Author Topic: Mismatched bandwidth between scope and probe?  (Read 6310 times)

0 Members and 1 Guest are viewing this topic.

Offline knotlogic

  • Regular Contributor
  • *
  • Posts: 165
  • Country: sg
Mismatched bandwidth between scope and probe?
« on: September 25, 2015, 01:32:44 pm »
So AFIK, we normally use a probe that's matched to the bandwidth of our oscilloscope.  eg: 100 MHz probe with a 100 MHz scope.  And I'm not aware of any issues if the two are slightly different - the stock probes on a Keysight DSOX2000 are 150 MHz on the 70 and 100 MHz models.  But what happens if that difference gets larger?  eg 100 MHz probes on a 350/500 MHz scope or higher?  Or vice versa?  Would it just be a case of the acquisition bandwidth would be limited to the lowest component?

Would we measure a lower noise level when using a 100 MHz probe on a 500 MHz scope as compared to a 500/500 combination?  And are there any "gotchas" we'd need to be aware of?
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 8992
  • Country: de
Re: Mismatched bandwidth between scope and probe?
« Reply #1 on: September 25, 2015, 01:40:27 pm »
There is no need to match the bandwidth of probes and the scope. The slower of the two sets the frequency limit - so it might be economincal to use parts with similar bandwidth, but it depends on the type of probe:
Typical 1:10 probes are rather cheap up to about 100-200 MHz, so its no problem to use a 150 MHz probe with a 20 MHz scope. However differential probes are often extensive when fast. So a 1 MHz differential probe can be OK with a 100 MHz scope if its fast enough and you know that it's only 1 MHz.
The switchable probes 1:1 and 1:10 are usually much faster in 1:10 mode and seldom more than 10 MHz in 1:1 mode. 
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 21134
  • Country: nl
    • NCT Developments
Re: Mismatched bandwidth between scope and probe?
« Reply #2 on: September 25, 2015, 01:49:32 pm »
Actually there is a lot more to probing signals over 100MHz so it is better to read/learn about it. As a rule of thumb a generic passive probe is useless for signals of 100MHz even if it says 500MHz due to the capacitance at the tip.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 13562
  • Country: gb
    • Having fun doing more, with less
Re: Mismatched bandwidth between scope and probe?
« Reply #3 on: September 25, 2015, 03:03:08 pm »
It is good to see someone thinking, recognising what they don't know, and asking questions.

Would it just be a case of the acquisition bandwidth would be limited to the lowest component?

No. Risetime is square-root-of-sum-of-squares.

Quote
Would we measure a lower noise level when using a 100 MHz probe on a 500 MHz scope as compared to a 500/500 combination?  And are there any "gotchas" we'd need to be aware of?

Many many gotchas.

There are useful references w.r.t. the theory and practice of probing, plus pointers to homebrew probes at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's 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 MarkF

  • Super Contributor
  • ***
  • Posts: 1978
  • Country: us
Re: Mismatched bandwidth between scope and probe?
« Reply #4 on: September 25, 2015, 03:09:00 pm »
The issue that arises is that the probe compensation adjustment will not cover the input capacitance of the scope.  Therefore you would not be able to properly adjust the probe.  You may want to review this video.

« Last Edit: September 25, 2015, 03:21:46 pm by MarkF »
 

Offline knotlogic

  • Regular Contributor
  • *
  • Posts: 165
  • Country: sg
Re: Mismatched bandwidth between scope and probe?
« Reply #5 on: September 26, 2015, 09:17:46 am »
Actually there is a lot more to probing signals over 100MHz so it is better to read/learn about it. As a rule of thumb a generic passive probe is useless for signals of 100MHz even if it says 500MHz due to the capacitance at the tip.

Well, I do understand that as frequencies go up, we can't think of the signals as just voltages but as dealing with RF and transmission lines.  I'm less certain where that crossover point is however.
It is good to see someone thinking, recognising what they don't know, and asking questions.

