Don't know how to use a scope...

Thread Starter

s_mack

Joined Dec 17, 2011
198
Actually no I haven't. I was going to give that a go (as well as your other tips) tomorrow or Friday. I've got other work and family responsibilities and haven't had the time yet.
 

Thread Starter

s_mack

Joined Dec 17, 2011
198
Hi again. I had some time to experiment more with this. That FFT function (Spectrum Analyzer in the QA100 software) is neat.

But before we get to that, this is what I noticed experimentally: My setup has zero on-screen noise issue if I use an alternative motor controller... however, that's not a great option. I need to use the one I was originally using. So I'd like to filter the noise rather than resort to swapping components. So I thought a good idea would be to plot the FFT with the "good" component and again with the "bad" component and then I expected to see a spike or two that I could go filter... but here's the thing... to my untrained eye, it looks nearly identical?!!? I can't see anything special about the "bad" chart (red lines) that aren't on the good one (yellow).

- Steven

ps. wow, after uploading I see the red lines are hard to see :( sorry about that.
 

Attachments

theamber

Joined Jun 13, 2008
325
If you have the time, and really want to learn the basics of scopes (mainly analog scopes), take a look through the 14 videos in this playlist:
http://www.youtube.com/playlist?list=PL746BF38BC2E068E0
Hi, Alan I am Miguel Rodriguez from youtube. I agree those videos are GREAT. His videos are in terms we ordinary folks can understand. I need some videos on Digital Modulation: costellations, eye diagrams, EVM, etc...

If you dont see the signal it is because probably is to small increase the Vertical sensitivity and leave it in auto triggering.
 

thatoneguy

Joined Feb 19, 2009
6,359
Hi again. I had some time to experiment more with this. That FFT function (Spectrum Analyzer in the QA100 software) is neat.

But before we get to that, this is what I noticed experimentally: My setup has zero on-screen noise issue if I use an alternative motor controller... however, that's not a great option. I need to use the one I was originally using. So I'd like to filter the noise rather than resort to swapping components. So I thought a good idea would be to plot the FFT with the "good" component and again with the "bad" component and then I expected to see a spike or two that I could go filter... but here's the thing... to my untrained eye, it looks nearly identical?!!? I can't see anything special about the "bad" chart (red lines) that aren't on the good one (yellow).

- Steven

ps. wow, after uploading I see the red lines are hard to see :( sorry about that.
I can't really see the red, is there an option to change the color to blue or green? Is the vertical scale on Volts, mV, or dB?

The peak is at 10kHz, which is a pretty high freq, like sibilant sounds, cymbals, etc. The other peak is around 4kHz (a girl's shriek or there about), which is also a high annoying sound.

The only low noise is the minimal hump on the left from maybe 600 to 800 Hz (mid to low voice range).

Are any of these in the range of the noise you are hearing? Are these measurements at the output of the motor controller, or at the input of the audio power system? Are the readings before or after the filter?

Try all of the points above and label them.

Make sure your probes are switched to 10x and the scope is set for 10x probes/10M input impedance as well!

Best would be:
Right at motor controller power input, w/o filter
Right at audio input, w/o filter

Change motor speed while watching display, see if any peaks move to the left or right with the speed change.
 

Thread Starter

s_mack

Joined Dec 17, 2011
198
Sorry for the confusion... no audio involved here. The 'noise' I speak of are in the form of horizontal banding on video.

I'm using 1x, not 10x. Why 10x?

There's no color options that I can see... I can adjust in photoshop though. I'll see what I can do.

The vertical scale? Hmm... good question. If it is consistent with the scale on the main screen then its volts.

Yes, if I increase the speed the whole pattern shifts. However, on the video I don't see it getting any "worse" (or better). If its on... regardless of speed... there's similar (bad!) banding. If its off... no banding.

The readings are without a filter. I removed all filtration because my goal here is to build a better filter. Currently I'm following the earlier advice... I have the ground connected to the negative terminal of the battery and the probe to the positive.
 

thatoneguy

Joined Feb 19, 2009
6,359
S

I'm using 1x, not 10x. Why 10x?
1x Increases the "load" on the circuit you are probing, sometimes drawing more current than something like the input to a video amplifier would, therefore you don't see the exact same signals the video amp does.

