Testing the TK Tar Baby

[TinselKoala]

polln8r, thanks for posting that.

As you can see the difference, either way, is very small. And that's the point. If there were a large "reverse" total to the current, or even a "forward" one, it would make more than a few pixels difference out of many thousands of total pixels in the areas. The net current is tiny, probably below the resolution of our technique here.

But this is an AC signal! It dissipates power in its load in both phases of the oscillation! If you plug your toaster into the wall, does it heat up only on the Red phase and return POWER to the wall during the Green phase? I don't think so.

These AC oscillations are going nowhere, really. If one side of the slosh is a little greater than the other side of the slosh, that is because some  has sloshed out of the bucket, as heat in a resistance or radiation as RF. But no matter... there is plenty more being dribbled in from the battery to replace that which is sloshed out, so the overall amplitude of the oscs doesn't decay, as long as the battery is able to supply a little bit to replenish them.

But the oscillations aren't even important!! They are a Red Herring in the masterful tradition of red herrings. Look at SCRN0235 !!
Massive power is transferred, heating the load and the mosfet massively, and this is the origin of the massive claims about load heating and current and all the rest. And the oscillations aren't even important in making this heat ... It's right there!

[poynt99]

Ok.... so what about SCRN0235, then?

Please....anybody...  look at this trace and compute the DC power level.

Sorry TK, about all I can motivate myself to do right now is calculate the overall average power of about 13W.

[TinselKoala]

More weirdness.
Here's a timeline that goes from about March 27 to April 8 2011, as chronicled in the locked thread.


She fixed the two blown mosfets but still couldn't get the required power products.

She couldn't get it working and thought she had a problem, a bad FG (offset apparently not working), so she replaced the FG with a new one.

Then she still couldn't get the circuit working and had to call in help.

After several days she reported that the circuit was now working but didn't explain why.

I think only one person actually asked her what was wrong and how it was fixed... and of course she did not deign to explain.

Right in here she also admitted that she omitted data that didn't fit her expectations from her reports, and was challenged on this by evolvingape and others.

Then she reported the battery fire and meltdown of the battery clips... and who knows what else.

Then... the next scopeshots she posted were these, the Turnips. I have not seen anything like these on Tar Baby that I can recall.


Of course I haven't exploded any batteries either.
I don't know what is making the strange envelopes, but I can tell that there isn't any DC current during the non oscillation portions. But without a gate drive trace who knows whether that is expected, or not. More Etch-a-Sketch Ainslie scoposcopy rather than clear display of real data.


Compare, contrast, discuss.

[TinselKoala]

Sorry TK, about all I can motivate myself to do right now is calculate the overall average power of about 13W.

Can you please show how you got that?
Do you agree with the numbers I extracted from the scope trace and put into the unsolved equation, or did I misread something or misapply the algebra somehow? Are there not 4 amps flowing around at 72 volts during the Gate HI time?

[mrsean2k]

I make it 130W


(or 130.064516129 :-))


I agree with your method concerning how you arrive at that figure given the duty cycle etc. reported (relieved, huh?), and I can follow some of what you've done from the scope now you've done it, but I wouldn't be able to read that from scratch


S



ETA: I didn't want to divert from more knowledgeable people who've followed more closely and can interpret those screenshots by asking to be fed the meaning of each division on the graticle etc. but I wouldn't mind if it doesn't disrupt.