Would it just be a case of the acquisition bandwidth would be limited to the lowest component?

No. Risetime is square-root-of-sum-of-squares.

Quote
Would we measure a lower noise level when using a 100 MHz probe on a 500 MHz scope as compared to a 500/500 combination?  And are there any "gotchas" we'd need to be aware of?

Many many gotchas.

There are useful references w.r.t. the theory and practice of probing, plus pointers to homebrew probes at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Thanks, that will make for some interesting reading.  How does the risetime relate to system bandwidth?  (Specifics I mean)  I would have thought that if a probe is rated as 100 MHz, then it effectively acts as a 100 MHz lowpass filter at the scope input.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 21455
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Mismatched bandwidth between scope and probe?
« Reply #6 on: September 26, 2015, 10:02:11 am »
Well, I do understand that as frequencies go up, we can't think of the signals as just voltages but as dealing with RF and transmission lines.  I'm less certain where that crossover point is however.
Understand EVERY measurement system has an effect on the DUT and the measurement itself.

At low frequencies there is normally sufficient signal power/drive that the DUT and the measured waveform are barely affected IF the measurement system is matched. There is NO cross-over point as you put it, every incidence is different and is assesed on a case by case senario.

The measurement system is as accurate as it's weakest link. This maybe resistive or capacitive loading or BW roll off beyond the industry standard -3 dB as is likely to happen by using probes beyond their rated BW.
« Last Edit: September 26, 2015, 10:05:02 am by tautech »
Avid Rabid Hobbyist
 

Online EEVblog

  • Administrator
  • *****
  • Posts: 32913
  • Country: au
    • EEVblog
Re: Mismatched bandwidth between scope and probe?
« Reply #7 on: September 26, 2015, 10:38:07 am »
Basically the higher the bandwidth probe the better.
Although you have to watch out for the trimming capacitance range. A high bandwidth probe that is grossly mismatched to the scope in terms in input capacitance may not be able to be trimmed into compensation.
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 13562
  • Country: gb
    • Having fun doing more, with less
Re: Mismatched bandwidth between scope and probe?
« Reply #8 on: September 26, 2015, 04:19:01 pm »
Actually there is a lot more to probing signals over 100MHz so it is better to read/learn about it. As a rule of thumb a generic passive probe is useless for signals of 100MHz even if it says 500MHz due to the capacitance at the tip.
Well, I do understand that as frequencies go up, we can't think of the signals as just voltages but as dealing with RF and transmission lines.  I'm less certain where that crossover point is however.

nctnico's point is that a 10Mohm *10 probe isn't 10Mohm at 500MHz: the load is dominated by the ~10pF tip capacitance. Do the arithmetic, and then observe that a 500ohm "low impedance" Z0 probe will load the circuit far less.

As for the crossover point: the higher the frequency, the shorter the line at which reflections become important. The exact length depends on the degree of uncertainty you are prepared to tolerate.

Quote
It is good to see someone thinking, recognising what they don't know, and asking questions.

Would it just be a case of the acquisition bandwidth would be limited to the lowest component?

No. Risetime is square-root-of-sum-of-squares.

Quote
Would we measure a lower noise level when using a 100 MHz probe on a 500 MHz scope as compared to a 500/500 combination?  And are there any "gotchas" we'd need to be aware of?

Many many gotchas.

There are useful references w.r.t. the theory and practice of probing, plus pointers to homebrew probes at https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/

Thanks, that will make for some interesting reading.  How does the risetime relate to system bandwidth?  (Specifics I mean)  I would have thought that if a probe is rated as 100 MHz, then it effectively acts as a 100 MHz lowpass filter at the scope input.

