Shop OBEX P1 Docs P2 Docs Learn Events
Scope Noise — Parallax Forums

Scope Noise

RDL2004RDL2004 Posts: 2,554
edited 2013-02-12 17:57 in General Discussion
Anybody have a clue what might be causing this?

attachment.php?attachmentid=99187&d=1360599756

That's with both the ground clip and the probe tip attached to the ground terminal of the square wave reference signal, so I guess that's some kind of common mode noise? The scope is set to the smallest divisions possible, but that still seems like a lot of noise. If I disconnect the probe from the scope it goes away, so I don't think it's any kind of internal problem. The signal kind of dances around a lot but the frequency is right around 95-96MHz. I had an old toroid core about 1/2" cross section, 2" diameter and with 5 or 6 turns of the probe cable wrapped around it the peak-peak dropped to about 12-15 mV. That's the best I've been able to do. Does this seem normal? I am far from an expert when it comes to scopes.
ScopeNoise.jpg
640 x 468 - 76K

Comments

  • LawsonLawson Posts: 870
    edited 2013-02-11 12:30
    does the noise show up when the probe ground lead is clipped to the input lead? Does the noise change if you wave the probe around? (still with the ground lead hooked to the input) These tests will help narrow down the noise source.

    Anyway, my guess is that the noise is from a noisy clock inside the oscilloscope. If the noise disappears with the ground lead clipped to the input, I wouldn't worry about the noise. (btw, about 1-2 mV of random noise is typical for the oscilloscopes I use)

    Lawson
  • evanhevanh Posts: 15,949
    edited 2013-02-11 13:53
    Yeah, what does happen with the probe still attached but not hooked to the test ground pin? 96MHz is pretty extreme for an external source.
  • mojorizingmojorizing Posts: 249
    edited 2013-02-11 14:13
    I have a rigol DS1102E and with the same setup I get 97mhz +/- signal at 780uvolts v-max. Probably noise coming from the scope.
  • RDL2004RDL2004 Posts: 2,554
    edited 2013-02-11 14:32
    Well, plugged in but not clipped to anything, with the probe just lying on the bench top, this is what I see:

    attachment.php?attachmentid=99195&d=1360621348

    Not too bad, but as soon as I connect the probe to just about anything I get that noise in the 25-50 mV range. It's not bad if what you're looking at is at a fairly high voltage, but I'm trying to measure noise on a power supply circuit and this is just not working. You don't even want to know what it looks like plugged into a plastic breadboard, because that's really ugly.

    I've tried several different probes, and I've turned off all the electronic stuff within about 10 feet of the scope but it doesn't seem to make any difference. I may have to haul out my old Tektronix analog scope to see what I get with it.

    edit: yeah, this is a DS1052E. The 1102 goes down to microvolts? 20 millivolts is as low as this one goes.
    ProbeBench.jpg
    640 x 468 - 60K
  • pmrobertpmrobert Posts: 675
    edited 2013-02-11 15:04
    Do you have a commercial FM radio station nearby at, say, 97mHz? A few years ago I had a zero crossing IC (LM1815) circuit on a breadboard that worked just fine if I wasn't at my home garage. It turned out that an AM radio station antenna is on a golf course about 1/2 mile away from my home and you could clearly see the noise on the VR sensor input. Shielded it, aka placed it in sturdy metal project box, made sure common grounds were simple and robust, and all is well. Figured this out because everything started working magically well immediately after a couple of hurricanes took that station off the air for a few weeks. Station back on, random triggers returned to the breadboarded circuit. Serendipity is nice when it happens...
  • ElectricAyeElectricAye Posts: 4,561
    edited 2013-02-12 06:29
    I don't know much about radio but it seems to me that this could be leakage from the local oscillator of a radio receiver you own. Or from the internal clock of a digital device you have nearby that isn't shielded very well?

    I wonder if you might be able to detect it with an FM radio tuned to ~97 MHz?
  • frank freedmanfrank freedman Posts: 1,983
    edited 2013-02-12 13:55
    Does this happen with other probes or only this one? Is there anything near the probe that could be generating this signal? My first thought would be ground loop somewhere which makes me think probe may have a problem. Does this scope have an internal low Z setting or can you put say a 50ohm terminator and see if the noise is still present. Not familiar with this scope so can't speak to shielding internally, or if a DSO the problem is aliasing causing the signal to be reconstructed in the manner shown.

