Hi Doug,

It was great to chat with you.
Here is a brief summary of the technical details we spoke (Doug, correct me if I cite something incorrectly).

Staying with the question how could someone show and demonstrate the amount of current the oscillator pumps into the circuit, there are a couple of simple possibilities: - measuring the current flowing in the wire between the oscillator output and the probes.  With this we need to be careful so that the measuring probe and instrument wont alter the setup's behavior too much. - measuring the change of DC supply current going into the oscillator between unloaded state (its output disconnected from the probes), versus when the oscillator output is connected to the probes.  Here we have a better chance to create a setup that only minimally alters the original scenario: we can use a miniature battery powered current meter.  We can also measure the DC supply current on the oscillator separately when we connect a tuned circuit to its output drawing approximately 40mA at specific harmonics.

When we do one or both of the above tests while we rearrange the probe cables to get maximum signal on the oscilloscope input, we can notice that the AC current supplied by the oscillator has two distinct states: there is a case when the oscillator supplies several times ten milliamperes, but there is also another state, when the oscillator hardly supplies any AC current.  This letter case can further be tested by inserting a very small capacitance between the oscillator output and the probe loops: we will still get about the same big signal on the oscilloscope input.  When I reproduced Doug's experiment at home, this latter case was accidentally the first I stumbled across.  This case corresponds to a parallel resonance instead of a series resonance, requiring only a very small amount of current feeding the circuit.  The Q of the tuned circuit will amplify the small injected current, so the current flowing in the ground lead wire of the oscilloscope probe is comparable to what we get in the series resonance case.  As we keep moving the probe cables, we can 'tune' the circuit through a set of series and parallel resonances.

Thanks again Doug for sharing your interesting experiments!

Istvan Novak

Istvan Novak wrote:
Great! Lets talk over the phone, will call you after work.

At the end though I think SI-list readers may also be interested in ways
we come up with to demonstrate further aspects of your intriguing


Istvan Novak

On 4/6/2018 1:49 AM, Douglas Smith wrote:
Actually the oscillator is producing about the same current it would into a 
short circuit, 40 mA) because it is driving a very low impedance of a series 
resonant circuit!
Istvan, let’s talk on the phone to save s lot of typing.
Doug Smith Sent from my iPhone IPhone: 408-858-4528 Office: 702-570-6108 Email: 
d...@dsmith.org Website: http://dsmith.org
On Thu, Apr 5, 2018 at 21:07, Istvan Novak <istvan.no...@verizon.net> wrote:

Thank you for sharing your interesting experiments. These always
trigger my curiosity to look a little further. In my basement lab I
came up with a small series of experiments to separate the possible
different coupling mechanisms from the oscillator to the oscilloscope
input. Though I did not intend to replicate your exact setup, I believe
it was conceptually and essentially similar enough that the results may
apply to a large number of setups, including yours. From all of the
different tests my conclusion is that the oscillator is not supplying a
significant amount of current to produce the effect in question. I have
not had the good fortune yet to see your full live demonstration, plus
as you say you have much more data than what you publish, so it very
well could be that you already have done similar experiments and came to
similar conclusions. Therefore I would like to hear the thoughts of
those list members who did not attend your demonstrations, what kind of
experiments would they suggest to prove or disprove my conclusion.


Istvan Novak

On 1/30/2018 12:06 PM, Douglas Smith wrote:
I did that experiment a long time ago, almost 30 years ago, when
developing this experiment, which is described in my book.

Reducing ground lead length to near zero eliminates the effect almost
completely with no other changes in the experimental setup. You still
see a little effect about 1/50 of before, due to the shield transfer
impedance of the probe cables. A tiny ground lead always swamps shield
transfer impedance of practical shielded cables. I do that experiment
for my classes as an extension of this experiment.

My live experiments are always more complete than the versions I
publish both to keep published versions reasonably short and to
provide extra value to live experiments. I usually have ten times the
data I actually publish!

Doug Smith
Sent from my iPhone
IPhone: 408-858-4528
Office: 702-570-6108
Email: d...@dsmith.org
Website: http://dsmith.org

On Tue, Jan 30, 2018 at 6:42, Istvan Novak <istvan.no...@verizon.net>


I see what you mean, but to me this does not seem to prove either
way whether the mechanism bringing the signal into the signal
paths is the inductance of the ground leads or is it happening
through finite surface transfer impedance. The convincing
argument/experiment would be to show that by eliminating the
suspect element, the signal pickup goes away.