Read the refs, and that, and many other questions will be answered - and you will learn far more easily from good source material than from any half-cocked responses on a bulletin board.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's 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 briselec

  • Regular Contributor
  • *
  • Posts: 94
  • Country: au
Re: Mismatched bandwidth between scope and probe?
« Reply #9 on: September 26, 2015, 09:20:48 pm »
Argh - I saw an equation for this recently in a white paper I was going to read later but now I can't find it.
Isn't it the case that a 100Mhz probe with a 100Mhz scope does not make a 100Mhz instrument. The actual bandwidth is 70Mhz because the actual bandwidth is the inverse of the square root of the sum of the inverse of the square of the scope bandwidth plus the inverse of the square of the probe bandwidth.
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 13562
  • Country: gb
    • Having fun doing more, with less
Re: Mismatched bandwidth between scope and probe?
« Reply #10 on: September 26, 2015, 10:25:53 pm »
Argh - I saw an equation for this recently in a white paper I was going to read later but now I can't find it.
Isn't it the case that a 100Mhz probe with a 100Mhz scope does not make a 100Mhz instrument. The actual bandwidth is 70Mhz because the actual bandwidth is the inverse of the square root of the sum of the inverse of the square of the scope bandwidth plus the inverse of the square of the probe bandwidth.
Essentially yes.

Look at the refs in https://entertaininghacks.wordpress.com/library-2/scope-probe-reference-material/ for a whole lot of essential background to scopes and probes.
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's 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 nbritton

  • Frequent Contributor
  • **
  • Posts: 443
  • Country: us
Re: Mismatched bandwidth between scope and probe?
« Reply #11 on: September 27, 2015, 01:40:47 am »
So AFIK, we normally use a probe that's matched to the bandwidth of our oscilloscope.  eg: 100 MHz probe with a 100 MHz scope.

From what I understand about system bandwidth you actually want to use a probe that has a higher bandwidth rating than the oscilloscope.

There is a math formula used to calculate system bandwidth:

1/sqrt((1/scope bandwidth)^2+(1/probe bandwidth)^2) = system bandwidth

Thus a 100 MHz scope with a 100 MHz probe has a system bandwidth of 70.7 MHz

As the probe's rated bandwidth approaches infinite bandwidth you will get closer and closer to the unadulterated 100 MHz rated spec of the scope. This is a convoluted way of saying that you should use the fastest probes you can afford, or upgrade to a higher bandwidth scope, ether solution will work to increased the total system bandwidth.
 

Offline nbritton

  • Frequent Contributor
  • **
  • Posts: 443
  • Country: us
Re: Mismatched bandwidth between scope and probe?
« Reply #12 on: September 27, 2015, 01:49:58 am »
Dave talks about system bandwidth here:

 

Offline knotlogic

  • Regular Contributor
  • *
  • Posts: 165
  • Country: sg
Re: Mismatched bandwidth between scope and probe?
« Reply #13 on: September 27, 2015, 01:55:10 pm »
I would have thought that if a probe is rated as 100 MHz, then it effectively acts as a 100 MHz lowpass filter at the scope input.

So I think I know where I'm wrong with that reasoning, and why it doesn't fit with the system bandwidth formula...  Because (in general) saying that a probe is rated at 100 MHz might mean that it's -3 dB down at 100 MHz.  If the manufacturer didn't "derate" the probe.  Feel free to correct me if I'm going down the wrong path.


So AFIK, we normally use a probe that's matched to the bandwidth of our oscilloscope.  eg: 100 MHz probe with a 100 MHz scope.

From what I understand about system bandwidth you actually want to use a probe that has a higher bandwidth rating than the oscilloscope.

There is a math formula used to calculate system bandwidth:

1/sqrt((1/scope bandwidth)^2+(1/probe bandwidth)^2) = system bandwidth

Thus a 100 MHz scope with a 100 MHz probe has a system bandwidth of 70.7 MHz

As the probe's rated bandwidth approaches infinite bandwidth you will get closer and closer to the unadulterated 100 MHz rated spec of the scope. This is a convoluted way of saying that you should use the fastest probes you can afford, or upgrade to a higher bandwidth scope, ether solution will work to increased the total system bandwidth.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf