measure E-field

Hi all, I've got a customer who is seeing a lot of fluorescent light interference/ pickup. I'd like a way for us to compare how bad things are in our labs. (Maybe it's really bad where he is, or it could be a grounding shielding issue.) So I took my scope and stuck 6 inches of 18 AWG bus wire into the input.

formatting link
I was getting some pickup from the 'scope display hence I added a bit of "shielding" But I can still peak in around the side and see things. I found that I got ~50% more signal with the 'scope on its back. (I see about 300 mV p-p on the lab bench.) Now in my lab the lights "go" at 25 kHz. In his lab the frequency is closer to 45 kHz. How will this antenna compare at these different frequencies. (A few pf of wire in series with the 'scopes 20pF and 1Meg ohm?) Any other ideas on how to measure local E fields?

George Herold snipped-for-privacy@gmail.com (Google now won't post anything unless it has an email to warn me about.)

Reply to
George Herold
Loading thread data ...

Why concentrate only on the E Field? I know, conducted EMI back up the mains plays havoc with a lot of electronics. But, the magnetic field coming out of those things is incredible too.

Reply to
RobertMacy

tiny MOSFET with open gate?

I found approx 15fF between the lab wiring and the electronics if that helps model things a bit. Depends a great deal on type of wiring, too.

Reply to
RobertMacy

Something like an aluminum pie plate and a 10:1 scope probe would be good. A little math takes care of the probe loading.

I recently needed to estimate the e-field at the surface of a quartz window into a process chamber. We held a dime against the center of the window with a scope probe.

I've seen ghastly noise from fluorescent lights with electronic ballasts, in FTMS systems where stuff like 40 KHz is where the signals are.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Turn off the lights and retest in the dark. You might be surprised to find that the interference is coming from other sources. I was fooled once by a big DC motor with arcing brushes.

You seem to be reinventing the wheel. You're also not going to see much with an oscilloscope. You need a much more sensitive spectrum analyzer. Done correctly, you should see something like these:

Search for Mil Std-461 and FCC part 15 for conducted and radiated emissions test procedures and specifications. There's also standard LISN (line impedance stabilization network) test fixtures that allows you to connect your test equipment to the AC power lines without precipitating an insurance claim: If there's fluorscent lamp noise around the lab, it's probably being either radiated by the power lines or conducted by the power lines. The EMI/RFI will be much stronger and easier to measure with the test equipment connected to the power lines, than trying to pick them up with an antenna. However, if you must do it over the air, there are various FCC approved and expensive test antennas available which produce calibrated and reproduceable numbers. For example: This should offer a clue on how it works:

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

10:1 ?? so instead of 20pF you load with 2pF but reduce the signal 10:1, to measure a source that has capacitive impedance, where's the advantage?
Reply to
RobertMacy

A 10:1 reduction in the highpass corner frequency. There's usually gobs of signal in a typical room.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

more slowly, the 10:1 probe 'adds' a bit of C to the input base making the probe have a 'higher' capacitance to gnd than the original input of 20pF thereby, for capacitive impedance sources, like trying to pick up E Fields, you actually LOWER the sensitivity using the 10:1 probe, Does NOT improve sensitivity.

You can verify this, by comparing the two readings. You 'should' find the sensitivity is higher with the 1:1 input, and find it is better to not use that 10:1 probe.

Reply to
RobertMacy

The reason to use a pie plate is to get a lot of capacitance to the world, where the signal is. And to get a lot of signal.

As I noted, sensitivity isn't an issue when probing typical local e-fields. A pie plate will deliver volts of local crud, sometimes tens of volts, into a 10:1 probe. And you may as well ignore capacitive probe loading, since the c-ratio will be high.

And you get to eat the pie first. I make a dynamite peach and blueberry, and a pretty good chicken pot pie. Mo just made a St Patty's Guniness cake, which is about as good as food gets.

formatting link

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Thanks Robert, but I'd like something that any idiot can copy in their lab. George H.

Reply to
George Herold

. A

ow into

scope

Hmm OK, My question is *very* basic. Should I model this as the attenna ca pacitance in series with the 'scopes 20pF//1 Meg ohm? So that above the ~10kHz corner (of 20pF and 1 Meg) The response if fairly flat.

Though not a pie plate, I tried a 8" X 8" piece of copper clad. About 1.0V p-p directly into 'scope and 130mV p-p with the x10 scope probe. (16 pf a nd 10M ohm) The nice thing about the x10 probe (besides the lower bandwidt h) is that I can move it away from the interfering noise from the 'scope.) With the x10 probe I can also see ~300mv p-p of 60 Hz. OK off to push some numbers around.

George H. snipped-for-privacy@gmail.com

in

Reply to
George Herold

You have to know the capacitance of the pie plate to ground, to figure out the effects of the probe loading.

Connect it all up and add a zot of charge to the plate, and look at the step response on the scope. Charge up a small piece of metal, or maybe a wad of aluminum foil, and toss it into the pie plate. Knowing the scope probe R and C, that is enough data to calculate the pie plate capacitance.

