Testing the TK Tar Baby

[Rosemary Ainslie]

MilesAsEverOffCentre

The cyan is the gate voltage, correct?  Roughly 20 uSec "high" and 20 uSec "low" with the superimposed oscillations.  The "high" is really bad and resembles an exponential decay.  I am at a loss to explain that.

The gate voltage in this capture is pretty gross.  20 uSec is a relatively long time so you wonder what is going on during the "high" phase.

Honestly, in seeing that weird gate waveform I would have investigated it further.  I am assuming that Rosemary and the NERDs just stared at the screen and made the capture and did nothing beyond that.

I can't really venture an explanation for what is going on there beyond stating that we know that when the gate voltage is high we should see current flow.  When we don't see current flow we should assume that something is amiss.  I think that "kindest as always" Rosie Posie did battle with PW over that one.  That was an irrational fight, poor Posie Posie should have simply acknowledged the issue.

Poor Rosie had to point out that the READING that PW was making was based on the DC coupled value.  He concluded that the the LeCroy had bombed or the MOSFET had blown.  He was wrong on both counts.

I've sent you a PM MileHigh.  You need to do some rather elementary corrections to one of your posts.  It'll spare you some embarrassment. 

Rosie Pose

[MileHigh]

Just some more musings.  The heart of this issue is the apparent recording of more power being returned to the battery than supplied to the battery - "COP infinity."

The root of the problem is that Rosemary and the NERDs never tried to investigate this further.  They never tried measuring the net current flow with a digital multimeter - as basic a measurement as that.  Instead, they put too much reliance on the DSO itself.  Also, it's readily apparent that they did not have the knowledge, experience, or skill set to investigate what was going on.  This is not a simple pulse motor, this is an oscillator that runs in the megahertz frequency range.  At this frequency and looking at the size of the setup and the lengths of the interconnect wires, inductive and capacitive effects come into play that you normally don't have to deal with when you are working with a pulse motor.

Anybody that has followed the Dr. Stiffler clips knows that he often makes reference to keeping lengths of wires short, etc.  This is not a place for newbies to be playing.  Figuring out exactly what is going on when you are playing in the 1 MHz to 10 MHz frequency range is not trivial.   The actual full bandwidth of the signal (say 90% of the power) probably extends out to 50 MHz.

The simple DC current flow measurement showing clockwise current is allowing you to infer that the batteries are discharging.  So you can do a simple dim bulb test with smaller batteries to avoid all of the complications of understanding the dynamics of the two-megahertz oscillator and see what the 'real truth' is.

MileHigh

P.S.: Rosemary I read your PM and I have nothing to change.

[MileHigh]

I've sent you a PM MileHigh.  You need to do some rather elementary corrections to one of your posts.  It'll spare you some embarrassment.

Oh really?  How is that Dale Carnegie course coming along?

[MileHigh]

Rosie:

Poor Rosie had to point out that the READING that PW was making was based on the DC coupled value.  He concluded that the the LeCroy had bombed or the MOSFET had blown.  He was wrong on both counts.

The LeCroy is always supposed to be DC coupled for all of your signals, and it's critical for the current sensing to be DC coupled.  So your point doesn't make sense.  The reasons for the missing current flow can be speculated upon, as was done by PW and others.

I know that you had some kind of strange fight over AC vs. DC coupling and it's another one of those cases where there is no issue.  I don't remember the details because I was 'filtering' myself.   

MileHigh

[TinselKoala]

@MH: I recommend that we use the color-code of the NERD device to indicate where the bias supply (FG or 555 or battery or whatever) connects. In the video, the FG's RED alligator clip, the center conductor, the "positive" or "probe" as Ainslie calls it, is shown connected to the gate of Q1 and the sources of Q2. And the BLACK alligator clip, the outer conductor, the "negative" or "terminal(?)" as Ainslie sometimes calls it, in the VIDEO was connected to the common ground bus BUT in the "approved" schematic we have been using is on the transistor side of the CVR.
Right?
As shown in the first diagram below. OK?

Now, for the negative bias situation, the FG or other bias source has effectively  switched polarity and is supplying a (relatively) Positive voltage to the BLACK input point (which is really held at "ground" potential, but voltages are always relative to some reference.) As in the second diagram below.  Right?
The _relative_ positive voltage at the BLACK input point causes the RED input point to become "even more negative" than the negative pole of the main battery supply, thus biasing Q2 (Festus) into the linear response region and passing some DC current and some AC current too. Oscillations happen. But.... look at the zener in Q1 (Moses), with respect to the polarity of the Bias Supply. I don't know if there is any other "end" to this current path or if it contributes to load heating, but I can think of some easy ways to find out.

Certainly in the Positive bias mode the "opposite" transistor's zener plays an important role, as GL has demonstrated both theoretically and confirmed by several experiments independently replicated. I am not that much of a theoretician.... so I'll just test, vary variables and hold constants constant, and record the results whatever they might be. It's wonderful not to be burdened by the necessity to prove a "thesis" of some kind... rather, I examine the data and draw conclusions based on it. In stark contrast to collecting only data that could support my idee fixe of a "thesis".

NOTE: With a scope attached as shown in the second diagram below DO NOT TRY TO TAKE OTHER READINGS, because here the scope is using a different REFERENCE than usual. Other readings require the reference at the negative rail, and most oscilloscopes have all their probe ground leads connected together inside the instrument.