Testing the TK Tar Baby

[TinselKoala]

So.... The C-H appears to be confirming what I have thought since these several days. It appears that in the "q2 oscillation mode" using a negative going gate drive pulse from the 555 timer, and the water heater element load of 10.3 Ohms and 75 microHenry inductance, at 48 V nominal input voltage.... the Q2 mosfets and the rest of the circuit are actually dissipating quite a bit more power than is getting to the load itself.

The mosfets get warm. I think I just might have to do this: I'll suspend the q2 stack in 250 mL mineral oil, insulated like the load is, and run the thing and see which jug of oil gets warmer faster !

How I would like to see some real test reports from the NERD device.

This is unique in the history of "open source" projects, I think. The "inventor" and claimant is being totally uncooperative in performing requested tests and in releasing clear and unambiguous information. Meanwhile... back at the ranch... the evil rival debunker Grand Inquisitor replicator is building who-knows-what and just might beat the NERD RATs to the vaunted OverUntied Prize.

If he can just figure out what those squiggly lines really REALLY mean, that is. Everybody knows that you can't use analog scopes for anything.

Oh never mind.

[MileHigh]

TK:

Yes indeedy in negative oscillation mode more power is dissipated in the MOSFETs themselves.  Bandwidth issues notwithstanding it looks like the C-H is in the ballpark.

The key thing to think about is voltage drop.  Whomever is "gobbling up most of the voltage drop" is the guy that dissipates most of the power.  So if you put your multimeter across the inductive resistor you might get a decent average DC voltage measurement across that component.  Again, having no direct experience and it being so long, I am not 100% sure.  Poynt has some wonderful clips on YouTube that show how the dual-slope integration hardware for getting an average DC voltage measurement works amazingly.  I am just not sure if it extends out into the megahertz range.  Of course you could put a simple low pass filter between your multimeter and the inductive resistor that doesn't disturb the oscillations.

So if you have a 48-volt voltage source and you notice that your 11-ohm load resistor is only responsible for about 7 or 8 volts of that voltage drop, then you know that your MOSFETs are dissipating the bulk of the supplied power.  I suspect that this simple logic could have eluded the RATs.  Hence this might be another surprise for Team Rosie Posie - that the magic negative oscillation mode that generates an alleged "COP infinity" actually results in more battery power being dissipated in the MOSFETs than the load resistor.

MileHigh

P.S.:  The reason I have no "direct experience" is that I never had a logical reason to make some of these krazy-kooky measurements in the real world working on a real bench working on a real project.

[TinselKoala]

@MH...yep, I concur. They had no idea what was happening in their circuit.... but at some point they realised that the mosfets needed heatsinks.... it's just too bad that they didn't compare the heating of the mosfets with the heating of the load... properly.
In the oscillation mode the thing works like a disabled PWM controller; the mosfets are partially on, I guess. The thing makes a pretty good motor driver (no oscillations) when you just hook a small DC motor where the load goes. Control motor by input voltage applied to the 555. Very inefficient.

Well, I have six batteries now, and all IRFPG50s as the transistors (although I see that as a waste of money). I'm making waveforms and heating the loads, both the parallel ceramic resistors and the stock water heater element, with oscillations. I have phase relations. (Don't tell my neighbors, please.) I have the LEDs of Doom, Deluxe Edition. I have every damn thing _except_ perpetually charged-up batteries.

Could it be that the NERDs are deploying anti-replication technology against me? Yes.. that must be it, otherwise my exact duplicate of their circuit, except for the batteries, must work in COP INFINITY mode.

Or could there be another reason? Stay tuned for the next episode in the continuing saga of TarBaby vs. the NERDs from Bizarro Universe.

[fuzzytomcat]

Hi TK,

I was curious on the Clarke-Hess 2330 Power Analyzer that your using and found a PDF file ( 2330.pdf ) on it's operating specifications what a nice unit !!  The one thing that really surprised me was the 30 minute warm up time for full accuracy ... like a older analog scope or maybe even a little longer. 

Fuzzy

[TinselKoala]

Yeah, long warmup times are pretty standard in the test equipment field, even with today's digital equipment.
Some of that stuff like precision counters even has a little oven in it to keep its crystal oscillator at an even temperature as long as it's plugged in.