Istvan Novak


On 1/29/2018 10:55 PM, Douglas Smith wrote:
Hi Istvan, If I connect the oscillator directly to the scope
chassis, I get similar readings at different frequencies. I
generally get less amplitude as well.

The small capacitance between the scope and oscillator is capable
of tuning the whole system to a low input impedance on the probe
end. I have seen this in many scenarios that ended up loading the
oscillator to a near short circuit! The whole system is resonant.
I only get enough current to get large probe voltages at discrete
frequencies. I had to tune around to get the results you saw.
These is a lot more information when I do the demo live for a
class that are not practical to put in a video (would last way
too long).

I think amateur radio operators can identify with the amazing
characteristics of resonant systems.

A loop is not nearly dangerous as a loop terminated in a small

Tesla coils resonate using stray capacitance. I built a 600 Watt
coherent 300 kHz RF driven Tesla Coil in the 9th grade, but that
is another story......


Doug Smith
Sent from my iPhone
IPhone: 408-858-4528
Office: 702-570-6108
Email: d...@dsmith.org
Website: http://dsmith.org

On Mon, Jan 29, 2018 at 22:24, Istvan Novak
<istvan.no...@verizon.net> wrote:


I see your point, but I think we would need a little more
argument for the amount of current you assume in this example.

I agree with the approximate calculation of the L*dI/dt
voltage drop,
where the ground-lead inductance is assumed to be 100nH and I
assume you
measured the rise/fall times of the HC240 to be around 2ns.
If we
assume 40mA current delta, we approximately get the voltages
you see on
the screen. I am just not convinced that the current is
really in the
order of 40mA. If I am not mistaken, you say (and the video
the same) that the oscillator's return (its 'ground') is left
In this case it matters much less what is the inductance of the
ground-lead loop (you say it is 100nH, not disputed), but for
current to
flow through the oscillator output, it has to find its way
back to its
return by going through the parallel of the two probes' braid
(we can call it common-mode impedance) and eventually it has
to close
back to the floating return of the oscillator. This last
portion of the
current loop, closing back from the oscilloscope
chassis/ground to the
oscillator return would need to be in the order of 50 pF or
higher in
capacitive coupling to not limit to current to lower values.
crudely from the video, the 'stray' capacitance of the
oscillator might
very well be much lower. However, here is another possible
to the waveforms on the oscilloscope screen: the oscillator
imposes a
common-mode voltage onto the scope probes, which creates
current in the
shields and these being passive probes, through the finite
transfer impedance of the cable braid, voltage is induced at
the probe
connections. To prove/disprove the existence and amount of this
potential contributor, you could redo the test by eliminating
ground-lead loop (for instance by using coax receptacles for
the probe
tips), leaving everything else the same. If the displayed
voltage drops
significantly, I would consider as a proof that in fact the
source of the waveform is related to the ground-lead
inductance. If the
displayed waveforms would not change considerably, it still
would not
prove that the finite surface transfer impedance is the man
cause, but
it would prove that there should be another major contributor
beyond the
ground-lead inductance. In this letter case further tests
could be
devised to nail down the major contributor.

What do you think?


Istvan Novak


On 1/29/2018 3:13 PM, Doug Smith wrote:
OK Everyone, here is the explanation:

