveform a bit.  You can give it a try.  Have a look at his schematic. 

.99, possibly you could post your burst osc schematic here?  We could discuss its validity as a possible alternative to Rosemary's circuit.

The nice thing about this method of bias is that the bias voltage applied to the gate will draw very little power (insignificant), as at DC, the gate is a very high R.  Using a 555 or similar in the source leg will require the 555 circuit to carry/handle Ibias.  That will likely require a buffered circuit and a pretty good battery or another supply to operate the 555 circuit for the duration of a rundown test.   If this alternate bias method is not acceptable for whatever reasons, the 555 may have to be dealt with at some point.

Rosemary,

I read your comments and I apologize but I really don't have time to respond at length right now.  I enjoy a good technical discussion but will do what I must to stay out of the "fray".  When I read this thread from the beginning, as soon as I read any "attitude" (from anyone), I scroll on until the conversation returns to a more technical nature, in doing so, it makes this rather lengthy thread a very short read!   

In any event, if you consider my discussions as "off topic", I will certainly honor that opinion and move on.  I mean no disrespect to anyone. 

Have to go for now,

PW   

   

[TinselKoala]

LOL.  That would be a miracle of some considerable proportions.  Golly.  We now have the advent of current flow that is entirely unrelated to voltage.  And the joke is that I'm the one accused of introducing new science concepts.  If there was any basis of truth is this proposal then I assure you that over unity is a doddle - compared to this claim.  Guys - our TK is branching out into new physics which he's trying impose on us with the same abandon as he imposes his definitions of Alpha rays.  Perhaps he should go back to the wiki definition and look up inductive laws.  With luck it will be as badly explained as alpha emissions.  Then he can, indeed, invent his new standard for our standard model.  What a joke.

Rosemary

Once again you are deliberately or stupidly misinterpreting the discussion of the voltage floor that we are seeing in the sim and the real circuit. You are obfuscating real understanding of this phenomenon.
And alpha "rays" are particles: helium nuclei stripped of all electrons and accelerated to moderate velocities by electric fields (since they are charged) and/or nuclear processes including decay of unstable nuclei.
Beta particles are ORDINARY ELECTRONS, moving rapidly, charged, slightly massive and are produced by many processes among which is Beta decay of a nucleon which is mediated by the weak nuclear interaction (aka force).
GAMMA RAYS are photons of electromagnetic radiation JUST LIKE LIGHT, are not charged, are massless yet carry momentum due to their velocity, and move at the velocity of light. The only "particulate" character of EM PHOTONS is due to their photonic nature... which in turn is only due to our lack of understanding. OURs, not yours, you've got it all explained using zipons at 2c, don't you.
Idiot.

[TinselKoala]

@PW: have you not seen the 555 circuit in operation? It is working fine. I can try your variation too, of course... but at this point I don't really see a reason to, since the basic astable 555 works so well. Your point about the power supply or battery is well taken... I think that the 555 is injecting some power from its supply into the circuit and will need isolation somehow...unless that kills the oscs.
Since I have these H11D1 optocouplers here I might try them first.

ETA: Load is at 110 F, inline current 310 mA, 555 supply at 10 volts, everything is rock-stable, Q2s are slightly warm on their heatsinks, Q1 is stone cold... but needs to be in the socket or the oscillations stop. Drain oscillations as shown in the screenshot above in an earlier post.

[picowatt]

TK,

I see you've already jumped on the 555 circuit!

Never mind,

Carry on

PW