Around here, we have a lot of nerf guns, which might accomplish the same function.

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

ood. A

indow into

h a scope

capacitance in series with the 'scopes 20pF//1 Meg ohm?

ly flat.

.0V p-p directly into 'scope and 130mV p-p with the x10 scope probe. (16 p f and 10M ohm) The nice thing about the x10 probe (besides the lower bandw idth) is that I can move it away from the interfering noise from the 'scope .)

Hi John, thanks for the 'hints'. So I should be able to estimate the capacitance. As a physicist I'm almost required to treat the pie plate as a sphere. Then the C of a sphere is 4*pi*epsilon_0*R (in MKS) for my 8" square I'll a ssume R is like 4 inches, 10 cm And that gives something near 10pF. So I've got two x10 probes. An old TEK listed as 8.5 pF, and a Rigol 17pf (+/- 5pf) I measured 200mV p-p with the 8.5 pf, 145 mV with the 17pF and 108 mV with both attached. Cranking through the numbers that works out to be ~11 .5 pF... hey not too bad. George H.

(Sorry for the late reply... )

Reply to
George Herold

Well, yeah, sometimes math is useful.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

You want something "dirty"? Use any 120VAC LED lamp...

Reply to
Robert Baer

On a sunny day (Thu, 20 Mar 2014 01:56:18 -0800) it happened Robert Baer wrote in :

Basically anything that has a diode rectifier in it. In the old days the diode bridges has small caps over the diodes, to decrease RFI. Not sure they still bother these days, Those were called 'rattle caps' here locally (translated).

Reply to
Jan Panteltje

The point to the capacitors was to prevent the carrier from a local LF/MF AM broadcast from being modulated with the rectified line frequency and sending as buzz to outside.

Is there still an AM trnasmitter in Hilversum?

--

Tauno Voipio
Reply to
Tauno Voipio

On a sunny day (Thu, 20 Mar 2014 14:16:31 +0200) it happened Tauno Voipio wrote in :

The transmitters were in Lopik, there was a Radio Wereld Omroep shortwave AM transmitter close to Hilversum?, been there, been in Lopik too. Hilversum is where the program material originates (Omroep centrum). If I am righ the AM transmission stopped altogether from Lopik, this is the current situation:

formatting link
Have not foloowed tha trado / tv business for years, 'wereldomroep' got less money probably closed some stuff.

Internet, satellite, cable. mp3players, but there are still FM radio stations.

Reply to
Jan Panteltje

Back when FCC and VDE rules were just coming into play and manufacturers started measuring the conducted EMI coming out of those old 100 W multi-output LINEAR supplies; they often found emanating noise at horrific levels. [You remember the old Telex +/- 12 Vdc and 5 Vdc?] They were VERY surprised to find that those linear supplies often put out MORE noise than the 'new' switchers they were comparing the linear supplies to. Many of those linear supplies actually FAILED compliance. Why? seems the old rectifiers shut off rather abruptly and produced AC mains related transients. That combined with the designer knowing/assuming the supply was linear did NOT add any AC line filtering, thinking just not necessary since the supply was linear and could not make noise. Solution? users [and some PS manufacturers] started putting little caps around the diodes to 'soften' that shut off and/or adding simple EMI line filters. I think the semi companies responded by changing their diodes, making it so the shutoff became 'softer', not so abrupt, and the problem went away. Plus, designers quit running 'looping' wires all over the insides translation: inductance, coupling, etc. and just became more aware of EMC and its implications.

Not sure true about semi-manufacturers, but it is more difficult to find newly manufactured rectifiers to have these same type shutoff symptoms.

Reply to
RobertMacy

One explanation that I have seen for these capacitors in LW/MW/SW domestic receivers, is that people usually connected some random wire into the antenna connector but do not have any good place to connect the ground connector. Typically the cold end of the input coil is connected to chassis ground, but this is not a good counterpoise for the unbalanced antenna.

I receiver connected to a grounded outlet will get "real" ground through the PE connector. However, Class 0 or Class 2 connector does not have a direct connection to the device chassis.

The mains transformer has a significant capacitance between primary and secondary, so it helps to have some "real" ground connection through the neutral wire. This works well with two diode center tapped rectifier, in which the secondary CT is connected to the chassis.

However, in a four diode bridge configuration, the situation is a bit different. When the diodes conducts, there is a low impedance connection between the Neutral and chassis ground (or the bypassed B+ line). However, when the diodes does not conduct, there is no low impedance connection to Neutral and hence "real" ground, only the diode capacitance. Thus, the ground connection from chassis ground to real ground is chopped 100 or 120 times a second, causing problems to the reception.

Putting capacitors across the diodes, will significantly reduce the grounding impedance (at RF) when no diodes connect, thus significantly reducing the grounding impedance variation.

Reply to
upsidedown

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.