Author Topic: What oscilloscope bandwidth do i need for measuring Vp-p(ripple, noise)?  (Read 4941 times)

0 Members and 1 Guest are viewing this topic.

Offline spfordTopic starter

  • Newbie
  • Posts: 6
  • Country: kr
Hello.
At first, My english skill is very low. Understand please.
I am a newbie about electronic circuits, oscilloscopes.
This measure is  hobby at home.

## Question 1
What oscilloscope bandwidth do i need for measuring Vp-p(ripple, noise) of a 5V USB charger with a switching frequency of 500 Khz?
1. Only use the 'five times rule' -> 500 Khz x 5 = 2,500 Khz?
2. Use the 'five times rule' + Consider harmonic components -> 500 Khz x 5 x 5 = 12,500 Khz?

Which is correct at bewteen 1 and 2? or other way?


Note.
In the links below, they measure both low and high frequencies.
http://www.righto.com/2012/10/a-dozen-usb-chargers-in-lab-apple-is.html
 
## Question 2
What is the most appropriate for me on the four Mini Oscilloscopes?



------------------- List of four products (low price to high price) -------------------- ------

###### 1 - JYE TECH DSO 150    (Price $ 26, Resolution: 12 bit, Bandwidth: 200 Khz, 5 mV / Div, Sampling: 1 Msps)
https://www.banggood.com/Original-JYE-Tech-Assembled-DSO-SHELL-DSO150-Digital-Oscilloscope-Module-9V-p-1211151.html?rmmds=search&cur_warehouse=CN

 
###### 2 - DSO FNIRSI PRO    ($ 40, 8 bit, 5 Mhz, 50 mV / Div, 20 Msps)
https://www.aliexpress.com/item/DSO-FNIRSI-PRO-Handheld-mini-portable-digital-oscilloscope-5M-bandwidth-20MSps-sampling-rate/32973233356.html


###### 3 - MINI DS212    ($ 85,: 8 bits, 1 Mhz, 20 mV / Div, 10 MSa / s)
https://www.banggood.com/MINI-DS212-Digital-Storage-Oscilloscope-Portable-Nano-Handheld-Bandwidth-1MHz-Sampling-Rate-10MSas-p-1202288.html?rmmds=search&cur_warehouse=CN
 

###### 4 - HANTEK 2C42    ( $ 102, 8 bit, 40 Mhz, 10 mV / Div, 250 MSa/s)
https://www.aliexpress.com/item/Hantek-3in1-3-in-1-2c42-2d42-2c72-2d72-Digital-handheld-Oscilloscope-Portable-40mhz-70mhz-bandwidth/32953177457.html

-------------------------------------------------- -------------------------------------------------- ---
Thanks for reading.
I've been browsing the internet, Youtube for a few days, but I do not know if I know what I'm sure of.
so I would like to get help from experts here.
I would appreciate for your a reply. :)
« Last Edit: May 12, 2019, 04:24:09 pm by spford »
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8175
  • Country: us
I don't know about the "5x" rule, but it can be hard to determine ahead of time what sort of noise may appear in such a device.  A lowpass filter that is effective at the fundamental will be even more effective at the higher harmonics, so I'd expect any ripple to be primarily at the switching frequency--or below it due to interaction with line frequency or other internal processes. 

Of the 4 scopes you listed, only the Hantek one is up to your task, IMO.  The first has clearly insufficient bandwidth, the second has low sensitivity and the third is marginal at best on both bandwidth and sensitivity.  If you look at the Hantek's detailed specs,  you'll see that it has 6 MHz bandwidth with the 1X probe.  In order to get the the 10mV/div sensitivity (which isn't all that great either, but the best of the lot), you need to use 1X probing and live with the 6MHz bandwidth--which seems sufficient in this case.

I don't know your circumstances, but $100+ for a very modest (at best) scope seems a poor investment when pretty good 2-channel DSOs are available in the $250 range.  Just a thought!  :)
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: spford

Offline radiolistener

  • Super Contributor
  • ***
  • Posts: 4135
  • Country: 00
You're need to multiply on 10 for pulse and square waves.

