Testing the TK Tar Baby

[TinselKoala]

Will somebody else besides me please put a meter in series with the function generator hooked up to Ainslie's circuit, negative pulse drive mode, extra long period, just to SEE if she is right in her assertion?

.99, you can do this in the simulation, I think, and probably already have, and likely, RA has already denied your sim's validity because of it.

We have been talking for some time here about how the bias supply.... whatever it is.... injects power into the circuit. We have shown that the bias supply, whatever it is, is in strict series with the main battery. And we have shown by instrumental measurements -- not handwaving conjectures and assertions by someone who makes unsupported claims every day and is refuted every day on something or other that she's claimed --- actual MEASUREMENTS have shown the existence, the magnitude and the direction of this current from the bias source.  And these are measurements that ANYONE can make (except of course Rosemary Ainslie) and see for themselves: The bias source, whether it is FG, 555, a nine-volt battery or probably even zipons, adds power to the circuit. Even running the bias off of one of the running batteries itself, AS ONLY I HAVE SHOWN, adds extra power to the circuit that would not otherwise be there and drains that battery faster than otherwise.

A nine-volt battery used for the bias source will RUN DOWN fairly quickly.... this has been established many times.

So, since Rosemary is always right, there must be something wrong at Agilent. Those fools... what do they know about test equipment.

Or maybe she just can't open .pdf files because her computer has been hacked again and can't read or understand their explanation of what a function generator IS and how it is USED.

[TinselKoala]

TK, PW.

I would suggest that rather than decoupling across the entire battery array, that each battery receive ample decoupling. In fact, only one battery need be decoupled for this test. Assume all batteries are of the same SOC and internal resistance, and use only ONE for the calculation.

By decoupling only the batteries (and not the interconnecting wires between them), there will be no effect on the oscillations, they will persist. There won't be any need to add more wire.

Alternately, when they are fully charged, there should be much less AC evident across each individual battery, and therefore the measurement could be taken directly across ONE battery's terminals. This is what I would suggest be done first thing. To summarize, I would suggest the following TK:

1) With fresh batteries, confirm the neg pwr measurement by sampling VBAT from the load side of the battery line. This is the "reference" measurement.

2) Do you have a differential probe available? If so, take a VBAT measurement directly across ONE of the batteries, and do the power computation in the scope. If the pwr computation is now positive, then multiply by the number of batteries to get a final value of the whole battery array.

3) If the pwr is still negative, apply several high quality low value capacitors across this ONE battery. At 2MHz, a 0.47uF cap has a Z of 0.17 Ohms, 4.7uF has a Z of 0.017 Ohms etc. Don't bother with any electrolytics IMHO. If you can get about 10-20uF (the more the better) of high quality capacitance across that one battery, it should smooth out the AC quite a lot. Now do the same test as in 2).

She's doing it again, .99, filling up the thread with ridiculous conjectures and insults against me and my "acolytes"..... I've now been elevated to the Pope of Bunkum, I suppose.... so that the real issues can  be buried from sight.

No, I can't seem to find a diff probe here, but I might be able to get around the issue other ways. But I also see no problem with moving the inductances away from the battery stack and just decoupling the whole thing as PW suggested, or individually if I can find enough "highquality" caps. I do have a few, like the little orange ceramic one in this photo....

The only problem I can see with your proposal -- and PW's and MileHigh's etc. is that... of course.... RA will not see them as valid tests nor will she allow them to be applied to her device.

[poynt99]

TK,

Here is a shot of the FG current and M2 (Q2) Source current. Even though the M2 Source current has a higher negative peak current, the average current of both is about -182mA as shown.

[poynt99]

As I suggested TK, I recommend keeping it simple, practical, and easy with as little disruption to the circuit as possible.

Alter ONE battery only (by adding filtering across it if necessary) and do the measurements using that ONE battery. No changes to the wiring necessary, and the oscillations will be unchanged.

[TinselKoala]

As I suggested TK, I recommend keeping it simple, practical, and easy with as little disruption to the circuit as possible.

Alter ONE battery only (by adding filtering across it if necessary) and do the measurements using that ONE battery. No changes to the wiring necessary, and the oscillations will be unchanged.

Thanks for the current shot. What does Rosie Poser say about that?

And sure, I have no problem with the test of a single battery, but she won't accept the measurement of a single battery, you know that. I can do the most negative in the stack without the need for a diff probe, right?