Author Topic: Analogue V digital scope trace question  (Read 1961 times)

0 Members and 1 Guest are viewing this topic.

Offline Chris Wilson

  • Super Contributor
  • ***
  • Posts: 1002
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Analogue V digital scope trace question
« on: July 02, 2015, 01:07:08 pm »
I am, with kind remote assistance, investigating display noise on an old HP8568B spectrum analyser. I am using an even older Tektronix 7854 storage `scope,. I am seeing noise spikes only when using the acquire function, they are not clearly visible whilst the `scope is free running. I also tried to look for them with an HP54501A digital scope. This is proving useless to see the spikes, but maybe it's the operator? Before I go further, should I expect to use the digi `scope to see such noise spikes as in the image hopefully attached? I am using a X10 probe, and the Tek `scope does not reflect that in the V/Div reading. Thanks.

The Tek is a 500MHz scope, the HP a 100MHz, and the HP specs are below:

oduct Identifiers
Brand   Agilent
Model   54501A
Type   Digital
Bandwidth   100 MHz
MPN   54501A

Acquisition System
Sample Rate   10 MSa/s
Memory Depth   501 points

Vertical System
Vertical Sensitivity   5 mV/div to 5 V/div
Rise Time   3.5 ns
Input Impedance   1 M?
Vertical Resolution   8 bits

Time Base System
Time Base Range   2 ns/div to 5 s/div
Time Base Accuracy   ±0.005%
« Last Edit: July 02, 2015, 01:14:52 pm by Chris Wilson »
Best regards,

                 Chris Wilson.
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 1705
Re: Analogue V digital scope trace question
« Reply #1 on: July 02, 2015, 01:30:21 pm »
What do you hope to get by seeing the spikes more clearly?
they are not clearly visible whilst the `scope is free running.
This suggests, as you would expect, that the spikes are correlated with some other activity in the analyser. My next step would be to try to work out what the spikes are correlated with. Then if you really want a  better picture of the spikes you could trigger off the event that causes them.

Edit: Having found the spikes with an analogue scope, my preference would be to stick with analogue - and use a second channel to "probe around" - just be careful about chop/alternate sweep modes.
Curious that the spikes only appear on positive signals - it's almost as if a power-supply was dropping out.
« Last Edit: July 02, 2015, 01:37:59 pm by Andy Watson »
 

Offline Chris Wilson

  • Super Contributor
  • ***
  • Posts: 1002
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Analogue V digital scope trace question
« Reply #2 on: July 02, 2015, 02:07:04 pm »
Consensus seems to be the spikes are CMOS originated, but digital stuff is a bit of a mystery, and this thing is complexity personified. I have done all the PS tests I know of and all the supplies seem clean and correct. Another opinion is something is causing big current spikes and they are then reflected throughout many circuits.  In an ideal world i am told the use of a second SA to "look at things" would be ideal, but i don't have one. I am limited to sig gen, frequency counter, scopes and meters. MAYBE I could use an SDR dongle and SDR receiver software as a very poor man's SA, but i am getting ahead of myself. I am hoping to divide and conquer, but the spikes seem to appear all over the damned place. The SA display randomly produces noise spikes, but otherwise the SA works seemingly OK. Sometimes it shows no noise. I see no real correlation between the SA started from cold, or whn hot, as to when the noise appears. The signal is from the 20MHz test port on the SA. Injecting a clean 20MHz from a sig gen still shows the noise. I made a video of the display noise at:

http://www.gatesgarth.com/noise.avi  (biggish file).

Thanks Andy.
Best regards,

                 Chris Wilson.
 

Offline Chris Wilson

  • Super Contributor
  • ***
  • Posts: 1002
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Analogue V digital scope trace question
« Reply #3 on: July 03, 2015, 10:16:53 am »
Curious that the spikes only appear on positive signals - it's almost as if a power-supply was dropping out.


Hmm, I should have posted the other screen shot Andy. I presume this is showing that when i change the span of the SA the spikes change to appearing on negative signals?
Best regards,

                 Chris Wilson.
 

Offline Chris Wilson

  • Super Contributor
  • ***
  • Posts: 1002
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Analogue V digital scope trace question
« Reply #4 on: July 08, 2015, 02:13:02 pm »
After having wasted good helpful people's valuable time looking at these traces on another forum I have now discovered to my embarrassment that they are caused by a fault condition in my Tek `scope. it's a known issue, and likely fairly easily resolved. I discovered it by accident when I connected a Tek time / mark generator to the `scope to check calibration. If use used the acquiescence mode I saw similar noise spikes.  I then tested with a decent sig gen, and at higher sweep speeds the noise was again present. Something to do with the unusual way the waveform acquisition is done on the complex 7854 `scopes... I have offered to crawl under a stone on the other forum, I posted loads of screen shots oblivious to the possibility the noise spikes were generated within my own `scope :(

So, if anyone reads this, never assume funny looking noise spikes are actually something seen at the probe! I live and learn, the joys of playing with older analogue `scopes I guess.  A salutary lesson learnt.
Best regards,

                 Chris Wilson.
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 1705
Re: Analogue V digital scope trace question
« Reply #5 on: July 08, 2015, 02:28:02 pm »
A salutary lesson learnt.
:) live and learn 

But, presumably, it doesn't solve the problem with the SA.
 

Offline Chris Wilson

  • Super Contributor
  • ***
  • Posts: 1002
  • Country: gb
  • Race car engineer, dog lover, hoarder.
Re: Analogue V digital scope trace question
« Reply #6 on: July 08, 2015, 03:17:44 pm »
Unfortunately not, it just adds to the list of faulty "bargains" :)
Best regards,

                 Chris Wilson.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf