Author Topic: Differential probe substitute  (Read 725 times)

0 Members and 1 Guest are viewing this topic.

Offline veedub565

  • Regular Contributor
  • *
  • Posts: 65
  • Country: gb
Differential probe substitute
« on: January 12, 2019, 05:38:51 am »
Quick question, for measuring ripple and noise on a PSU, could I use a matching pair of scope probes instead of spending a fortune on a differential probe? I already have a differential amplifier with two BNC inputs (and single BNC o/p)

From what I can tell, all the differential probe seems to be is a matched pair of scope probes. Or is there more to it than that ?

Thanks
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 8165
  • Country: us
  • DavidH
Re: Differential probe substitute
« Reply #1 on: January 12, 2019, 06:30:42 am »
That can work and I have done it myself but it is actually more difficult with a differential amplifier.  The trick with a differential probe is that the DC and AC characteristics of each side match closely resulting in high common mode rejection.  This is more difficult to do with separate probes.  Some old x10 probes intended for this application have DC and AC trims but they are very rare.  The problem with a differential amplifier is that they are not made to be adjustable either.  This leaves a couple of solutions:

1. Use x1 probes with your differential amplifier.  x1 probes all match since they are effectively straight through connections without attenuation.  Unfortunately, this will limit the input range of your amplifier although some old amplifiers had build in switchable precision attenuators (Tektronix 7A13, 7A22, AM502, a bunch of earlier ones) and can be used like this.

2. Use x10 probes with an oscilloscope in A-B mode.  Now the variable function of one or both channels can be used to match the attenuation of the x10 probes.  This works surprisingly well with an analog oscilloscope (or DSO which has an analog front end) and you can *see* the point where the attenuation of the probes match because the noise decreases.  It works less well with a DSO which does the variable attenuation and subtraction digitally because of increased quantization noise.
 

Offline veedub565

  • Regular Contributor
  • *
  • Posts: 65
  • Country: gb
Re: Differential probe substitute
« Reply #2 on: January 16, 2019, 12:29:48 am »
Thanks David, sorry for the late reply. I do have switchable attenuators on my diff amp (1Mohm and 100Mohm I think) so that might be worth a try with probes in x1 mode instead of x10. I'm going up to about 40v with it.

Not sure if my scope has a digital front end or not, it's an old HP infinium, basically a PC board inside with a scope front end.

At least I know now it's a workable idea to use a pair of scope probes instead of a diff probe.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 8033
  • Country: gb
    • Having fun doing more, with less
Re: Differential probe substitute
« Reply #3 on: January 16, 2019, 01:00:14 am »
If one side of the PSU can safely be connected to a probe shield (and hence protective mains earth), then consider a single-ended measurement with the scope's channel coupling set to AC.
There are lies, damned lies, statistics - and ADC/DAC specs.
Gliding aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 8165
  • Country: us
  • DavidH
Re: Differential probe substitute
« Reply #4 on: January 16, 2019, 01:34:01 pm »
If one side of the PSU can safely be connected to a probe shield (and hence protective mains earth), then consider a single-ended measurement with the scope's channel coupling set to AC.

That usually does not work with noise measurements on switching power supplies because they produce so much common mode noise.  It might work with an oscilloscope that has isolated inputs or with an isolated probe.
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 5536
  • Country: 00
Re: Differential probe substitute
« Reply #5 on: January 16, 2019, 03:20:39 pm »
That usually does not work with noise measurements on switching power supplies because they produce so much common mode noise.  It might work with an oscilloscope that has isolated inputs or with an isolated probe.

How about arrangement like in this ATX standard document stated, as they used P6046 probe which is not isolated ?


Offline ycui7

  • Newbie
  • Posts: 2
  • Country: us
Re: Differential probe substitute
« Reply #6 on: January 16, 2019, 04:13:03 pm »
How about getting a battery powered oscilloscope? They are not expensive these days.

My colleague sometime float them on kV rails to measure small noise. You do need to be very careful to do that though.
 

Online blueskull

  • Supporter
  • ****
  • Posts: 10583
  • Country: cn
  • Power Electronics Guy
Re: Differential probe substitute
« Reply #7 on: January 16, 2019, 04:36:04 pm »
If you only need to see ripple on a DC majority bus, how about using a capacitor and transformer coupled probe to capture the AC component?

BTW, at a first glance, I thought the title was differential prostitute |O. I need to go to bed.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4395
  • Country: au
Re: Differential probe substitute
« Reply #8 on: January 16, 2019, 05:47:43 pm »

BTW, at a first glance, I thought the title was differential prostitute |O. I need to go to bed.
Interesting concept! ;D
 

Online Marco

  • Super Contributor
  • ***
  • Posts: 3877
  • Country: nl
Re: Differential probe substitute
« Reply #9 on: January 16, 2019, 08:45:03 pm »
From what I can tell, all the differential probe seems to be is a matched pair of scope probes. Or is there more to it than that ?
It varies. High voltage differential probes have matched dividers, but say the chipwhisperer differential probe does not. Matched dividers will have poor CMRR at higher frequencies.

A power supply is a very low impedance source though, it makes far more sense to simply AC couple the scope.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 8033
  • Country: gb
    • Having fun doing more, with less
Re: Differential probe substitute
« Reply #10 on: January 16, 2019, 09:01:34 pm »
If one side of the PSU can safely be connected to a probe shield (and hence protective mains earth), then consider a single-ended measurement with the scope's channel coupling set to AC.

That usually does not work with noise measurements on switching power supplies because they produce so much common mode noise.  It might work with an oscilloscope that has isolated inputs or with an isolated probe.

Agreed.

It isn't clear that the OP has an SMPS. I would be very cautious about assuming a ground can be safely attached to a mains-DC SMPS, but a DC-DC SMPS should be easier to assess.
There are lies, damned lies, statistics - and ADC/DAC specs.
Gliding aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Online Wolfgang

  • Frequent Contributor
  • **
  • Posts: 955
  • Country: de
  • Its great if it finally works !
    • Electronic Projects for Fun
Re: Differential probe substitute
« Reply #11 on: January 17, 2019, 10:51:59 am »
... please tell me about the voltage levels and the maximum frequency you want to measure.
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 4395
  • Country: au
Re: Differential probe substitute
« Reply #12 on: January 17, 2019, 11:11:55 am »
... please tell me about the voltage levels and the maximum frequency you want to measure.

Best answer of the lot, Wolfgang!
People often go off at tangents when they don't have the full information.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 8165
  • Country: us
  • DavidH
Re: Differential probe substitute
« Reply #13 on: January 18, 2019, 05:31:48 pm »
How about arrangement like in this ATX standard document stated, as they used P6046 probe which is not isolated ?

The P6046 is a differential probe.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf