Author Topic: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE  (Read 3982 times)

0 Members and 1 Guest are viewing this topic.

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« on: June 30, 2015, 06:09:38 PM »
Dear All

Please see screen shots attached, I have a RIGOL DS1054Z opened up to 100MHz and has software 00.04.02.SP4 board version 0.1.1.

This is a fantastic bit of kit but I am seeing needle spikes from my circuit when the display intensity is set to 100% on the Rigol, if I turn it down to 50% the spikes disappear as you might expect.

My question is are they real because as you can see there is no evidence of these spikes on my Gould DSO 465 100MHz scope with the display intensity set to full?

 

Offline Dave

  • Super Contributor
  • ***
  • Posts: 1019
  • Country: si
  • I like to measure things.
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #1 on: June 30, 2015, 06:19:36 PM »
How are you probing your circuit? Could you take a photo of your probing setup?
<fellbuendel> it's arduino, you're not supposed to know anything about what you're doing
<fellbuendel> if you knew, you wouldn't be using it
 

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #2 on: June 30, 2015, 06:38:17 PM »
Both the Rigol and the Gould are attached to the same small point with probe hook tips, the ground leads are also connected together so there both seeing the very same signal.
 

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #3 on: June 30, 2015, 06:44:09 PM »
I have now attached the scope lead connections, this is to do with the differences of the scopes I fear and how they sample a signal, but I would like everyones input
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #4 on: June 30, 2015, 06:48:48 PM »
Don't use ground leads, use ground springs instead:



Digital scopes can see more noise than analog ones.
 

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #5 on: June 30, 2015, 07:04:06 PM »
I have just tried this and yes this does indeed solve the Rigol needle spikes, but why is the Gould no showing these spikes when BOTH the Rigol and the Gould are looking at the very same signal?
 

Offline ebastler

  • Super Contributor
  • ***
  • Posts: 1113
  • Country: de
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #6 on: June 30, 2015, 08:51:26 PM »
The Gould has a lower sampling rate it seems (200 MHz?), but should still be able to pick up at least some of the spikes. Have you set both probes to 10x, or is the one on the Gould set to 1x? In that case, the lower input resistance or larger capacitance might dampen the spikes.
 

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #7 on: June 30, 2015, 09:14:14 PM »
Yes both probes are set to x10 and the Gould never sees a spike when both probes are connected to the very same electrical node, scope grounds also.

I have to also conclude this is the sampling regime of the Gould in the absence of further information.

The the spike are indeed true, but generated by the length of scope ground, the spring ground method of signal observation shows no spikes :D
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 1438
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #8 on: June 30, 2015, 09:20:54 PM »
Don't use ground leads, use ground springs instead:
+1. At 100MHz those few inches of ground-lead wire may as well not be there.
 

Offline H.O

  • Frequent Contributor
  • **
  • Posts: 474
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #9 on: June 30, 2015, 09:34:43 PM »
The 20MHz BW limit on the Rigol is ON and the sample rate is "only" 4MSa/s.
The screenshot of the Gould scope doesn't show what the sample rate is at that timebase. Do you have documentation on the scope where you can see what the sample rate is at various timebase setting? If that's not available then what IS the timebase set to and what's the memory depth of the scope?

EDIT: Oh, it says TB=1s on the Gould so that's got to be 1s/div. What's the memory depth?

EDIT2: This operators manual for the Gould 400 (doesn't mention the 465 though) seems to indicate the acquisition memory is a whopping 512bytes which would make the sample rate for a 10 second "sweep" 51.2Sa/s....
« Last Edit: June 30, 2015, 09:43:27 PM by H.O »
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 9203
  • Country: nz
    • Taupaki Technologies Ltd.
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #10 on: June 30, 2015, 09:46:57 PM »
Time for some study of Interpolation errors as they relate to sampling speed and memory depth.
Analogue CROs are immune from this.

Answer: False
« Last Edit: June 30, 2015, 10:42:06 PM by tautech »
Avid Rabid Hobbyist & NZ Siglent Distributor
 

Offline PICmonster

  • Regular Contributor
  • *
  • Posts: 61
  • Country: gb
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #11 on: June 30, 2015, 10:47:07 PM »
Thanks to all who have contributed, I have attached photos of the Gould scope, not sure is it has any memory depth, it has averaging but this is off.

