Testing the TK Tar Baby

[MileHigh]

TK:

When I look at the voltage bump on the batteries when the current is reversed it makes me think that that AC battery impedance test that PW suggested would be interesting.

Let's play Devil's advocate and assume that if you did the number crunching based on your scope traces that it would show "more energy being returned to the battery."  In other words, you have a SUCCESSFUL replication of the NERD circuit!!!

So let's assume that the LEDs of DOOM or your ammeter still show that the batteries are discharging.  So right there you have some contradictory data.

PW suggested using a sine wave at a high frequency with voltage drops across resistors to determine the AC impedance.

I can suggest another test that might be useful.  If with your signal generator you could somehow inject a current pulse train that is positive going only (meaning back to the battery positive terminal) with similar timing characteristics to the reverse current (i.e.; back to the battery positive terminal) that you see on your scope trace it would be interesting to see the battery's response.  You could do this on a single battery only with very short wires to reduce inductance effects.

The desire would be to see how much that current pulse train raises the battery voltage every time it is hit with a short current pulse.  You can assume that most of this increase in battery voltage is due to the charging input resistance of the battery.  The delta-V component is therefore not "returned energy" but more akin to energy burned off inside the battery.  So right there that would allow you to attach a weighting factor to the apparent energy returned to the battery.

So perhaps between voltage swings due to wire inductance and voltage bumps due to the battery internal resistance you can start to get a clearer picture of what is actually going on.

A possible scenario is that when you factor out the "return voltage bump" effects due to wire inductance and battery input resistance then the picture of the energy dynamics will change.  Something like 100 units of energy flow out of the battery, and then it appears that 120 units of energy are returned to the battery.  But when you factor in what was said above, in reality only 80 units of energy are returned to the battery.

Even if there is a resistor inside the battery itself, it's still a resistor just like a resistor in the regular circuit and it represents lost energy.

Don't forget the standard disclaimers!  lol

MileHigh

[TinselKoala]

@MH:
What you are describing is an interesting test.... one that should be done by anyone who claims that their batteries aren't discharging while usefully powering a load. Wouldn't it be nice if the CLAIMANT performed these comprehensive tests, before attempting to rewrite some of the basic laws of physics?

Besides, I've seen nothing in the behaviour of this circuit that leads me to believe for an instant that the batteries aren't discharging at a normal rate for the current they are sourcing.

Oh.... sure.... if I believed that a 12 volt battery is fully charged as long as its no-load voltage is over 12 volts, or if I believed that it was impossible to reach current spikes of 8 amps in the negative direction without _something_ unusual happening, or if I believed that One Joule is One Watt Per Second (the terms are interchangeable according to RA), or if I believed without question anything that a digital instrument tells me...... then things might be different.

I've given a lot here. A ten minute video takes at least an hour or more: setup, a basic check to make sure the demo will actually work, one or two false starts, an interruption by the dog, the actual demo, downloading from the camera, converting from the 1 GB high-res file into something that will upload to YT in a reasonable amount of time, the actual upload, reviewing to make sure it's legible, writing the description, posting a link.... all that takes time. Then there's the driving around getting parts, the time spent shopping, the actual _money_ spent on parts and supplies, the lost time from my actual paid job.....

I'm starting to want something back.

People have suggested all kinds of things I "should" or could do to test various aspects of the behaviour of Tar Baby..... just as if I were some OVERUNITY claimant who is trying to apply for a monetary prize and needs vetting. And I've performed and reported on many or even most of these suggested tests. But let's not forget something here: My only claim is that Tar Baby performs just like NERD in all significant respects. If NERD, meaning RA, hasn't shown the results of some suggested test, why should I have to do so?

Much of what I've done should in fact have been totally unnecessary, but I had to do it because "some people" don't understand analog scopes, much less digital ones. They don't understand basic power measurements or the mathematics involved, they don't understand the basics of scientific experimentation, and worst of all they think that they know enough that they don't need to know anything new. Therefore I have had to establish "credibility" for my instrumentation and methodology, things which normally would not need doing.... and something, strangely, that has NOT been done by the NERDs.

All evidence... from their video demo and their "papers"... indicate just the opposite: they are not careful about facts, they make many errors in concepts and execution, and they clearly do not use their equipment properly or to full advantage. SO what am I doing proving MY credibility and instrument capability? This is the FIRST thing that should have been insisted upon in the case of the RA claims: they MUST correct all the obvious mistakes, typos, misrepresentations, incorrect schematics, blown mosfets, and get rid of the "theoretical" distracting musings. FIRST.

Show that the basic data leading to the claim is at least correct and properly obtained. THEN. Show a basic understanding of power, energy, and instrumental measures of the same. At that point, an unusual claim might deserve to be taken seriously and examined further.

But a mishmash of mistakes, errors, scoposcopy, instrument abuse, abuse of persons investigating the claims, and bloviating about some "papers" full of errors and a "theory" that makes no mathematical predictions at all..... this does not even deserve a second glance. Not UNTIL the claimant actually produces the needed evidence: batteries that will power a light bulb for just as long AFTER they have been heating up a load for some time, as they would BEFORE doing that work.

In other words, I want something back: I want to see the NERDs testing their claims, correctly and comprehensively, and SOON. I've even done their home work for them. I've shown that they do not need a FG or even a 555 timer to make the negative-going bias signal that produces the "lowheat" mode of operation.

Either that, or I want to see them formally drop their claim of overunity performance.

Actually, it appears that RA has backed off from this position: lately she claims that she ISN'T claiming OU performance at all... they only "measured" OU performance. Of course this is more prevarication from RA. They most certainly have been claiming OU and COP exceeding infinity. But if they want to officially retract those claims and stop applying for prizes and awards.... then they obviously will no longer be under the OBLIGATION to prove what they claim.

[poynt99]

Well done on the wave forms TK.

Any idea why the difference in AC "battery" voltage?, i.e. 70Vpp (yours) vs. over 200Vpp (RA)

[TinselKoala]

By the way...if there is anyone else out there doing this.... that potentiometer in the gate bias battery (or powersupply) circuit should be a 2 Watt or better pot.

The cheapo 1/2 Watt carbon film pots are too easy to damage. Last night after shooting the last demo I just let the thing run, at whatever current I had it set, 150 mA or something on the Hickok, to heat the load. After half an hour or so I looked over and the current had dropped to 30 mA or so, and no adjustment of the Power Supply proper (not the pot) could get it back--it was very unstable. So I moved the pot slightly and then readjusted the PS... stability reestablished and high current again. What happened was that the power through the pot finally burned another little spot in the resistance layer right where the slider was, and so the pot couldn't make good contact there any more. Moving the slider a little bit to an unburned location, then readjusting the input voltage to reestablish the correct current worked to reestablish stability... for a while.

So use a 2-watt pot here, of good quality. They are a lot more expensive than the generic Chinese 1/2 watt pots that are everywhere, but they will last a lot longer in this kind of service.

[TinselKoala]

Well done on the wave forms TK.

Any idea why the difference in AC "battery" voltage?, i.e. 70Vpp (yours) vs. over 200Vpp (RA)

Thanks.

Could it have something to do with the fact that they are using a 72 volt (or 60 volt) battery pack and I am using a 48 volt one? Or perhaps that the bandwidth of their scope far exceeds that of mine? Or that they probably still have more lead inductance, especially inter-battery, than I do? Or the fact that I am using a limited bias supply voltage, and they have their FG cranked all the way to the negative offset stop? (Oscillation amplitude depends on bias current in this setup, as I have shown several times.)

I think all these things combine to account for that particular difference.

More importantly, though.... look at the current trace amplitudes in the TB and NERD. What accounts for the strong _similarities_ of the amplitudes here? The slight difference in waveshape can be a result of probe compensation and bandwidth. Wouldn't you like to see NERD and TarBaby tested side-by-side with the same instrumentation?

Look very carefully at the ascending slope of Tar Baby's current trace. You can even see the little "knee" there, right where NERD shows a fast spike. The bandwidth of the HP180 isn't fast enough to catch that little spike... but the knee is there and of course with a higher bw scope we'd surely see that same little spike. (At least I think I can see the knee there ... it shows up better at higher drives and timebase settings.)