Author Topic: Scope channel Noise difference? (Read 18175 times)

Wmacky · « **on:** March 21, 2015, 08:05:51 am »

Is there an acceptable limit for a noise difference between channels on a brand new 2 channel scope, and with nothing connected to the inputs? Does it matter?

rs20 · « **Reply #1 on:** March 21, 2015, 08:22:34 am »

Check the noise relative to the specs of the scope, rather than between channels. If the noisier channel is out of spec, then there's a confirmed problem. Otherwise, your scope is in spec. I don't particularly see a reason for defining acceptable limits on noise equality between channels.

Having said all that, the channels are supposed to be identical circuitry so it is some sort of cause for concern. Although, at the same time, I'm always skeptical of these people who test the noise floor of their scopes by leaving the inputs unconnected. Unconnected means open circuit, high impedance, extreme noise sensitivity. It's entirely possible that the noisier channel just has a noise source physically closer to it on the PCB, but the problem goes away once a reasonably low-impedance signal is attached. Of course, 10x probes hardly affect the impedance of the node, but still. It'd make more sense to me to connect a terminator to the input to test noise floor, and would at least help with diagnosing your issue.

Wmacky · « **Reply #2 on:** March 21, 2015, 09:16:07 am »

I just found the difference odd, as my upstanding is both channel are built the same. That, and the fact that all the Youtube videos I've watched for this scope seems to indicate identical noise levels in the all that showed unconnected inputs. I switched in the scope 50 Ohm input with no difference. Channel 1 seems consistently noisier than channel 2.

T3sl4co1l · « **Reply #3 on:** March 21, 2015, 09:40:01 am »

Measurements? Model?

Wmacky · « **Reply #4 on:** March 21, 2015, 09:54:36 am »

Quote from: T3sl4co1l on March 21, 2015, 09:40:01 am

Measurements? Model?

Rigol 2302a

T3sl4co1l · « **Reply #5 on:** March 21, 2015, 10:05:14 am »

Measurements?

JohnnyBerg · « **Reply #6 on:** March 21, 2015, 10:05:44 am »

Quote from: Wmacky on March 21, 2015, 08:05:51 am

and with nothing connected to the inputs?

I thinks you should put both inputs to ground, with no probes attached. Noise injection could be different, depending on the location, and how the scope is placed on the bench.

Wmacky · « **Reply #7 on:** March 21, 2015, 10:14:21 am »

Quote from: JohnnyBerg on March 21, 2015, 10:05:44 am

Quote from: Wmacky on March 21, 2015, 08:05:51 am
and with nothing connected to the inputs?

I thinks you should put both inputs to ground, with no probes attached. Noise injection could be different, depending on the location, and how the scope is placed on the bench.

How best to achieve this?

Wmacky · « **Reply #8 on:** March 21, 2015, 10:15:10 am »

Quote from: T3sl4co1l on March 21, 2015, 10:05:14 am

Measurements?

Which measurements? Kinda new to DSO's

Do you mean total noise levels? At full 300 MHz bandwidth?
If so, and by visual guesstimation Channel 1 - about 5.5mv Channel 2 - 6 mv

dom0 · « **Reply #9 on:** March 21, 2015, 10:21:11 am »

That thing probably has measurement functions. For now peak to peak voltage and RMS voltage are of interest. Also check whether the noise is always different or only on the lowest vertical setting.

Some compact scopes (the typical Hameg or Tek 2xxx series design with the tube left and the controls and inputs to the right) had issues with the channel closer to the CRO being a bit noisier and the like. Given that many displays radiate EMI like there were no tomorrow I wouldn't be surprised if some DSOs have similar issues.

JohnnyBerg · « **Reply #10 on:** March 21, 2015, 10:27:58 am »

Quote from: Wmacky on March 21, 2015, 10:14:21 am

Quote from: JohnnyBerg on March 21, 2015, 10:05:44 am
Quote from: Wmacky on March 21, 2015, 08:05:51 am
and with nothing connected to the inputs?

I thinks you should put both inputs to ground, with no probes attached. Noise injection could be different, depending on the location, and how the scope is placed on the bench.

How best to achieve this?

Huh? Just select GND as the input coupling on the scope.

Wmacky · « **Reply #11 on:** March 21, 2015, 10:31:39 am »

From a single shot:

Channel 1

VRMS 620uV
VPP v 1.8 mV

Channel 2

VRMS 542uV
Vpp 1.56 mV

JohnnyBerg · « **Reply #12 on:** March 21, 2015, 10:35:43 am »

There is not that much difference between the channels.

