Auroratek demonstration from Bill Alek at TeslaTech conference

[TinselKoala]

MarkE, Every Tek digital scope screen that I have ever worked with indicated the channel coupling by a small symbol next to the channel vertical setting indications at the bottom of the trace display.

An AC-coupled channel displays a small sine wave symbol here, and a DC coupled channel either displays nothing as your screens show, or the dotted-over-dashed DC symbol.

I can't find any channel coupling symbol on your scopescreen shots, though. How does your scope indicate the channel coupling on its display? It would be very unusual, in my experience, to find a modern scope that did not, in some manner, always show the channel coupling on the main screen somewhere, since it is so easy to make mismeasurements when AC-coupled. The operator needs to have this information "in her face" so to speak.

For example, note the sine wave symbol next to the channel vertical setting for Ch1 on the DPO4034 screen below. The AC-coupled channel shows the sine symbol and the DC coupled channels show nothing in that position, as your scopeshots show.

(Finding good screenshots of AC-coupled TEK scopes is very difficult since AC coupling is so rarely used, I guess. So I have to present you with one of my own.)

Here's the Operator's Manual for Alek's Fluke 105B scopemeter. There are some interesting details involving the use of the coupling settings, particularly in Section 8. Everything they say in the manual supports what I have been saying all along.
http://assets.fluke.com/manuals/scopemt2umeng0000.pdf


In addition, your experiment, which shows ac vs dc coupled measurements that appear to be made on an input _with no DC offset_ is incomplete. Could you please repeat the measurements on a signal from your FG that includes a significant DC offset.

(By the way.... note the _decreasing_ energy integral in the Tek shot. This is the integral of the instantaneous VxI load power trace of the DUT.)

[TinselKoala]

TK, Who did say, that Bill Alek used the AC-coupling function on his scope ?
Is this verified ?

Surely one should only use DC coupling for the measurements, but are you shure he used AC coupling
during the demo ?

Bill posted the screenshot I took from Alek's demo showing the scope channel couplings. I put a link to the Fluke 105b scopemeter operating manual in the comment above, so that anyone can check to see for themselves that I have identified the right indicators on the scope's display. There are also references to the proper use of channel coupling in the manual. Alek's Fluke 105b Scopemeter is set to have both channels AC-coupled in that video demonstration. No doubt about it. Apparently if one is in "running auto" mode and selects a measurement that is "AC" such as "RMS AC" rather than "RMS" which normally does RMS AC+DC, the scope's channel coupling auto-switches to AC coupled. In short the manual supports what I have been saying all along: the AC coupling destroys DC offset information and also most other vertical measurements made on the signal, except p-p amplitude. In MarkE's scopeshots above, the fact that the AC coupled measurements coincide with the DC coupled measurements is just that: a coincidence, caused by the fact that he truly did not have any DC offset in the input signal from his FG. Had there been waveform distortion and/or a DC offset in the input signal, the differences between AC and DC coupled would have been very plain in his data.

[TinselKoala]

Here's another Tek scopeshot showing an AC coupled channel symbol, item #11.

[MarkE]

MarkE, Every Tek digital scope screen that I have ever worked with indicated the channel coupling by a small symbol next to the channel vertical setting indications at the bottom of the trace display.

An AC-coupled channel displays a small sine wave symbol here, and a DC coupled channel either displays nothing as your screens show, or the dotted-over-dashed DC symbol.

I can't find any channel coupling symbol on your scopescreen shots, though. How does your scope indicate the channel coupling on its display? It would be very unusual, in my experience, to find a modern scope that did not, in some manner, always show the channel coupling on the main screen somewhere, since it is so easy to make mismeasurements when AC-coupled. The operator needs to have this information "in her face" so to speak.

For example, note the sine wave symbol next to the channel vertical setting for Ch1 on the DPO4034 screen below. The AC-coupled channel shows the sine symbol and the DC coupled channels show nothing in that position, as your scopeshots show.

(Finding good screenshots of AC-coupled TEK scopes is very difficult since AC coupling is so rarely used, I guess. So I have to present you with one of my own.)

Here's the Operator's Manual for Alek's Fluke 105B scopemeter. There are some interesting details involving the use of the coupling settings, particularly in Section 8. Everything they say in the manual supports what I have been saying all along.
http://assets.fluke.com/manuals/scopemt2umeng0000.pdf


In addition, your experiment, which shows ac vs dc coupled measurements that appear to be made on an input _with no DC offset_ is incomplete. Could you please repeat the measurements on a signal from your FG that includes a significant DC offset.

(By the way.... note the _decreasing_ energy integral in the Tek shot. This is the integral of the instantaneous VxI load power trace of the DUT.)

This model does not show the little AC coupling icon.  I guess Tek figured that $2200. wasn't enough to include that firmware feature.  Here is a shot with same set-up at 1 Hz but the Channel 1 settings displayed.

I do not dispute that what Bill Alek did was poor practice.  I just don't think that there is any direct evidence at this point that AC coupling is contributing to the obvious, gross phase shift error in his measurements.  Bill Alek has much to do to establish any validity to his measurements.  One of the obvious steps would be to eliminate the possibility that AC coupling is distorting the phase relationships by switching to DC coupling.  There isn't any reason I can think of that he should not DC couple.  If for example he has an unstable supply feeding the current sensor and is using AC coupling to suppress an unstable baseline, then AC coupling is masking a bigger problem.

Below I added captures per your request showing the waveforms with a +/-100% pp offset.  I had to reduce the pp amplitude so that the waveforms would fit on the screen.  As you can see there is no perceptible phase shift that results from the offset, as the math trace remains flat.  Remember that any transformer is by its nature and AC coupling device.  Bill Alek is not going to be able to increase the secondary side power by introducing a DC bias to his primary drive.  If he has some DC offset, that is just going to be power lost as I²R losses in the primary.  There are no visible signs of saturation in his waveforms.

[TinselKoala]

MarkE, Great, thank you for doing that test. That confirms that phase data for sine waves isn't affected by the coupling setting even when there is a DC offset present, which makes sense for symmetrical waveforms like sine. I wonder if this is generally true though or if it is only true for sine or other vertically symmetrical waveforms.

Do you have a delay line or some other way to delay a signal from the FG by 3 microseconds? I think Alek's Honeywell Hall-effect based current monitor has an unavoidable 3 us delay built in due to the ADC conversion circuitry.

I agree, of course, that DC power on the input to a transformer won't affect the output as long as the core doesn't saturate. This is the case for plain transformers that we know about. However, we are being presented with a "new" type of transformer with allegedly unique characteristics. It remains an untested _assumption_ (albeit on very good grounds) that DC power on the input to this device will not affect the AC output, lack of core saturation notwithstanding. In addition, there is that bit of circuit board  material, g10 or FR4 glassfilled epoxy unless I miss my guess. This seems an interesting choice for a mere transformer support board and I would like to see it displayed in full, just so I can be confident that there aren't any strange wiring twists or capacitive couplings going on in that bit of costly structural support.

Once again I would like to stress that I am not accusing Alek of fakery by hiding DC current in the input. What I am saying is that his measurement technique is faulty and if he _DID_ want to hide some additional power, using AC coupling without calling attention to it would be one technique that I would think of instantly. Later on down the line in Alek's claimed development process he will have to be making measurements on "AC" signals with DC offsets: he is building battery chargers and self-charging scooters that do have batteries in them. Will he, or has he, used AC coupled channels while working on such systems, while quoting power data to investors and supporters? Skeptical minds want to know.