In 10x, the impedance is also increased by 10x, which disturbs the measured circuit less, and can give you a bit more information, by "seeing" more like the video amp input does, if that makes sense.

Impedance matching, but trying to make your scope such a high impedance that it has zero effect on the circuit, but measure all of it. It's nearly impossible to measure something without perturbing it, but reducing that change all you can helps, if that makes more sense.

If your scope has input options of 50 Ohm and 1Meg Ohm, use the 1Meg input, which will be effectively 10 Meg with the x10 probe.
 

Thread Starter

s_mack

Joined Dec 17, 2011
198
I have 1x and 10x selectable on the probe itself and on the controls... that's it. I'll give that a try tomorrow. Thanks for the explanation. I thought that was only for use when I was trying to measure voltages that would otherwise be too high to safely probe. For example, 120VAC.
 

thatoneguy

Joined Feb 19, 2009
6,359
x10, x100, and x1000 probes are also used for measuring higher voltages or to have greater impedance, but the x10 is a "standard" setting for probes.

Unless there is a situation where I need to match a 50Ω or other load, I use x10 for most general purpose measurements. Since the scope goes down to 1mV to a few uV/div, the division isn't crippling with "normal level" signals. :)
 

vk6zgo

Joined Jul 21, 2012
677
You really have two problems:
(1) You have noise on your system.
(2) You are learning to drive your Oscilloscope.

I suggest you go back to (1),putting the 'scope to one side for the moment.

If the video system is an analog one,& you have horizontal banding on your Monitor/TV screen,any interfering signal would normally be lower in frequency than the line rate of 15kHz.
Frequencies above that usually appear as vertical banding.

You can check whether the problem is in the video camera,or the transmitter by turning the camera off.
You should then see a black screen on the TV.
Sometimes you will get a blue screen,but usually you can turn this function off.

Run the motor.
If it is upsetting the transmitter you should see noise bands on the screen.
If not,it looks like it is in the camera or its supply.

Now try looking at the video output of the camera with your Oscilloscope,at field rate.

If,as I believe is likely,your 'scope reduces its sampling rate at long
time/div settings,you may not get a usable display,but try,anyway.

One field of NTSC video is about 16mS duration,so you will probably have to set your time/div to around 5mS/div.
A PAL field is 20m/S.

If you get a usable display,run the motor,& if the noise is getting into the
camera,you should be able to see noise signals running through the video signal.

Try line rate, too,you may see something----- line duration is around
64 uS.

Check all your connections,particularly the power wiring.
If the return path of one of your devices is via the coax shield,rather than
the supply earth lead,you may well have problems.
 

Thread Starter

s_mack

Joined Dec 17, 2011
198
Thanks. All noted and I will try. Two immediate comments:
  • No coax. Actually none of the lines are shielded in this simple setup. While I realize that may be an alarm bell... others with similar setups have no visible noise and nobody uses shielded cables.
  • The monitor I'm using goes immediately to "no signal" if I disconnect the camera. But I understand your test, so I'll try to find a display that will show me static or whatever instead of blue-screening it
 

vk6zgo

Joined Jul 21, 2012
677
Another idea might be to run the transmitter off a separate 12volt supply
or battery (car battery,or maybe two 6V lantern batteries in series),for a brief test.

If you still have the interference,it is not caused by the transmitter power feed.

Does your monitor have a direct analog video input?

If so, you can just run the camera output into that with the motor running & check for interference.
Obviously,even if you have such an input,you will have to move stuff around to do this,which may not be convenient,or even possible.
 

Thread Starter

s_mack

Joined Dec 17, 2011
198
You know... that is what I ended up doing... running the transmitter and camera off one battery and the motor, etc off another. It cleaned it right up. I'd still like to tackle it with one battery but I haven't had a lot of time to work on it.

I did try the camera direct to the monitor. While powered the same way as initially, it still had the interference.
 
Top