The box is superfluous, it just hides what is actually
going on. The box contains two heavy gauge wires. One shorts
the two tips sig and gnd together and the other connects the
center point of the first wire to the center pin of the BNC
connector. This is equivalent to removing the box and putting
a stiff wire into the BNC center pin on the generator and
connecting both probe tips and both ground leads to this
stiff wire.
Doing so causes the generator to push a current out the
prob e ground leads creating a voltage across the inductance
of the leads that create a loop at the front of the probe.
The current flowing through the ground leads creates magnetic
fields that for the most part are captured by the loop formed
by the ground leads and probe delivering the induced voltage
to the probes.
The probes only respond to voltage between their tips and
ground lead attachment point, and the ground lead induced
voltage is delivered to each probe by loop they form.
Since the probes are lying apart from each other, one on
the plastic table with a loop in its cable, and the other
over a highly conducting (center layer) ESD mat, their common
mode impedance will be different, varying at different
frequencies. That results in different common mode currents
on the probes and the ground lead induced voltages are
therefore different and that is what is displayed on the
scope. It is interesting that the current output of an HC240
Octal Inverting Buffer can induce volts across the ground
leads when its output is only 5 V P_P, but entirely
understandable if you calculate e = Ldi/dt = 100 nH*.040
A/2ns) as an approximation.
For the generator to push a current onto the probe cables,
it must form an image current somewhere and that is
displacement current to the nearby scope chassis and some
radiation from the oscillator itself into the "ether,"
although it is on the small side to radiate itself
efficiently at 40 MHz.
I suspect the total current being pushed onto the probes is
about 40 mA, the short circuit current of the HC240 IC being
used. At some frequencies, each probe cable will resonate
with the capacitance back to the scope from the oscillator.
By changing the frequency, I can make either probe register a
larger signal than the other probe.
I love experiments like this. I have been doing this one
for over 20 years for my classes and have have tons more
demos to challenge engineering minds. Most of my demos have
an unexpected result and the discussion that follows
elucidates some engineering principle or addresses a myth.
University of Oxford, Course Tutor
Department for Continuing Education
Oxford, Oxfordshire, United Kingdom
Doug Smith
P.O. Box 60941
Boulder City, NV 89006-0941
TEL/FAX: 702-570-6108/570-6013
Mobile: 408-858-4528
Email: d...@dsmith.org
Web: http://www.dsmith.org

On Mon, 29 Jan 2018 14:40:17 -0800, "Tom Dagostino" wrote:

Why is everybody looking for tricks or complex answers.
This is really a
very simple, basic issue. If you look closely you will see
the difference.
Tom Dagostino

Teraspeed Labs
9999 SW Wilshire Street
Suite 102
Portland, OR 97225

-----Original Message-----
From: si-list-bou...@freelists.org
[mailto:si-list-bou...@freelists.org] On
Behalf Of Austin Mack
Sent: Monday, January 29, 2018 2:28 PM
To: d...@emcesd.com; 'si-list'
Subject: [SI-LIST] Re: ***UNCHECKED*** Measurement Dilema

Hi All,

It looks to me like the oscillator and its power pack (with
very long leads)
are in close proximity to the CH2 probe cable. Since the
oscillator output
is tied to GND at the probes, it and the floating power
pack will be
radiating like crazy at 40MHz and electromagnetically and
coupling to the CH2 shield which BTW is close to a quarter

-----Original Message-----
From: si-list-bou...@freelists.org
[mailto:si-list-bou...@freelists.org] On
Behalf Of Doug Smith
Sent: Friday, January 26, 2018 2:53 PM
To: si-list
Subject: [SI-LIST] ***UNCHECKED*** Measurement Dilema

Hi All,

Can you explain the result in this video I just made? Scope
plots of the
same two nodes are completely different. Probes and scope
are operating
normally, no problem with the equipment itself.

If you have been to my seminars you know the answer, please
do not post the
answer unless you have not seen this experiment until now.

Hint 1: There are no EM fields radiating from the shielded
box affecting the
Hint 2: There are no active components inside the box.


I do a lot of experiments in my classes that give
surprising results. Each
one addresses a design or troubleshooting problem that
engineers do not
realize their troubleshooting efforts. Next
one:http://emcesd.com/bcsem_hfmeas.htm on March 13-16.


This message is from the IEEE Product Safety Engineering Society emc-pstc discussion 
list. To post a message to the list, send your e-mail to <emc-p...@ieee.org>

All emc-pstc postings are archived and searchable on the web at:

Attachments are not permitted but the IEEE PSES Online Communities site at 
http://product-compliance.oc.ieee.org/ can be used for graphics (in well-used 
formats), large files, etc.

Website:  http://www.ieee-pses.org/
Instructions:  http://www.ieee-pses.org/list.html (including how to unsubscribe)
List rules: http://www.ieee-pses.org/listrules.html

For help, send mail to the list administrators:
Scott Douglas <sdoug...@ieee.org>
Mike Cantwell <mcantw...@ieee.org>

For policy questions, send mail to:
Jim Bacher:  <j.bac...@ieee.org>
David Heald: <dhe...@gmail.com>

Reply via email to