Here is some more work on driving the Tar Baby with an internal timer rather than an external function generator.

Fig.1:  The Tar Baby Revision C schematic reproduced again, showing the working, floating, externally powered 555 timer circuit that produces a strict negative-going gate drive pulse with oscillations, which I have been using for several days now.
ETA: The diagram shows 3 x 12 v main batteries, but I've been using 4, and of course with proper transistor heatsinking up to 6 batteries (72 volts nominal) can be used.

Fig 2: A 555 - based voltage inverter circuit that will allow the Fig.1 timer circuit to operate NON-floating and powered by the most negative battery in the main circuit's supply, with no external battery needed. This works but makes a bit of noise on the oscillations.... perhaps a filter stage to keep the oscillations from affecting the inverter itself  may be needed.

Fig 3: The Voltage Inverter installed on TarBaby and powering the 555 timer board for testing. Input power at the top, power connections to the timer board can be seen clearly. Red is positive, black is negative.
(I didn't have any 470 microFarad caps so I used 2 ea. 220 microFarad 35 volt caps in parallel for each of the 470s. Obviously component values aren't too critical.)


HOWEVER: since the TarBaby (and NERD) need that negative bias current on the Q2 source pins in order to make the oscillations..... neither Rev C alone nor the Rev C + Inverter will work in the positive pulse mode. That is... they work fine to turn mosfets on in that mode, but since they can't also provide the negative bias current at the same time... no oscillations are made.

I explain this need for negative bias current in the latest video, and show why a simple 555 circuit isn't going to work for both modes without some means for providing that negative bias current.

http://www.youtube.com/watch?v=XkNDYaFVc7g

(I'm posting this here _provisionally_ .... if the noise in the thread continues, then I will carry on further discussion on the YT comments and PMs and leave this thread for the fishes.)

TK,

What is the current draw on your 555 circuit while the TB is oscillating?

PW

TK,

What is the current draw on your 555 circuit while the TB is oscillating?

PW

200 mA is not unusual, but in the "quiescent" mode with no or minimum oscillations it's a few tens of mA. The 555 gets hot, depending on operating mode, hence the heatsink.

I'm still not completely sure about this; there is massive potential for groundloops in this circuit and test arrangement.
Yesterday my _BNC probe connectors_ at the scope were heating up at one point... I realized the whole  system was actually getting its power from one probe's ground lead at the neg batt terminal, up to the scope, thru the connectors, to the other probe's ground lead, and down into the circuit, because I had forgotten to hook up the main power negative battery lead to the ammeter at the right place.

I'll be doing more measurements on this later on, but I've got to do some "real work" today so I won't get to it before late tonight.

ETA: OK, here's a quick measurement. The above values were without the Voltage Inverter feeding the clock. So by increasing the voltage on the supply to the clock, I could get the oscillations to grow larger and the system could draw more current.

Now I've inserted the Voltage Inverter, and still using a regulated supply, I can take the input to the Inverter up to 15 volts (as high as I dare for chip protection) and I get about 55 mA draw indicated on a good moving-coil meter in series with the regulated supply. This only gives me about 90  mA indicated on the main inline DMM ammeter at the main battery, and appears to produce only a little load heating. So the Inverter isn't allowing the required current to flow through the clock, apparently. I still get nice oscillations (in the negative gate pulse mode) but not of sufficient amplitude to allow much load heating, apparently. I've not yet taken current readings when the Inverter is powered by the most negative main battery. That will have to wait, I've got to get real here for a while.

MH,

I do not believe it is acceptable to ignore the current draw from whatever is used to bias on Q2.

Regardless of whether an FG or a 555 circuit, etc is used, something must provide the current necessary to bias Q2 on and if ignored, represents an error term in the final power calculations.

A pair of center-tapped batteries or an isolated DC to DC converter could be used to power the 555, but the current necessary to bias Q2 on has to come from somewhere.

Even with an efficient isolated DC to DC converter operating off the main battery string, the converter will have to draw both its quiescent current and the Q2 bias current from the batteries.  Without that current, Q2 cannot be biased on and the circuit will not oscillate. 

Personally, I would consider using an FG as OK, if its power contribution was both measured and calculated into the final power calculations.

PW   

PW:

Personally, I would consider using an FG as OK, if its power contribution was both measured and calculated into the final power calculations.

Indeed, I agree with you.  And I was composing my last posting as you were composing your posting.  The first go round when I forgot about the current draw requirements to bias the Q2 array I felt dumb, like I was slipping.

Like anything as you get older, if you don't exercise the brain cells, then they start to atrophy.  When I first looked at the free energy forums, it had already been 15+ years since I worked as an engineer.

MileHigh

This seems to be getting confusing. Q2 does not require any significant "bias" to turn it on. It's a MOSFET remember guys? What IS needed is a relatively low impedance AC path for the MOSFET SOURCE and GATE legs to ground.

The real DC bias and AC paths can be completely separated, as I have done with my design. This circuit is not that complex. You give the MOSFET a lot of stray inductance on its leads, a little forward DC bias, and off she goes.....oscillation.

No, perhaps we aren't being clear. We are not talking about turning Q2 on. We are talking about supplying a negative potential to the Q2 _source_ pins, with sufficient oomph to put 200 mA in there, to induce and maintain the oscillations.

As Stefan has inquired and as I have just confirmed with live circuitry, this can be done with a 9V battery and a pot, no switching just continuous oscillations. This has to be "negative" though: the positive from the pot/battery has to go to where the "negative" FG lead is connected to the NERD circuit: the gate of Q1 and the sources of Q2. And the negative from the pot/battery goes to where the "positive" FG lead connects to the NERD circuit: the gates of Q2 and the source of Q1. Thus, no mosfet sees a positive gate potential at all, but the source bias causes the oscillations and allows some current through the oscillating mosfets... and the source bias _source_ is contributing significant power to the system, at anywhere from 60 to 200 plus mA.

I've got to hit the road, so I won't be able to post again until later this evening. Thanks for your comments, and there is a new video up with a bit of a puzzler that I could use some help with.

.99,

I have suggested a few times that a 50R be placed in the source leg of Q2 and that the gate be biased at a positive voltage to keep Q2 operating (similar to your circuit).  A 1meg pot across the first Batt+ with the wiper tied to the gate circuit (AC decoupled) would allow the operating bias of Q2 to be set/varied.  This would only draw 12-14ua. from the first battery (depending on its voltage).  Alternately, the gate coud be tied to the first Batt+ directly through a series resistor/cap to ground (for decoupling) and only the decoupling cap and gate leakage current would be drawn from the first battery.  The value of the 50R at the source could be modified to set the Q2 operating (bias) current.

PW