Testing the TK Tar Baby

[TinselKoala]

Tk,

okay, your last circuit diagramm
http://www.overunity.com/12182/testing-the-tk-tar-baby/dlattach/attach/98077/
is not much different than mine, only that you connect the plus pole of the 9 Volt battery still
behind the shunt and I before the Rshunt resistor...

I guess there is still some misunderstanding going on. My corrected diagram (sorry about the wire colors in the video, it's been corrected now) has the battery polarity inverted from your diagram. This seems like a big difference to me... and I'm fully responsible for the error in your first diagram, because of my confusion of the battery wire color coding.  But that's fixed now, as soon as I was aware of it. So the battery in my circuit is completely floating, yours has a connection to ground, and my circuit has the battery _positive_ connected to the "FG minus" point of the circuit, as noted. These may be small differences on paper but they make a "big difference" in the circuit.

Okay, so the voltage on the Rshunt seems to matter...

Okay, so lets use your circuit from now on.

But in your circuit, when you have the pot tap at about the center there also can not
flow 200 mA !

I'm sorry if I gave that impression. But I thought further posts should have cleared this point up. The oscillations START at a relatively centered pot position, with about 12-14 mA indicated on the Hickok meter in series with the 10R inline resistor. As the pot is turned more toward the END of its travel, the oscillation amplitude increases and the current indicated increases, until _with the depleted battery of 7.9 volts AND the 10R series resistance AND the pot turned _all the way to one end_... the most "positive" end of the pot--- the oscillations reach a nice strong amplitude and the series meter reads around 100 mA. If I use a fresh battery or an external power supply, I can indeed reach 200 or even 350 mA by using more voltage from the potentiometer's slider. The position of the slider, hence the overall resistance, hence the current seen by any meters or CVRs,  will of course depend on the voltage applied across the potentiometer's end terminals.  Ohm's law still works, don't worry. It works especially well if you use the right numbers as input data.
I don't think I ever said anything about there being 200 mA with the pot centered, as you are assuming. If I did, I misspoke or perhaps you misunderstood. If you point out where I said it, I'll correct it right away.
This is not the case... unless you want to apply more voltage across the pot.

At best about maybe 10 milliAmps as the 9 Volt battery then sees at maximum 5 Kohm resistance
if you don´t turn the knob on the pot almost to one of the ends... but as you said,
it needs -4 Volts it surely is in the center. so no low resistance for the 9 Volts battery...

Once again.... as the pot setting is turned towards the positive side.... the _indicated_ voltage it is putting out goes more and more negative until it reaches a true -4 volts and the oscillations begin with about 14 mA from the bias battery. Now... further turning of the pot doesn't cause an APPARENT increase in the voltage to more than -4 volts... because the mosfets are now turning ON so the voltage stays LOW, and further "voltage increases" on the pot setting only send more _current_ into the system. If you set the pot at some value near the end, where you still read -4 volts on the oscilloscope channel or a simple voltmeter, and you have that big current flow like 100 mA.... and then you unhook the pot wiper from the main circuit and measure the open circuit voltage at the pot wiper... it will be much more negative than -4 volts. So with the pot IN CIRCUIT, you might read -4 volts at the wiper and the pot might be near the center, when the oscillations just barely start. But as you turn the pot to increase the amplitude of the oscillations, the _indicated_ voltage at the wiper, in-circuit, will stay at that "voltage floor" and will remain there _even when the pot is fully turned_ and you know you are applying the full 9v battery voltage, or that of whatever power supply you have there instead.
This is a "voltage drop" due to a heavy (low resistance) load being placed on the battery or other bias power supply like the FG output. My Interstate F43 FG will do 40 volts p-p into 50 Ohms, but even it is drawn down to only -4 volts _indicated_ when it should be putting out -20 volts... and if it's disconnected, at the same knob settings, it will read -20 volts.

ALso this Q1 and Q2 is mostly a multivibrator circuit in selfrunning mode (without LRC units, just the wire lenghts),
so you would only draw mainly AC from the 9 Volt battery.

I am afraid I don't know how to draw AC from a DC battery. So I suppose I am failing to understand just what you mean. The polarity of the current coming from the battery never reverses, at least I don't think it does. I'll check it again later, perhaps on a video, if I have time.
A rising and falling voltage or current waveform doesn't indicate "AC" unless the _polarity_ actually reverses at some point in the cycle. It's a common error though; I've even had to explain it to working electrical engineers. An "AC" or RF ripple on top of a large DC offset or constant voltage does not add up to "alternating current" as we normally think of AC, which means a net current flow in one direction, followed by net current flow in the other direction, in a repeating cycle. The ripple actually means that the _magnitude_ of the current, always flowing in the same direction, varies regularly at the "AC" or RF frequency. Think of it as a pulsating garden sprinkler. Even though the water is flowing strongly, then off, then flowing strongly again, then off, and so on... the water never flows back into the hose.

