Claimed OU circuit of Rosemary Ainslie

[PaulLowrance]

Don't you think people will wonder why everyone here will not do this the correct way, by measuring how fast the components heat up?

I'm in this thread every once in a great while for one reason, to save legit researchers time & money. Given by how few people are replicating this, I'm glad legit people are finally getting it: Wait for sincerity before spending time & money replicating.

[poynt99]

Examine this one.

1) You have two long croc-type leads to your resistor that run to a connector. I would guess around 24" each. Solder those leads to your resistor mate and get rid of the croc ends. Also, those leads appear under-gauged and are too long. Use multi-strand #12 throughout.

2) You have from this connector another long run to your battery bank, I would guess about 36" to the floor. The "battery" voltage is being taken from some point in the middle of this long run as shown on the butt connector. Incorrect. To properly measure the voltage across the load your two probes need to be right on the resistor ends as I show in my setup. The high side is your "battery" voltage, the low side the Drain voltage (reduce those lengths). See my test plan. That is how the wiring and measurement points should be set up.

3) You have two different probe ground points; one on the butt connector from the battery, and one star point ground going to your shunt (and you say I have ground loops ). This is a no-no, as I have found and emphasized here. One good ground point, and this should be at the -'ve side of the shunt with a heavy multi-strand wire. That little white ground wire going to your shunt is not adequate as I have noted.

You want to know why your results are skewed, this is most likely it. Clean up your grounding as I have and as I've just suggested, then see if your results are the same.

There is more I could suggest, but that will keep you busy for a bit should you choose to clean up your setup. I will also be trying a few more things to clean mine up as well.

.99

[MileHigh]

Glen:

I don't agree with your comments on .99's setup.  It looks pretty good to me overall.  The one point that you make that I think has some merit is about the load resistor being directly over the circuit.  However, the magnetic field generated by the load resistor is synchronous with the operation of the circuit.  In that sense it is not like an unrelated magnetic filed that would induce an asynchronous disturbance into the circuit.  Since the varying magnetic field is synchronous with the operation of the circuit, whatever possible effects it could induce will seamlessly blend in with the operation of the circuit.  This is also the distinct possibility that any effects would be minimal one way or the other.  There are no big wire loops in .99's setup to "catch" the changing magnetic flux from the load resistor.  None the less, it would not hurt matters to displace the load resistor so it is not directly over the PCB.

.99 and Hoppy:

I am surprised the grounding improvements had much of an effect.  There are essentially no currents in the ground connections associated with the probes, so you would assume that the probe grounds can't induce any ripple effects in the ground voltage.  There are a few pretty sharp edges though starting with the 555 waveform on down.  Perhaps there are some localized very high harmonic frequency effects associated with the sharp edges that were reduced in amplitude with the improved grounding arrangement.

This is starting to make me think of my two microwaves courses, where the professor reminded us that all general circuit analysis is just a generalization of microwave theory where the frequencies are normally low enough so you could ignore the microwave transmission line effects.  I should never have taken Microwaves II, big mistake!  lol

Paul:

Don't you think people will wonder why everyone here will not do this the correct way, by measuring how fast the components heat up?

Just measuring the temperature of the load resistor at thermal equilibrium is essentially the same thing with the added benefit that it is much easier to do.  Measuring how fast the components heat up is just the thermal power integrated over time.  Most people are just measuring the thermal power.

MileHigh

[poynt99]

MH,

Indeed I have "directed" the resistor end away from the pcb with no effect. I've moved all the wires around also with no effect on the wave forms. The biggest effects due to proximity to objects is electrostatic.

Regarding the grounding, yes a huge difference when reading "in between the lines". From a bird's-eye-view, you may not see much difference, but zoom in to the spike and you see a world of difference.

It has nothing to do with the probe ground affecting the circuit grounds, it is the other way around. As you just mentioned, high frequency effects must be considered in this case. Where is ground really ground All these leads are TL's at these frequencies. They have their own inductance and capacitance to earth ground. Arbitrarily attaching your scope ground in the middle of a 36" piece of wire that is carrying pulsed currents, isn't exactly going to act as a good ground any more. Since the probe ground is the reference point the scope uses, this greatly affects the "voltage" measured between the probe tip and its own probe ground. This "voltage" can not be representative of the real voltage being measured, unless all the inductance and TL effects can be eliminated. Well, we can't eliminate the effects entirely, but we can minimize them by keeping all wires as short as possible (including probe ground leads) and utilizing good grounding practices, which I somewhat failed to do from the start, but have since cleaned up my act.

Glen has some work ahead of him as well in this regard if he wants to improve the integrity of his data.

.99

[PaulLowrance]

Paul:

Just measuring the temperature of the load resistor at thermal equilibrium is essentially the same thing with the added benefit that it is much easier to do.

Huge difference. One is accurate & easy to perform. The other is prone to great errors giving false readings.

Measuring how fast the components heat up is just the thermal power integrated over time.  Most people are just measuring the thermal power.

And that's what you want to measure, but so far I don't see anyone doing it.


Paul