    FF

    RDL2004 wrote: »
    Anybody have a clue what might be causing this?

    attachment.php?attachmentid=99187&d=1360599756

    That's with both the ground clip and the probe tip attached to the ground terminal of the square wave reference signal, so I guess that's some kind of common mode noise? The scope is set to the smallest divisions possible, but that still seems like a lot of noise. If I disconnect the probe from the scope it goes away, so I don't think it's any kind of internal problem. The signal kind of dances around a lot but the frequency is right around 95-96MHz. I had an old toroid core about 1/2" cross section, 2" diameter and with 5 or 6 turns of the probe cable wrapped around it the peak-peak dropped to about 12-15 mV. That's the best I've been able to do. Does this seem normal? I am far from an expert when it comes to scopes.
  • RDL2004RDL2004 Posts: 2,554
    edited 2013-02-12 15:17
    Well, I'm really more interested in working on the power supply project than tracking down weird noise sources, so I've given up on this for now. I'm almost certain it's some kind of RF problem. If I clip the ground on the probe to the probe tip, then the closer I move it to a grounded metal object (such as the reflectors of the two fluorescent lamps over my bench-turned off of course) the stronger the noise signal. Moving it around all sides of the scope itself doesn't cause it to change much.

    I have to point out that I only see this noise when the probe tip is connected to something. The big problem is that the power supply circuit I have built on a plastic breadboard, which isn't ground referenced, also picks up this noise and it's even worse there. Anyway, I was reading in the scope's manual and found out it has some digital filtering capability that I can turn on to hide this noise for now.

    I have indeed tried several different probes, five to be exact, and both channels on the scope. All see pretty much the same signal. I also turned off all nearby electronic devices with the exception of the cable modem, wireless router and my little HP Microserver which are about 12 feet away. At some point I'll probably have to drag out the Tektronix scope to see what I can find with that. Thanks for all the ideas and suggestions :)
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2013-02-12 15:23
    There are two FM stations in Mississippi broadcasting on 97.1 MHz, one in Meridian, the other in Glen. Are you near either of those locations?

    -Phil
  • RDL2004RDL2004 Posts: 2,554
    edited 2013-02-12 15:45
    No, those are pretty far away. There are a few "local" stations in that MHz range, but I think they're all near Jackson, which is about 15 miles away.

    I increased the time/div and noticed this:

    attachment.php?attachmentid=99215&d=1360712464

    Looks like some kind of AM at around 1.25 MHz (ignore the scope saying 50 MHz, it wasn't locking on)
    ProbeNoise2.jpg
    640 x 468 - 71K
  • jmgjmg Posts: 15,173
    edited 2013-02-12 17:01
    RDL2004 wrote: »
    I have to point out that I only see this noise when the probe tip is connected to something. The big problem is that the power supply circuit I have built on a plastic breadboard, which isn't ground referenced, also picks up this noise and it's even worse there.

    If you are on good terms with a local PCB shop, get a sheet of Copper PCB laminate, and place that on your workbench, (fibre-glass side up ;) )
    - we found that very good for quieting things down, when facing RF breakthrough...
    Another approach is to make your own LPF-Probe, with well bonded coax, and a cap right at the end.

    If you have feed thru caps, and only want to check PSU noise, then a 10:1 probe with 10K will likely be ok,
    1k//1n feed thru is ~ 1MHz or a 1K:100ohm//1n is ~10MHz - and a ferrite bead will roll off more at 96MHz.
  • RDL2004RDL2004 Posts: 2,554
    edited 2013-02-12 17:47
    Hmm, I don't know if there's anywhere around here that sells copper in sheets. My workbench is covered with that blue anti-static conductive rubber, I wonder if a couple of layers of heavy duty aluminum foil under that would work as well?

    attachment.php?attachmentid=85404&d=1317070438

    That's from when I first put it down, it's no longer as uncluttered :)
  • LawsonLawson Posts: 870
    edited 2013-02-12 17:57
    RDL2004 wrote: »
    Hmm, I don't know if there's anywhere around here that sells copper in sheets. My workbench is covered with that blue anti-static conductive rubber, I wonder if a couple of layers of heavy duty aluminum foil under that would work as well?

    That's from when I first put it down, it's no longer as uncluttered :)

    Yes the Aluminum foil under that blue mat should work just as well.

    Is your PC case still open? A PC mother board can put out a considerably amount of RF and there is usually at least one clock running at 100MHz.

    Lawson
Sign In or Register to comment.