This scope could not detect the spikes but the Rigol did
 

Offline ebastler

  • Super Contributor
  • ***
  • Posts: 1113
  • Country: de
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #12 on: July 01, 2015, 12:40:13 AM »
Have you tried to "zoom in" on the spikes by switching to a much faster time base, and trying to trigger on the spikes? (Using either scope, and putting the long ground leads back in place, of course.) That might tell you how fast the spikes really are, and whether a sampling/averaging issue on the Gould is plausible.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 1770
  • Country: ca
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #13 on: July 01, 2015, 01:14:36 AM »
I recommended this in another thread too : put a snapon ferrite choke on the probe lead close to the front panel connector, see if it helps.
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 3640
  • Country: au
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #14 on: July 01, 2015, 01:53:56 AM »
The 20MHz BW limit on the Rigol is ON and the sample rate is "only" 4MSa/s.
The screenshot of the Gould scope doesn't show what the sample rate is at that timebase. Do you have documentation on the scope where you can see what the sample rate is at various timebase setting? If that's not available then what IS the timebase set to and what's the memory depth of the scope?

EDIT: Oh, it says TB=1s on the Gould so that's got to be 1s/div. What's the memory depth?

EDIT2: This operators manual for the Gould 400 (doesn't mention the 465 though) seems to indicate the acquisition memory is a whopping 512bytes which would make the sample rate for a 10 second "sweep" 51.2Sa/s....

If the spikes disappear when  the contact spring is used,it appears that the OP is picking up high frequency signals from some external source.
Such signals may cause aliasing problems,if their frequencies are  greater than the sample rate,& still well below the 20MHz roll off.

"Ringing" at high frequencies may cause a similar effect.

Perhaps the Gould doesn't use real time sampling at such long time/div settings,or if it does,inserts a 4MHz roll off.
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 3640
  • Country: au
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #15 on: July 01, 2015, 03:15:23 PM »
What I am implying in the above is that the spikes are not "real",but are due to interaction between the sample rate & some external high frequency signals picked up due to your "long-ish" ground lead.

(You must have a lot of "grot" around 4MHz or so in your location,as normally pickup at such frequencies would be minimal with a normal ground lead)

I think if you try to "zoom in" on the spikes with a much shorter time/div,they will not be present,nor will they be present on a higher level DSO,or an Analog 'scope (although the latter may be battling to give a useable display at 500mS/div),using the existing settings.

By the way,do the narrow pulses on the LH side of the screen show up correctly on the Rigol when you use the cround spring?
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #16 on: July 01, 2015, 04:03:37 PM »
I agree, the spikes are not on the circuit, they are just picked up by the ground lead that acts like an antenna, same as if you probe your skin you'll see a 60Hz (or 50Hz) from your mains picked up by your body.

I could as well be harmonics too, so the actual signal interfering with your ground lead might be a higher frequency, from your router or who knows.

The short ground springs don't make the problem go away, they just wont introduce noise to your measurements.
 

Online vk6zgo

  • Super Contributor
  • ***
  • Posts: 3640
  • Country: au
Re: RIGOL DS1054 NEEDLE SPIKES TRUE OR FALSE
« Reply #17 on: July 01, 2015, 05:37:14 PM »
Just for fun,I decided to play around with the Tek 7613 to see if I could see signals at around the rate the OP was probing.
As I don't have anything which will produce pulses at such a low frequency,I tried just touching the probe & ground lead across a "C" cell,with the 'scope set to 500mS/div & dc coupled.

To my surprise,I saw no vertical deflection,just the spot doggedly scanning across the screen.

I changed to a faster timebase setting---still no sign of the dc shift expected,but lots of 50Hz hum.
It took a bit of fiddling before I hit the "Duhh!" moment.
An open circuit ground lead!

Fixed that,& had another try.
As I anticipated,the normal display didn't show much,but the "storage" mode did--for a while,before I stuffed it up trying to make it better!

Anyhow,all you would see would be a bit of a thickening of the trace if there was external grot on it-----none here at all.
I bet it would have picked up any HF stuff around,along with the 50Hz,if the probe's ground lead was still  faulty,though.

Maybe the OP should check the continuity of his ground lead,just in case!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/