Author Topic: 90 MHz noise on scope? (Read 10350 times)

cellularmitosis · « **on:** January 16, 2014, 08:13:02 am »

Hey guys,

While working on a little power supply project I kept noticing noise up around 90 MHz and I couldn't get it to go away. This is on a rigol 100 MHz scope.

Eventually I realized it wasn't my project, even just probing a piece of bare copper clad produced the noise.

What's strange is that when you zoom out it doesn't look random at all, it looks more like an MP3.

I'm using a spring tip ground, so I think this is as good as I can get my measurement setup.

Also, simply grabbing the cable or moving it around can cause 2x variations in the noise amplitude.

I also see a similar situation on my 40Mhz analog scope, except that it looks like just a solid band of noise blur, not a sinus. I even hooked up a BNC splitter to have the same probe driving both the CRO and DSO and I saw the similar fluctuations in amplitude from moving the cable around, so I don't think it's a poor BNC connector.

Any ideas?

Sorry about the photos, I just got my scope and haven't learned how to take screencaps yet.

AndyC_772 · « **Reply #1 on:** January 16, 2014, 09:21:30 am »

Sounds like you're picking up FM radio. Do you live near a transmitter?

MatCat · « **Reply #2 on:** January 16, 2014, 09:36:21 am »

Quote from: AndyC_772 on January 16, 2014, 09:21:30 am

Sounds like you're picking up FM radio. Do you live near a transmitter?

The idea came to my mind too when I saw the frequency but that's not FM modulation it's AM.

KJDS · « **Reply #3 on:** January 16, 2014, 10:00:05 am »

It's possible that the AM effect you're seeing is a sampling issue of an AM broadcast. It's also possible that it's an AM signal that's only appearing to be 90MHz because of a sampling effect.

rdl · « **Reply #4 on:** January 16, 2014, 02:37:43 pm »

Interesting. I have seen damn near the same exact problem. I never did figure out what was causing it. Best guess seemed to be that the scope was picking up a local FM transmitter of some kind.

http://forums.parallax.com/showthread.php/145932-Scope-Noise

cellularmitosis · « **Reply #5 on:** January 16, 2014, 03:19:20 pm »

So... anyone have any cheap ideas for building a desktop faraday cage? Something like a sandblasting cabinet with stainless mesh gloves or something?

cellularmitosis · « **Reply #6 on:** January 17, 2014, 05:30:08 am »

So I rigged up a desktop faraday cage out of an aluminum Dutch oven, but that hasn't entirely eliminate the problem. I still see around 5 or so mv of noise with no circuit in the cage, and I can still push that value around significantly by moving the cable and or my body, yet with nothing connected to the DSO I get a 500uv noise floor.

Hmm...

cellularmitosis · « **Reply #7 on:** January 17, 2014, 05:44:10 am »

Hmm...

"With your probe connected to a signal, move the probe cable around and grab it with your hands. If the waveform on the screen varies significantly, energy is being coupled onto the probe shield, causing
this variation. Using a ferrite core on the probe cable may help improve probing accuracy by reducing the common mode noise currents on the cable shield."

From: http://cp.literature.agilent.com/litweb/pdf/5989-7894EN.pdf

ConKbot · « **Reply #8 on:** January 17, 2014, 12:28:01 pm »

Hmm, the frequency of the AM signal is too high to be audio carrying radio, he's at 200ns/div on the slowest sweep there. Could also be an FM signal thats modulating in and out of what the scope can pass, hence the amplitude modulation.

rdl · « **Reply #9 on:** January 17, 2014, 01:02:25 pm »

Quote

Sorry about the photos, I just got my scope and haven't learned how to take screencaps yet.

Taking screenshots on these Rigol scopes is a bit of a pain. I don't know why they couldn't just have a dedicated button.

First, push the "Storage" menu button, then, on the screen select "External". From there you should be able to figure it out. Naming files on the scope is awkward, so I usually just transfer them to a computer and do it there.

Note that you have to have a USB flash drive inserted or the "External" option will be grayed out. My scope doesn't recognize the flash drive when it boots. I have to pull it out and re-insert it before the scope sees it. Maybe this has been fixed on newer firmware, I've never updated mine.

cellularmitosis · « **Reply #10 on:** January 18, 2014, 06:14:43 am »

Quote from: rdl on January 17, 2014, 01:02:25 pm

First, push the "Storage" menu button...
...Maybe this has been fixed on newer firmware, I've never updated mine.

Thanks rdl, I've now mastered the technique

And it looks like that has been fixed in modern firmware (my open menu updates from greyed out to enabled when I plug in a USB stick).

cellularmitosis · « **Reply #11 on:** January 18, 2014, 06:20:40 am »

Ok, another few data points.

I figured out that the center pin of my BNC cable didn't seem to be making reliable contact with the scope (I didn't figure this out until I took a DC coupled measurement and it didn't work).

So I tried a 1.5 foot cable I had, and here are the results.

The setup was the dutch oven, with internal alligator clips shorted to each other, with 2 lead weight bags on top to keep the lid shut securely, with an extra alligator clip from BNC shield to the bottom half of the oven (this didn't appear to make any difference).

It looks like I'm getting about 2mV (rms) of noise, and it appears to still be the AM modulated noise, at about 96 kHz.

cellularmitosis · « **Reply #12 on:** January 18, 2014, 06:27:41 am »

Next, I tried to see if I could rule out the effect of the dutch oven, and just focus on the cable.

I soldered up a DIY BNC cap to put on the end of the cable and just measured that (the cap shorts the tip to ground).

I've also overlayed channel two (in cyan), which has nothing connected to it and is uncapped. (This disconnected setup seems to be the lowest noise measurement I've been able to take thus far).

