EEVblog #21 – The Unusual Oscilloscope Phenomenon(3)Posted on July 28th, 2009 29 comments
Dave dons the lab coat for yet another installment of the Unusual Oscilloscope Phenomenon.
It all looks good. It appears you have found a good way to share lab experience on the web. The important thing is everyone is seeing what one is up against when trying to figure out what is going on. I guess that is part of the fun.
I just want to speculate a little bit. No, wait, not a little bit, a lot
You mentioned all sorts of things the effect does *not* or not entirely depend on (probes, termination, o’scope type, etc.).
What about the chair? You didn’t mention that at all.
You have a gas lift cylinder in the chair, don’t you? Could it be that the friction between the various parts of the cylinder create the signal?
And regarding the 140 MHz. These cylinders are manufactured according to national and international safety standards. The cylinder dampens the move of the chair when you stand up. Could it be, that the dampening is such that most mass-produced cylinders end up with a moving speed good for 140 MHz? Some kind of the cylinder’s resonance frequency, plus some bouncing when the piston hits its end position?
If you are still inclined to do experiments, what about shortening both parts of the cylinder?
Nice lateral thinking, but it’s not the chair mechanism, sorry, I can get the same thing just walking past the probe, or brushing my arm across the chair material etc. I just find that standing up from the chair is the most consistent way to produce the broadband static impulse required, presumably because of the larger surface area of my butt(!). I’ve tried other methods to generate the static but they just aren’t as consistent, but I still get the exact same ~120MHz sinX/x impulse. So the chair is a complete read herring, always has been.
The important thing is not so much that the probe picks up static, but rather that it rings. So your next blog, continuing from what you’ve just presented, should be on how fast edge ringing may not be down to the circuit, but the probe. You could show this by attaching a standard probe and a 50 ohm terminated probe to a fast rising edge pulse generator and showing the difference.
To show that the ringing is indeed from the probe, you could also change the frequency of the ringing by changing the length of the cable to change its inductance, or increase its capacitance by putting a capacitor across the input of the scope and across the trimming capacitor of the probe.
The edge ringing is a function of the filter on the oscilloscope input circuit. As such, it’s pretty much impossible to get rid of.
Have you tried this with an older type O’scope that has a fully shielded cabinet? Could it be that the plastic cabinets on these newer digital scopes is insufficient to block the ESD waveform? The frequency of the pulse could be a parasitic resonance of some part of the internal circuitry or groundplane.
No, I haven’t tried an older style scope, the problem is getting one that has the required single shot bandwidth.
If it is internal parasitic circuitry/groundplane effect then it’s hard to explain the fairly consistent result among all different types and brands of oscilloscope.
One thing is for sure, something is ringing somewhere, and the only really verifiably consistent thing among all these scopes I can think of is the input capacitance.
It would be interesting to obtain a medium power VHF signal generator and antenna and sweep the frequency band 100 to 200 MHz to see what responses occur. An O’scope is a receiver of sorts, and anyone who works around radios has witnessed that even a receiver without an antenna will pick up plenty of signals nearby.
I’m intrigued by the consistent frequency of this phenomenon. At 140 MHz, it puts a wavelength at about 2 meters … which I will guess is about your height. I’d also speculate that, as you rise from the chair or otherwise build up a static charge on your body, a spark occurs somewhere (perhaps through the soles of your shoes) and this impulse causes your body to become an efficient (resonant) radiator at that frequency. The amplitude of the pickup obviously changes with the efficiency of the receiving antenna (probe). I further speculate that it doesn’t disappear when you short or terminate the front-panel connector because the scope is not completely shielded. Full shielding is nearly impossible because of the large screen opening, even on older CRT-based, metal cabinet ‘scopes. Internal wiring then becomes the residual antenna.
Bill Whitlock, president & chief engineer
Jensen Transformers, Inc.
Chatsworth, California, USA
Very interesting thought Bill. Not sure how feasible it is, but I like the concept! (I’m 1.75m BTW)
The spark is almost certainly occurring between my butt and the chair. The quicker I stand up the greater the voltage produced. But the same thing happens when I say rub my jumper on the chair seat, or simply walk past in some cases. So the actual source of the static charge is not of any consequence (apart from amplitude).
Each scope seems to have it’s own particular frequency though, but once again, all I’ve tested have been very well bunched around that 100-140MHz range.
The most obvious consistent parameter in this respect across all scopes is the input capacitance, so I can’t help but think this is one of the main contributors.
Well, it looks to me like this is similar to the way a Theremin produces its music. I don’t know a lot about these scopes, tho I do quite like all those buttons and knobs.
i will leave you with this entertaining video from The University of Nottingham for more of the “how it works” type of thing.
Very interesting phenomenon. As an aside about the chair (and apologies if this repeats someone else’s comments; I’m new here!): it almost certainly is a huge source of ESD, I think due to the gas cylinder. At a previous job we had so-called “ESD chairs” that were grounded, but if you were wearing an ESD wrist strap with a meter, it would still go off the charts if you stood up from the chair. So if you’re killing ESD-sensitive parts in your lab, tell people not to handle the parts when they’re standing up or sitting down and see if there’s a change.
It doesn’t explain anything about the coupling of the signal into the scopes, but I do think that your idea about input capacitance is something to look into. That will help you see how it’s coupling in, but getting rid of it…
I use a Tektronix DSA70804 8 Ghz scope at work with 25 GS/s and I have seen this phenomenon not only standing up but sitting down. I work in a long narrow room and to get past the equipment that we are developing is a bit of a squeeze. Even that passing with no contact to the equipment or the scope can be detected from time to time.
Thanks for the Vblog. I’m a technician but I really enjoy the stories from the bench. You’ve given me a lot of ideas to think about. Keep up the good work.
Excellent work sir
Try wrapping the scope’s mains lead a few times through a big ferrite ring, at the scope end, and see if that helps. (The mains lead is about 2 meters long.)
My wild guess for the consistent frequency is the average scope probe length. It also fits with the frequency of the direct alufoil short being much higher.
I think the loop helps in getting the initial energy pulse, but the ringing occurs along the coax.
I found the same phenomenon using a TDS1002B today. however i could not get it to work by standing up from a chair. to cause the result i picked up my fleece coat from a table about 5 feet away from the scope.
as you can see my signal frequency is around 60Mhz and i forgot to compensate the 10X probe so it looks like the amplitude is roughly the same
Dave, I think your analysis is spot on.
No doubt the input circuit parameters of the scope factored along with the scope probe cable can sense the static discharge caused when you rise from the chair. I believe the length of the probe cable makes a pretty good QUARTER WAVE ANTENNA. Normally that length antenna would resonate at frequencies near half what you observe there. But the feedpoint of that unintended antenna is poorly defined and it’s 2nd harmonic could be what is being excited.
So, when you substituted the plain coax cable for the scope probe, what was it’s length? If it’s length is the same, would the transient frequency go lower with a longer coax cable?
Please, don’t beat me with Ned Kelly’s holster.
howdy there, just wished to say thanks alot for this article, it assisted me spot one thing I hadn’t given much thought to it before.
You seem to be assuming this is a static discharge phenomenon, despite evidence to the contrary (anti-static coat and wrist-strap). Could it possibly be created by the electrical impulse sent to or created by your muscles while standing? You’ve mentioned that it happens while walking past as well. One potential test would be to remove all static generating clothing, eliminating the chair, and attempt to trigger the effect with muscle movement only. I’ve learned that sharks use a similar phenomena, to sense prey, and I like the suggestion that your height is the basis of the consistent 140Mhz frequency.
P.S. I just realized I’m posting this almost a year after this video was posted. I wonder if this problem has been solved already.
Interesting. Australia is dry (even when it’s
wet). So static discharge is easy to create
and could be any part of your body or
clothing, say rubbing on the chair, or
I can’t get the phenomena with my
DS1052E in Maine, USA, where the humidity
is higher and static discharge is
That discharge is very broad spectrum, like
a spark gap transmitter.
The wire connected to the scope is coax
(standard or high impedance like on a probe).
The issue here is that it’s not behaving like
coax. There are currents flowing along
the shield and conductor of the coax in
a non normal mode.
Put ferrites around the wire next to the
scope and see if the impulse stops. They
will force the normal mode of the coax.
So the wire plus scope form a circuit that
responds to the impulse by ringing. The
resonant frequency is determined sort of
as you describe. This can be checked.
(a) Use radically different cable lengths.
(b) Use just a wire stuck in the inner
conductor hole of the BNC.
(c) Try and open or short on the
scope BNC (no wire). If the
frequency stays the same, then maybe
the wire is just an antenna feeding
energy to the input circuit and that
is what is ringing.
I find it interesting that the impulse
takes time to build up then decay.
I would expect very rapid build up
and slower decay.
So if you want to experiment further,
you need to find where the resonant circuit
is located. Is it the (1) cable/input?