20-40 MHz bandwidth will be enough for your purpose. But for your task sensitivity requirement is more critical than bandwidth.
I'm very skeptic about these links, which you're provided.

I think first three are just a toys and will be completely useless for your task. You will not be able to see voltage ripple with it.

Don't know about HANTEK, but according to specification, it's max resolution is just 10 mV/div.
Such device will be definitely useless for your task. Because it cannot show you ripple which usually is about 1-5 mV.

Look to this review, it seems that this is also just a buggy toy, but more expensive than other.




I don't recommend to buy any of these toys, because you will not be able to measure power supply ripple with it.

I can recommend Siglent SDS1102X-E, it has almost enough sensitivity for your task.
« Last Edit: May 12, 2019, 03:51:16 pm by radiolistener »
 
The following users thanked this post: spford

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Basically you need to be able to see all significant harmonics of the wave.  If you are concerned about rise time, you need more bandwidth than if you just want to verify operation.  As someone said, about 10 times the fundamental frequency is a rule of thumb but even that is open to question, depending on how important the fidelity is.

Since you want to measure noise, you need to make sure you can see all components of the signal that matter to you.  You don't need to worry about frequencies into the microwave region unless you have equipment sensitive to those frequencies.  So measuring ripple and stating a voltage value isn't as important as deciding how much you can tolerate.  A very fast wave will contain high frequency components regardless of its fundamental frequency.

So a 1 ns rise will require a wider bandwidth than a 100 ns rise, even if the wave is low frequency.  In other words, there is no simple answer and each case is unique.
 
The following users thanked this post: spford

Offline Adrian_Arg.

  • Frequent Contributor
  • **
  • Posts: 450
  • Country: ar
for hobby I recommend ahttps://www.ebay.com/itm/Hantek-DSO5102P-Digital-Storage-Oscilloscope-100MHz-1Gsa-S-2-CH-7-TFT-USB/292256381782?epid=2240906425&hash=item440bd67b56:g:KYkAAOSwR0xZuQg4,It's economic, a little higher but a much higher Rigol DS1054z, and of the last ones that have come out but of much higher price Sglent sds1204x-e
 
The following users thanked this post: spford

Offline bsfeechannel

  • Super Contributor
  • ***
  • Posts: 1668
  • Country: 00
The answer to the title of this thread is: as much as your money can buy. Do not skimp on bandwidth.

Watch this brilliant video by our fellow forum member w2aew to understand why. Although the title says "analog oscilloscope", the rationale is still valid for digital scopes.



And as if this is not enough, he published a version for digital scopes, too.



Hope that helps.
 

Offline innkeeper

  • Supporter
  • ****
  • Posts: 615
  • Country: us
Output Ripple and Noise is usually specified with a 20MHz Bandwidth.

many higher bandwidth scopes have a bandwidth limit function for this.

all that being said, it really depends on your need for knowing what noise is there, there are specs, and then there is real life.
Hobbyist and a retired engineer and possibly a test equipment addict, though, searching for the equipment to test for that.
 
The following users thanked this post: spford

Offline spfordTopic starter

  • Newbie
  • Posts: 6
  • Country: kr
bdunham7, radiolistener, bob91343, Adrian_Arg., bsfeechannel, innkeeper, blueskull

Thank you for the reply with many thought,  advice, oscilloscope recommend and helpul video links.



My background of this question is....
Several reviewer measured Apple 5W USB charger's Vp-p value, the value was 26mV ~ 66 mV.
I wanted to measure the Vp-p of my several chargers like any other review with a low investment cost.


Fortunately, I had a multimeter.
I measured the voltage of the Apple 5W USB charger in AC mode with my Fluke 177 multimeter.
but the result was only 1.6 mV (at no load) and 1.8 mV (at 1A load).
I could not figure out the Vp-p value. Lol



So I asked the experts here for help.
In fact, I thought the third product ($40, BW 5Mhz) was a measurable device.(It may not be very accurate like bench oscilloscope.)
I was going to buy a third product.



A few expert recommend a oscilloscope with $250 or more.  i also want to buy high grade oscilloscope.
But in reality ...

After looking at several answers, i found that measuring the Vp-p value was not easy(whether it's a price and a way).
I need more study, thought and review.

Thanks again for taking the time to answer.  ^-^
« Last Edit: May 14, 2019, 01:41:19 am by spford »
 

Offline bob91343

  • Super Contributor
  • ***
  • Posts: 2675
  • Country: us
Why do you want to measure the p-p voltage?  The value isn't a lot of use, other than to see if something has failed.

Ripple current in a charger isn't usually a problem.  If the instantaneous voltage is above the battery voltage, it will charge.  Sometimes it's actually a benefit, allowing slightly greater charge due to the higher momentary levels.

On the other hand, if you are just curious or trying to learn how to make measurements, I can understand that.
 
The following users thanked this post: spford

Offline innkeeper

  • Supporter
  • ****
  • Posts: 615
  • Country: us
I'm torn on this... on the one hand I feel if you can get your hands on a used scope at a decent price you will have a fairly usable piece of test gear that will do what you want now and give you a tool to learn on.

if you are going to measure power supply noise, its helpful to get one of the scopes that can go down to 500uv, and at least 20mhz of bandwidth.  there is a thread on what scopes can do that. https://www.eevblog.com/forum/beginners/500uvdiv-oscilloscopes/  buying a used CRT based scope is really a good option. Especially if you are on a budget.

on the other hand If you are willing to put some more money into this, at 499 usd Siglent SDS1104X-E 4-Ch 100MHz and at 379 usd Siglent SDS1202X-E 2-Ch 200MHz are stand-alone digital scopes but can also be hooked to your computer that offers good value for the money and are my current personal prefreces

both of those scopes offer some very advanced features which you might not appreciate now but will if you continue and grow in your knowledge you will quickly make use of.   for now, one of the important features is it has true 500 μV / div sensitivity (not software enhanced which is important when looking at noise)

These scopes will grow with you with most needed features already in there out of the box and the ability to add more if needed.  You get one of those, you will likely not be wanting for a different scope for a long time.

i used to be a BIG advocate of learning oscilloscopes with a CRT based scope because it illustrates exactly what is going on with the controls on your digital scope, it helps with the understanding of it.  but in this age of smartphones, starting with a digital scope like the siglent is not a bad place to start as it can perform many functions that you would traditionally need additional gear for, voltage meter, frequency counter, rudimentary spectrum analysis, protocol decoding of digital data. and you can add things like signal generation for sweep functions and various waveform generation, and the ability to add on logic analyzer functionality. This makes for a very compact workspace

for your use ... lets say you look at the power supply and you see noise. and you want to dig deeper and see what frequencies make up that nose. you can pass that thought the scopes build in FFT and see the frequencies and amplitudes that make up that noise. pretty sweet. you can't do that on an old CRT based scope!

i think your smart taking a moment and researching some more ... and i'm sure some of the terminologies is new for you, and that's fine. It gives you some things to research :)  Most of us have bought equipment/tools that we outgrow almost immediately or overspending on things we just don't need..

hope our experiences/insights help you make a good choice for you.. but start someplace, it is better than not starting at all :)

have fun!
« Last Edit: May 13, 2019, 09:24:28 pm by innkeeper »
Hobbyist and a retired engineer and possibly a test equipment addict, though, searching for the equipment to test for that.
 
The following users thanked this post: spford

Offline spfordTopic starter

  • Newbie
  • Posts: 6
  • Country: kr
### To bob91343

1. The reason to measure a Vp-p value is.....

I think the quality of the power supply or charger can be judged by the Vp-p value.(is it correct?)
i think that i can decide whether i can trust it and whether will keep use it by the Vp-p value.

If the Vp-p value is high, i can not trust it and i do not like to use it and maybe i will throw it into the trash.(various power supply and chargers around me at home)

So i wanted to know a Vp-p value.


2. "curious or trying to learn how to make measurements", You are right.
This measurement is based on curiosity that I want to measure directly and judge whether it is good or bad.
Maybe i have a little engineering character that I want to measure when i look at something. Lol

Thanks for the reply. :)




### To innkeeper

1. I agree. Used oscilloscopes are also a great opportunity for me. but CRT based scope is...i need more searching.

2.  Also I want to buy a Siglent SDS1102CML+ or SDS1104X-E or DS1054Z ETC. and I like use advanced features.
But maybe I will only use one or two features(eg, Vp-p and FFT). I do not want overspending.
Of course, if this measurement was my job, I would have invested a good oscilloscope without hesitation recommended by experts.
and if i was very rich....

