Noise question

[cs.dk]

I've thrown a simple "breakout harness" together, to take a measurement. The measurement though is very noisy.

First a (attached) picture from a Tek 2246 (sorry, dark surroundings, a bright screen and a phone isn't the shit to make pictures) - I does not help to put the HF, LF or Noise reject on to the triggercirciut, which i don't understand 100%.

A pic from a DSO Nano (Yeah, i know ) The Nano just don't measure the correct voltage. At the Tek it's measuring approx. 13V

It should be a 1 KHz squarewave at ~12V (car battery voltage). As i can see, my only other option is to measure the output direct on the VN05 driver, to see if it is noisy too?


It is a Citroën Xantia Activa HA2 suspension controller. The load is a solenoid measuring 4 DC ohms.

My question is, can this little breakout-cable pick up so much noise? Why is the noise gone at the Nano?

[Jon86]

Can you explain your probing setup in a little more detail?
Also, is your question 'why does the Tek scope pick up so much noise'? Sorry, I'm having a little difficulty understanding what you're trying to do.

[kfitch42]

I am a newb as well, but I think part of the answer is that you are comparing instruments of WILDLY different capabilities (200KHz bandwidth vs 100 MHz bandwidth)

0) Are the instruments configured the same? e.g. AC vs DC coupling?

1) Is your evidence for "noise" the thickness of the lines at the top and bottom of the traces on the Tek?

2) One thing to try is to bit the bandwidth limit on the Tek (labelled "SCOPE BW 20MHZ" on yours), that will bring it down to 20Mhz. Then the instruments will only be 2 orders of magnitude apart (instead of almost 3).

3) Assuming for a moment that the noise is in the circuit instead of an artifact of the measurement: Is this noise problematic in your particular circuit?  What specific problems is it causing?

4) How is the setting of the trigger relevant? Is the trace jumping left/right on the screen?

[Jon86]

Something to note is that the scope is mains earth referenced, and the Nano is obviously not. So that could possibly have something to do with it.
Try hooking your cable up to the test signal on your scope, and maybe as kfitch mentioned, have a play with the settings on your scope such as the bandwidth limit.

[cs.dk]

Thanks for the replys.

Probe setup is simple. coupling the little wire in circuit, and measure on the two loose ends. The probe it self is a fairly common 100 MHz 1-10x probe.
Yes, the question is; 'why does the Tek scope pick up so much noise'? - Sorry if it wasn't clear, English isn't my native language. 
Is it the length of the breakout cable that picks the noise up, or could the signal really be that noisy?

For AC vs DC coupling is doesn't do anything to the signal, just moves the curve vertical. Didn't try the BW limiter though.

The noise i mean is that the squarewave has 3 squarewaves side by side, if that makes sense?

I don't think the noise is problematic, it's just driving a solenoid/coil - I was just very surprised to see it, main reason for this thread.

[owl_uk]

I think I can see what is happening.

Your analogue scope doesn't show you one snap trace of the waveform.  It is scanning the waveform repeatedly.  So what you are seeing is that the time between the trigger event and your square wave is not constant.  It is sometimes quite short, giving you the first sqaure wave, then it is a little longer, which gives you the next one.

So there are a few possible delays between the trigger event and the square that you are seeing.  You've not really explained what the wave is.  If it is the output from a pwm (or class D amplifier) you might see something like this.  It also might be to do with how you are triggering the scope.

The digital scope gives you one trace of one set of samples, i.e. you are looking at one event.  You can think of the analogue trace as being like lots of the digital traces piled up on top of each other.

This is where a nice expensive scope is very usefeul.  I used to have a TEK DPO in my lab (where I was employed - it wasn't actually mine).  This acts like a  phospor analogue scope but has a very useful 'single seq' button for looking at a single trace.  So with that you could get both of your sets of results.

So in summary, I think this is a scope setup issue, not a noise issue.  I hope that's helpful.

[cs.dk]

Weird.. Could not post in this thread again, maybe because of attachments?

Sorry for the long reply time, had to measure a little more to be sure to test trigger delay, LF/HF/Noise-reject, BW-limit, and so on.

This measurement is right of the VN05N drivers output pin. I actually don't know whats creating the waveform, as mentioned above, it is a Citroën specific suspension controller. There is a couple of TI-controllers; TMS370C310 and a TMS370C756, not much more. I'll attach a picture, and a PDF explaining the system, if some should be courious. (Who knows)
The video will show the unstable waveform - Thats my main concern.. Shouldn't i be able to trigger a 1 KHz squarewave?

Videolink, as the file is too big; https://dl.dropboxusercontent.com/u/6429978/EEVBlog/Unstable_Wave.mp4

[qno]

I think it is the probe.
Did you use the grounding wire of the probe?
Check the forum for scope probe use.

[cs.dk]

Thanks for your reply - This topic seems to avoid people from answering. 
Yes, the groundclip is connected, as you can see by the picture above - I've got two other probes, but they are out of my reach at now. The only thing i can think of regarding the probe, is a broken BNC cable, which i doubt.

Should a 100 MHz scope not be able to trigger, and produce a stable waveform on a 1 KHz squarewave? Or is this event just to expect?

[sync]

It's not a stable signal. See at the Nano screen shot. The pulse width are varying. On the Tek try triggering on the positive edge. Maybe this makes it more stable.

[kfitch42]

Two things come to mind that haven't been mentioned yet:

1) Trigger hold-off. I think Dave made a video about it a while back.

2) Get a camera that can do a long exposure (even 1 second should _probably_ be enough ... search into menus, most cameras can do this). Set the camera pointing at the screen (tripod would be perfect). Hit the shutter button, then quickly hit the "single shot" button on your scope. This should give a result more similar to the nano.

[Rudane]

cs.dk;

I agree with the previous poster's reply, you are likely trying to view a pulse width modulated signal. If this is the case the signal is not periodic, and as such cannot be seen in the manner you are trying with the analog scope.

[cs.dk]

It's not a stable signal. See at the Nano screen shot. The pulse width are varying. On the Tek try triggering on the positive edge. Maybe this makes it more stable.

I think you are right. Haven't noticed that! I did play around with all triggermodes that i could think of, but nothing seems to get a stable readout. Here would a singleshot feature become useful i think.

Two things come to mind that haven't been mentioned yet:

1) Trigger hold-off. I think Dave made a video about it a while back.

2) Get a camera that can do a long exposure (even 1 second should _probably_ be enough ... search into menus, most cameras can do this). Set the camera pointing at the screen (tripod would be perfect). Hit the shutter button, then quickly hit the "single shot" button on your scope. This should give a result more similar to the nano.

Trigger holdoff does nothing on this waveform. I actually don't know if the feature works on this scope. Gotta follow up on that. May have to see the video.

Great idea. Haven't figured that one out. The singleshot is impossible to see with eye.

cs.dk;

I agree with the previous poster's reply, you are likely trying to view a pulse width modulated signal. If this is the case the signal is not periodic, and as such cannot be seen in the manner you are trying with the analog scope.

Yes, it makes sense to me, now the explanation comes.

Thanks for all replys. Leaning something new everyday