But I find the noise rather high. Can you show a picture on how you measure this?

Wmacky · « **Reply #13 on:** March 21, 2015, 10:36:54 am »

Quote from: JohnnyBerg on March 21, 2015, 10:27:58 am

Quote from: Wmacky on March 21, 2015, 10:14:21 am
Quote from: JohnnyBerg on March 21, 2015, 10:05:44 am
Quote from: Wmacky on March 21, 2015, 08:05:51 am
and with nothing connected to the inputs?

I thinks you should put both inputs to ground, with no probes attached. Noise injection could be different, depending on the location, and how the scope is placed on the bench.

How best to achieve this?

Huh? Just select GND as the input coupling on the scope.

No "ground" Option under input. Ground coupling only.....

dom0 · « **Reply #14 on:** March 21, 2015, 10:38:29 am »

Ground coupling is precisely that. It connects (couples) the input to ground.

JohnnyBerg · « **Reply #15 on:** March 21, 2015, 10:42:32 am »

Quote from: Wmacky on March 21, 2015, 10:36:54 am

Quote from: JohnnyBerg on March 21, 2015, 10:27:58 am

Huh? Just select GND as the input coupling on the scope.

No "ground" Option under input. Ground coupling only.....

YT on scope basics?

Wmacky · « **Reply #16 on:** March 21, 2015, 10:48:21 am »

Quote from: dom0 on March 21, 2015, 10:38:29 am

Ground coupling is precisely that. It connects (couples) the input to ground.

In that case, no noise with input to ground, All measurements =0

Wmacky · « **Reply #17 on:** March 21, 2015, 10:49:21 am »

Quote from: JohnnyBerg on March 21, 2015, 10:42:32 am

Quote from: Wmacky on March 21, 2015, 10:36:54 am
Quote from: JohnnyBerg on March 21, 2015, 10:27:58 am

Huh? Just select GND as the input coupling on the scope.

No "ground" Option under input. Ground coupling only.....

YT on scope basics?

Thanks, I'll get right on that.

rs20 · « **Reply #18 on:** March 21, 2015, 10:58:08 am »

I'm not sure GND coupling is a useful option for Rigol scopes; on my Rigol DS2202 GND coupling is quite clearly just a digital zero, there's no noise at all.

Wmacky · « **Reply #19 on:** March 21, 2015, 11:12:55 am »

Quote from: rs20 on March 21, 2015, 10:58:08 am

I'm not sure GND coupling is a useful option for Rigol scopes; on my Rigol DS2202 GND coupling is quite clearly just a digital zero, there's no noise at all.

Which led to my confusion. Your correct, ground coupling appeared to kill the internal noise completely, and really didn't offer anything. I wasn't sure I was understanding Johnnyberg's request completely.

Here's a pic.

Wmacky · « **Reply #20 on:** March 21, 2015, 11:23:56 am »

Lots of variability between these single shot numbers. The pics I'm posting are with DC coupling. The very first numbers were with AC coupling. Bigger channel difference with DC. Notice the central spike on channel 1 that's always there. It moves with the horizontal position control, and only exists on DC coupling?

Another

JohnnyBerg · « **Reply #21 on:** March 21, 2015, 11:24:04 am »

And what is on the inputs?
Open scope probes are antenna's

I have some of these, to "seal" the inputs.

Wmacky · « **Reply #22 on:** March 21, 2015, 11:26:59 am »

Quote from: JohnnyBerg on March 21, 2015, 11:24:04 am

And what is on the inputs?
Open scope probes are antenna's

I have some of these, to "seal" the inputs.

Nothing connected. I have no terminators.

I'm afraid my "return" window would close before some could arrive.................................

dom0 · « **Reply #23 on:** March 21, 2015, 11:27:22 am »

Yeah, stick a terminator in or short the input out with a small piece of wire from the center conductor directly to the shell of the BNC.

Interesting that Rigol tries to cheat with GND coupling. How dumb do they think their customers are?

JohnnyBerg · « **Reply #24 on:** March 21, 2015, 11:31:01 am »

Quote from: dom0 on March 21, 2015, 11:27:22 am

Interesting that Rigol tries to cheat with GND coupling. How dumb do they think their customers are?

I thought it was a assumption? Did you look look at the schematics?


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

Author Topic: Scope channel Noise difference? (Read 18175 times)

Share me