3. I will remember "500uV/Div and at least 20Mhz". thanks
Thank you for sharing your experience and insight.
it was very helpful for me.


bob91343 and innkeeper,
Thanks for taking the time. :)

« Last Edit: May 15, 2019, 02:47:10 am by spford »
 

Offline spfordTopic starter

  • Newbie
  • Posts: 6
  • Country: kr
I am searching about OWON VDS1022i. (USB PC OSCILLOSCOPE,  about $110,  BW 25Mhz,  100MSa/s,  5mV/Div,  5ns/div )

It is a bit inconvenient to use because of PC booting, line connection, etc,

But it seems to be a good compromise for me. :)
« Last Edit: May 18, 2019, 10:34:50 am by spford »
 

Offline innkeeper

  • Supporter
  • ****
  • Posts: 615
  • Country: us
Hobbyist and a retired engineer and possibly a test equipment addict, though, searching for the equipment to test for that.
 
The following users thanked this post: spford

Offline spfordTopic starter

  • Newbie
  • Posts: 6
  • Country: kr
1. Information.
I found that VDS1022i also has limited 5 MHz bandwidth with the 1X probe(5mv/Div).

[VDS1022i's Manual]--------------------------------------------------------------------------------------------------------
"Caution: When the attenuation switch is set to 1X, the probe will limit the bandwidth of the oscilloscope in 5MHz.
To use the full(25Mhz) bandwidth of the oscilloscope, the switch must be set to 10X."
-------------------------------------------------------------------------------------------------------------------------------


2. Question.
innkeeper said "usually noise measures 20MHz bandwidth."

Do I have to measure with 10x prob mode and 50mV / Div sensitivity?
Is it possible to measure the noise properly if the sensitivity is 50mV / Div?

Or Is it possible to measure the noise properly even if the bandwidth is limited 5 Mhz with 1X probe mode?

I would appreciate for your a reply. :)
« Last Edit: May 21, 2019, 01:06:32 pm by spford »
 
The following users thanked this post: sx57

Offline innkeeper

  • Supporter
  • ****
  • Posts: 615
  • Country: us
as you might suspect, Dave has done videos on both of these topics.

watch them both, you will learn a lot

but to answer your direct questions.

yes, a x1 probe has lower bandwidth then a x10 probe (or for dual x1/x10 probes you will get higher bandwith at x10)
preferably you want to use a x1 probe with 20mhz bandwidth, to get the best fidelity of the noise measured.
various probes are different and you have to look at the specs to see if it will go to 20mhz at x1.
However, the best type of probe for noise measurement is a differential type probe instead of your standard single ended probe. However, those kinds of probes are expensive.
there is a trick to use 2 channels of your scope as a poor man's differential probe - also described in the videos below u using 50 ohm coax and 50 ohm terminations.

yes, you can measure noise and ripple with lower bandwidth then 20mhz, however, it is just the defacto standard used for such measurements and if you don't use it you will not be able to compare your measurements to other documented specifications etc.

No, your likely not going to be able to measure noise at 50mv/div. unless the PS is very bad... you want to be able to measure 1mv/div or better.

lastly... there is a pitfall with probe bandwidth you should be aware of. a probe bandwidth is a frequency in which the signal drops by 3db, or about 30% lower than the real reading. in order to avoid this, you should use a probe with 3x or more bandwidth than the highest frequency your trying to measure.  there is a good video from Tektronix on this I will also post below. the reason for 3x bandwidth is that the 3db point is for pure sinewave, whereas you need to account for high rise time signals for non-sinewave elements.

this is all explained in the videos below




« Last Edit: May 26, 2019, 11:56:24 pm by innkeeper »
Hobbyist and a retired engineer and possibly a test equipment addict, though, searching for the equipment to test for that.
 
The following users thanked this post: spford, sx57


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf