another small breakthrough on our NERD technology.

[MileHigh]

TK:

I hope that my comments will help with the caveat that I am not going to be digging into the low-level nuts and bolts of the circuit, I have been there and done that.

What the RAT team focused on was the oscillation that took place specifically when the function generator was outputting -5 volts.  So they had the function generator negatively offset such that the square wave was outputting 0 volts and -5 volts.  They slowed the function generator down to see longer bursts of oscillation at -5 volts.

So what that means is that the function generator has absolutely nothing to do with the actual operation of the circuit.  When you look at the schematic you will see that with -5 volts output, that Q1 is always off and the Q2 array will be in oscillation with the only path for the current to complete the loop being through the function generator itself.  So I don't see a possible solution with opto-isolators.

So, you could completely eliminate the function generator with a 5-volt battery in series with a 50-ohm resistor.

What you will notice if you follow the voltage drops as you go around the main loop, is that the function generator's -5 volts is added to the main battery bank's voltage.  So if we assume five 12-volt batteries the total voltage that's powering the circuit is (5 x 12) + 5 = 65 volts.   In other words, 7.7% of the power provided to the circuit comes from the function generator itself.  There are three main power dissipation points in the circuit; the inductive resistor, the Q2 MOSFET array, and the 50-ohm resistor inside the function generator.

It just occurred to me in looking at Poynt's reference schematic in posting #1505, that you notice that the current flow completely bypasses the 0.25-ohm CSR.   Is it possible that this observation slipped through the cracks?  If yes it would mean that the CSR data is useless and all of Rosemary's DSO calculations are invalid.  I am not sure if this is true but it may need to be investigated.

So the bottom line is this:  -5 volts from the function generator will induce a forward bias on the Q2 array's gate-source voltage and the Q2 array will switch on.  Since MOSFETS are prone to oscillation, and not forgetting the stray inductance in the wires, and not forgetting the 50-ohm output impedance of the function generator, the gate input of the Q2 array is just itching to be 'tickled' and all of the ducks are lined up to tickle.  So you get spontaneous oscillation.

As we know, spontaneous oscillation is more an annoyance than anything else, but Rosemary thought that she discovered something.

All in all, it's a nonsensical circuit.

I hope my little treatise helped.

MileHigh

[TinselKoala]

@MH: what you say here agrees with my own findings, except that I don't believe the part about them only having the FG set for negative output. She has said that they used the offset control, and as my work has shown the scope trace won't reflect the true output that the circuit sees when the FG is loaded down by it. If you watch my latest video you can see what happens with symmetrical square wave, positive going only, and negative going only offsets. There IS a region right in the middle where ALL the mosfets are being turned on and the oscillations are present on both phases of the signal. This can be seen in the video, and the current measured from the battery by a simple inline DMM..... which WE HAVE NEVER SEEN FROM ROSEMARY....... duh, why hasn't anyone insisted on this?----- shows just under 2 amps at that moment, and the mosfets-- all of them-- get hot fast since they are improperly heat-sunk. FFS, I just superglued the heatsinks to the backs of the mosfets !! While the more typical settings that make traces that look just like hers give me a typical 170-200 mA from the battery... and yes only the Q2 mosfets are turning on, as attested to by my burnt fingertips. Oh...and the noncontact IR thermometer too.

It should be clear by now that the main function of the function generator is to... wait for it.... turn OFF the oscillations. You can clearly see that when the FG's output goes HIGH in my system's scope shots... and in Rosemary's too, the blue trace is her FG...the oscillations go off.

(And yes.. I certainly did notice, and have mentioned before, that the CVR isn't properly seeing all the current flows in the system. Of course her "data" here is invalid.)

ETA: I've been reading the thread on the other forum from exactly one year ago. It is like an exact image of this thread here and now, except that I am playing the "humburger" role with parts instead of sims, and all the other cast members are nearly the same. It's very eerie. I suggest starting with page 10 or 11; there's no need to go all the way back to 2009. Just start where the discussion of the video starts... and you will think you are here again, but without the troll crone Red Queen shouting "off with their heads" at the slightest provocation.

[MileHigh]

TK:

I think that it's worth it to mention that Rosemary and her whole team worked on this circuit for months, and once they discovered the oscillations when the function generator output -5 volts all of the focus was on this event.  All of the data captures for this most recent incarnation of the RAT circuit were related to this oscillation only.

Rosemary thought that she had some 'secret sauce' and hid the actual configuration of the schematic from the public when she presented her new circuit.  I believe it was Point that reverse-engineered the circuit from the demo clip and then Rosie Posie fessed up.

So, that means for months and months neither Rosemary or any of the other members of the RAT team were even aware of how the circuit was actually operating.  That's why I called them a nasty name suggesting gross incompetence in an earlier posting and I think that it was justified.  That's why I have very little faith that they could do a fairly complicated battery draw-down test properly.

MileHigh

[TinselKoala]

TK:

I think that it's worth it to mention that Rosemary and her whole team worked on this circuit for months, and once they discovered the oscillations when the function generator output -5 volts all of the focus was on this event.  All of the data captures for this most recent incarnation of the RAT circuit were related to this oscillation only.

Rosemary thought that she had some 'secret sauce' and hid the actual configuration of the schematic from the public when she presented her new circuit.  I believe it was Point that reverse-engineered the circuit from the demo clip and then Rosie Posie fessed up.

So, that means for months and months neither Rosemary or any of the other members of the RAT team were even aware of how the circuit was actually operating.  That's why I called them a nasty name suggesting gross incompetence in an earlier posting and I think that it was justified.  That's why I have very little faith that they could do a fairly complicated battery draw-down test properly.

MileHigh

Interesting spin you put on it. Very charitable.

I think they simply had no idea that they wired the circuit wrong in the first place. They tried a single mosfet first, then someone pointed out that it could never handle the current required for decent heat in the load, so they decided to add the other 4. Look at the board: it is NOT the single mosfet that is installed "wrong".... actually it is the 4 on big heatsinks !!  Then they saw the massive oscillations, creamed their jeans, and probably did not realise the difference in schematic diagrams until  .99 pointed it out and MADE her acknowledge it.

Of course they aren't going to do any tests at all, much less properly. The excuse will be either that Rosemary can't find two "academics" that she approves of to participate, or that she and .99 can't agree on some fundamental point of interpretation of instrument readings. This is my prediction.... PROVE ME WRONG. Please.

[MileHigh]

TK:

Yes I agree that if you have a positive swing in the function generator output you get oscillations also and this time the current flows through the CSR.  Magic oscillations abound!

You may want to double-check with Rosemary, I think the only analysis that she is interested in is when the function generator outputs -5 volts.

It sounds like there was always was and there is a real possibility of blowing the 50-ohm resistor inside the function generator.  I think it's two one-watt 100-ohm resistors in parallel.

MileHigh