(2) Is it just in the input circuit of
Nice demonstration of impulse response of
a resonant circuit. Now where’s the
Is it electrostatic discharge or just plain electrostatics? That doesn’t seem to have been addressed, yet.
Think of your body as one plate in a capacitor and the probe lead as the other. When you change the spacing between plates, you change the voltage on the capacitor. Thus, you have a voltage/E-field signal right there, without a discharge. It may have a somewhat high dV/dt. If in addition, you have a discharge than you have a current and an M-field signal as well, that may also be a more dramatic impulse (very high dI/dt and induced dV/dt) It could well be a spark, but the initial E-field issue still shouldn’t be overlooked because it may be what cases the M-field. Initial “bias” on the capacitor might affect the results. There is also the capacitor of your butt and the chair and numerous other stray capactances which change when you move.
I think you went a little overboard in dismissing the ground lead loop theory; indeed, by “disproving” it, in a way you proved it. IIRC, the voltage was about ten times less (or more) without it. I.E. that was the dominant effect. Once it was removed, however, there was still a residual effect. Thus it is the ground lead loop but it isn’t ONLY the ground lead loop. You exploited the ground lead loop to magnify the phenomenon, one shouldn’t ignore why it magnifies it. It is a good antenna, your probe cable is a poor antenna.
You missed a chance at a good rant over those %@#$%@$#% ground leads on scope probes. They cause all sorts of problems. There is a reason all the RF guys immediately suspected the ground leap. If you want to measure cleanly, you should use an adapter that connects directly to the concentric ground ring or a differential probe.
Bringing up this measurement noise source was a good thing. It is a reminder of how noise can affect your measurements and how the observer can interact with the experiment. About ESD. About impulse response.
Lots of further experiments are possible. I will have to play with this a bit when my 1052E arrives (my Tektronix 2236 isn’t as good for this). Of course, it is pretty damp (muggy) here.
I found an answer on Douglas C. Smith website:
I know that this is an older blog post, but it is so intriguing ! I wonder a lot about adequate shielding, and also about the impulse qualities of the signal n
I have not tried this yet, but Dave, you should take your scope down to the local high school physics lab and try using their van de graaff generator to cause this same phenomenon constantly, so you could catch a continuous waveform and then try foil or screen shielding techniques to see if it is a shielding issue.
A simple isolation transformer and a large toroid on the input lead from ac Mains should assuage any issues with signal getting in that port, and likely they already have a filtered iec socket internally as well, those are commonly available.
I am going to have to try this with my tek dso at work and with my 500mhz analog scope at home (if I can find a vdg to use) this is great blog material, very thought provoking!
Do you have some special brand cathode ray Lcd’s in your scope’s down in Austraila?
I became just searching for this information for some time. After six hours of continuous Googleing, at last I got it within your web site. I wonder what’s having no Google strategy that do not rank this type of informative web sites in the top of the list. Usually the very best web sites are rich in garbage.
This reminds me of when I first played with an o-scope so many, many years ago. I could hold the probe up over my head and pick up the 60 cycle hum. I played with touching the tip making the signal stronger, touching ground with my other hand and making it lower (but not go away). It was quite a bit more sensitive than the VOM and even VTVM’s of the day. And if you didn’t take care to ground everything together the hum might make measurements impossible, and even if you did some hum would still come through.
It’s just another reminder that what your test instruments are showing you is not necessarily what you are trying to measure. You have to consider every possibility when something doesn’t work out the way you expect. (Like in your cheap Chinese DVM show causing a complete rework–ouch!)
I’m loving your blogs. I just started at the beginning a couple days ago. I look forward to catching up with the current episodes. Thank you!
I think this really is expected behavior since all metal will act like both a receiving and transmitting antenna. It is a matter of the signal power, frequencies, and filtering/attenuation.
In our senior design lab, we were measuring audio signals from a low power electret microphone (30 to 3300 Hz) over some long cables. One member managed to borrow a 2 Gsps scope. There were some extra high frequency signals, so we decided to use the scope’s FFT function. Strangely, we were picking up signals in the 100MHz range! The phenomenon could not be replicated in other buildings. Whomever design the building and the university FM radio station were not aware of these kinds of problems! One of the antennas was pointing towards our building, and we were picking up an attenuated version of the high power signal over our long cables. Some coaxial cables, a second low pass filter after the cable and shortening the cables fixed that issue.
High voltage is required for static electricity discharges. Since it is an impulse, the frequency is very high. So even if the current is low, the power is still relatively high since there is more power in a high frequency signal and lower frequency. It really means the cables are not shielded enough.
Leave a reply