Maybe you can show just the voltage on your 10R to 50R Ohm pre-resistor, then you will probably see,
that there is not flowing 200 mA !

Regards, Stefan.

I am very sorry but I still don't think we are communicating, somehow. You still seem to be basing all your calculations on the assumption that the pot is centered when the 200 mA is flowing... and of course with 9v across the pot that can't be right. However your assumption is wrong: the pot isn't centered when that current flows, especially not with an almost-dead 9v battery, and with the depleted battery and the 10R series resistor, I can only get about 100mA max anyway. The 200 mA figure is what _can_ be attained without pushing too far, but of course using more than the voltage of a depleted 9v battery. 350 mA can be attained as well if you like, by  using more voltage at the bias source, if you like to see a scope screen full of garbage distorted oscillations.

I have indeed looked at the voltage drop across the series 10R on the oscilloscope. If I did this right... it was late last night and I was pretty tired... the current flowing in this resistor is "always" in one direction --it is not AC. There is a ripple on top of the voltage drop corresponding to the oscillations, which can be seen at any point in the circuit.
The value of the voltage drop across the 10R (actually 10.3 ohms measured) indicates actually a bit _more_ current than is indicated on the Hickok meter -- but of course this is an "eyeball" reading of a small scope deflection viewed at high gain. I have also looked at a 0.3 ohm CVR in series with the negative output of the battery/pot system... and it also more or less agrees with the meter's reading. I'm glad I have that rugged Hickok meter in there, because I'm always pegging it against the 100 mA stop. The bias current is real, even if where it's going is still something of a mystery, and in the circuit as I have it wired, I don't think it can go through the main CVR so it wouldn't be taken into consideration by the NERDs as power input. But it is power input and should be so considered.

Please, if you have more questions, ask them and I'll try to answer clearly. And certainly, if you or anyone finds anything that seems to be a discrepancy, let me know so I can address the issue and correct my mistakes, or even, if necessary ... correct yours.

[hartiberlin]

Hi TK,
http://www.overunity.com/12182/testing-the-tk-tar-baby/dlattach/attach/98075/

http://www.overunity.com/12182/testing-the-tk-tar-baby/dlattach/attach/98077/


are basically the same circuit, just you put the plus pole behind the Rshunt resistor and
I did put it before...

So the voltage drop at the shunt resistor seems to matter...

Just show us in a new video a scopeshot at your 10R - 50 R resistor
this will be interesting what current will be flowing there.

Please use a DC coupling on the scope so we can also see the DC current level.
and please show where the scope´s zero line is set and please show Volts per div.

Many thanks.

Regards, Stefan.

[TinselKoala]

Hi TK,
http://www.overunity.com/12182/testing-the-tk-tar-baby/dlattach/attach/98075/

http://www.overunity.com/12182/testing-the-tk-tar-baby/dlattach/attach/98077/


are basically the same circuit, just you put the plus pole behind the Rshunt resistor and
I did put it before...

Yes, I can see that now. Thanks... you are right, it's not as big a difference as I at first thought, but it does make a difference in the circuit's behaviour especially if there are probes hooked up here and there, I think.

So the voltage drop at the shunt resistor seems to matter...

Just show us in a new video a scopeshot at your 10R - 50 R resistor
this will be interesting what current will be flowing there.

I've just calibrated my Hickok meter (see video and photo below) against a known resistance using a regulated power supply and Ohm's Law. I've made a video illustrating the calibration, and it also shows the use of the oscilloscope to make a current reading in the calibration system... just to prove that one doesn't need numbers in boxes to make reasonably precise current readings on an analog scope. The test is a tough one too: I use a 0.3 ohm CVR and the scope is set to 0.005 volts per division, and less than one full division is used for the reading.... and I get within 5 percent of the true value. Twice.

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

Next, I'll do the demonstration you asked for, using both the Hickok and the scope across the 10R as a CVR. I've got a fresh 9volt battery too.

Please use a DC coupling on the scope so we can also see the DC current level.
and please show where the scope´s zero line is set and please show Volts per div.

Don't I always? How would it be possible to read a current value from the scope if I didn't do these things? Never mind, it was a rhetorical question.

(I think someone may have told you that I don't show quantitative measurements on my scope... but in fact I do, almost always. Sometimes I leave out the timebase setting, but I always try to show the coupling, the channel vertical amplifier settings, the probe attenuation, how the scope is triggered, and where the zero baseline is, and I do try to illuminate the graticle so the divisions can be seen. Things move fast in my videos; I encourage liberal use of the pause and rewind buttons.)

Many thanks.

Regards, Stefan.

You're quite welcome.
-TK

[TinselKoala]

Quantitative scoposcopy and Tar Baby's Bias Current

http://www.youtube.com/watch?v=OV0Qvzumhhk

[TinselKoala]

Tar Baby and the Coup de Grace 

(main CVR measurement)

http://www.youtube.com/watch?v=VWTcGFYqbF8