I'm now down to about 1.2 mV rms of noise (in yellow). This time around 93 MHz.

cellularmitosis · « **Reply #13 on:** January 18, 2014, 06:30:42 am »

The final data point for the evening was connecting the cap directly to the scope via a gender changer.

This yielded some different results: lower noise but the addition of what appears to be a 125 kHz switching noise. I had to enable some visual persistence in order to capture what I was seeing on the scope, because when you stop the capture, it turns out the blips aren't as regular as they appear.

cellularmitosis · « **Reply #14 on:** January 18, 2014, 06:32:18 am »

oops, missed the last screenshot.

So, I'm guessing the conclusion from all of this is that I need to find a better BNC cable? Because the cable seems to be the part which is imparting the 90ish Mhz noise.

Any advice on better quality BNC cables? I'm assuming the thickness and material (copper vs aluminum) of the braid shield is what determines the shielding quality? Is there a "Cadillac" in the BNC cable industry?

cellularmitosis · « **Reply #15 on:** January 21, 2014, 02:05:27 am »

I've ordered an RG-6 cable which has two layers of shielding, and while I'm waiting for that to arrive I wrapped my existing cable in a couple layers of al-foil. The result is similar to the plug above, where the 90ish Mhz noise is gone, and a smaller 125kHz switcher noise is seen.

I'm wondering where that switcher noise is coming from... (I've turned off all of the switchers in my house to no avail). Could this be the switcher noise of the Rigol itself? Has anyone ever hacked a Rigol to run off of batteries?

This was with a 50 ohm resistor as load. With the alligator clips shorted, the noise seems slightly less amplitude, but is otherwise the same.

cellularmitosis · « **Reply #16 on:** January 23, 2014, 02:39:04 am »

Ordered some shorted BNC caps, thinking that would reveal the ultimate noise floor of this scope. What I found was counter-intuitive...

Here's a measurement with BNC caps on all three front jacks, as well as the external pass/fail jack on the back:

Edit: note the presence of the 125kHz switcher noise, despite the BNC caps...

cellularmitosis · « **Reply #17 on:** January 23, 2014, 02:42:58 am »

...and here's the measurement with nothing connected at all. Our old friend the 125kHz switcher is gone!

How in the world am I getting a lower noise floor with nothing connected at all than I do with a shorting cap on? Shouldn't the shorting cap produce the lowest noise floor of all?

T3sl4co1l · « **Reply #18 on:** January 23, 2014, 04:23:51 am »

How far to the nearest radio station?

When I was in school, the school's radio station's antenna, on top of the dorm building, ensured that all test equipment within several blocks (the lab building being on the next block over) picked up erroneous signals. That was under 100W I think, but it was still pervasive. FM is 88-108 (at least in the US; I forget others), so it's a good first guess to suspect. Especially with the apparent voice/music modulation pattern. (Heck, get a wire, a small schottky diode and listen to it, see if you can find the callsign!)

I've heard that apparent amplitude modulation arises from multipath interference. Seems reasonable -- plenty of metallic structure in an urban environment, lots of ways to get position and frequency dependent interference patterns.

Tim

cellularmitosis · « **Reply #19 on:** January 23, 2014, 05:06:57 am »

Quote from: T3sl4co1l on January 23, 2014, 04:23:51 am

How far to the nearest radio station?

When I was in school, the school's radio station's antenna, on top of the dorm building, ensured that all test equipment within several blocks (the lab building being on the next block over) picked up erroneous signals. That was under 100W I think, but it was still pervasive. FM is 88-108 (at least in the US; I forget others), so it's a good first guess to suspect. Especially with the apparent voice/music modulation pattern. (Heck, get a wire, a small schottky diode and listen to it, see if you can find the callsign!)

I've heard that apparent amplitude modulation arises from multipath interference. Seems reasonable -- plenty of metallic structure in an urban environment, lots of ways to get position and frequency dependent interference patterns.

Tim

Actually, now that I think about it, I'm about 1 mile away from UT Austin's on-campus radio station, KVRX ("None of the hits, all of the time!"). Turns out they have a wikipedia page... http://en.wikipedia.org/wiki/KVRX

They are broadcasting at 3,000 watts, on 91.7 Mhz. Hmm...

T3sl4co1l · « **Reply #20 on:** January 23, 2014, 03:05:09 pm »

Yeah, that would probably do it...

As for abatement... do what you need to do for milivolts anyway. Probes suck, use direct coax connections wherever possible. Build deadbug/manhattan on copper clad instead of solderless breadboard, use lots of BNC connectors.

That'll at least reduce things to internal noise, which that switching noise could very well be. A shame, oh well. If you're desperate, you could try adding more filtering inside the scope -- ferrite beads and ceramic caps along the power supply wires to the main board, say, and shielding around the PSU if it doesn't have some already. (The power supply in my TDS460 is literally more than half inductors, it's ridiculous... but I bet it's at least as quiet as needed, too!)

Tim

rdl · « **Reply #21 on:** January 23, 2014, 03:19:00 pm »

When I first ran into the problem, I checked for any FM broadcast towers in the area. The nearest I could find was 20 miles away so I figured it was something else causing the noise. I never did track down the source. However, when you mentioned college radio something finally clicked in my brain. I had only looked for large commercial broadcasters, but there is a college nearby. Yesterday I checked into it and sure enough they have a radio station. I'd estimate it's only about 2 miles away and broadcasts at 6000 watts on 93.5 MHz, I'm now pretty sure that's the source of my noise problem.

cellularmitosis · « **Reply #22 on:** February 20, 2014, 03:24:29 pm »

I just realized I never updated this thread with the results of using the RG-6 cable. I'll do that tonight (it helped over the cheapy cable).


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: 90 MHz noise on scope? (Read 10350 times)

Share me