Author Topic: Siglent SDS1202X-E ground bounce between channels  (Read 1748 times)

0 Members and 1 Guest are viewing this topic.

Offline calvinglosterTopic starter

  • Contributor
  • Posts: 30
  • Country: za
Siglent SDS1202X-E ground bounce between channels
« on: June 20, 2023, 12:36:30 pm »
I am using a microcontroller to get an ADC reading from a 16 bit ADC (LTC2326). The SPI clock speed is set to 500KHZ and im getting massive ringing on the clock, MOSI and MISO lines. this ringing seems to be showing up on channel two which is what I'm using to trigger on. I have added 50 ohm resistors in series to try damp the lines but the ringing and crosstalk is still showing up quite significantly. Is there something wrong with my scope? I tried using a 10X probe on channel two and the crosstalk gets worse.

The setup is on a breadboard and I've tried to keep channel 1 and channel 2 leads separated as much as I can. Ive added a screenshot.

Could there be something wrong with my scope?

« Last Edit: June 21, 2023, 01:44:30 am by calvingloster »
 

Offline Grandchuck

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: us
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #1 on: June 20, 2023, 02:04:33 pm »
Try shorter oscilloscope ground leads.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7463
  • Country: hr
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #2 on: June 20, 2023, 02:29:24 pm »
Clock frequency (repetition rate) of 500kHz makes no difference.. That chip can use 100 MHz SPI clock.. Which means it will have sub 1ns edges... Those are seriously fast edges..
Like Grandchuck say, use very short probe wires (use direct spring connection if you can).

Title is wrong.. What you see is not crosstalk between channels but ground bounce...

Crosstalk between channels is what you get if you connect a signal to one ch, and nothing at all to other, and then try to see if you see anything where there is nothing connected..

Ground bounce is artefact of PCB layout and probing technique...

So I would suggest to change misleading (wrong) title..
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 
The following users thanked this post: Performa01, tooki

Offline calvinglosterTopic starter

  • Contributor
  • Posts: 30
  • Country: za
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #3 on: June 21, 2023, 01:43:52 am »
Thanks for the reply much appreciated. So is this an artifact of measurement or is it inherently in the circuit itself? If it's only showing up on the ossciloscope because of leads and it's not actually causing havoc in the circuit then I don't mind.
I just need the circuit to be as noise free as possible.

The ADC will be on an analog board recieving a differential signal from a femto bias current op amp thats measuring tunneling current of a Scanning tunneling microscope so having this bounce is def going to slow the scanning speed down somewhat if it is actually something inherent to the circuit. Is this ADC not suitable because it has such fast rise and fall time?
« Last Edit: June 21, 2023, 01:49:03 am by calvingloster »
 
The following users thanked this post: 2N3055

Offline BillyO

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: ca
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #4 on: June 21, 2023, 03:08:42 am »
I am using a microcontroller to get an ADC reading from a 16 bit ADC (LTC2326). The SPI clock speed is set to 500KHZ and im getting massive ringing on the clock, MOSI and MISO lines. this ringing seems to be showing up on channel two which is what I'm using to trigger on. I have added 50 ohm resistors in series to try damp the lines but the ringing and crosstalk is still showing up quite significantly. Is there something wrong with my scope? I tried using a 10X probe on channel two and the crosstalk gets worse.

The setup is on a breadboard and I've tried to keep channel 1 and channel 2 leads separated as much as I can. Ive added a screenshot.

Could there be something wrong with my scope?

This may not be a great image of what you are talking about, but I don't see anything of concern in it.
Bill  (Currently a Siglent fanboy)
--------------------------------------------------
 

Offline Performa01

  • Super Contributor
  • ***
  • Posts: 1729
  • Country: at
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #5 on: June 21, 2023, 06:06:04 am »
So is this an artifact of measurement or is it inherently in the circuit itself? If it's only showing up on the ossciloscope because of leads and it's not actually causing havoc in the circuit then I don't mind.
Even though the ground leads will act as a loop antenna, picking up magnetic pulses stemming from high currents caused by fast edges charging parasitic capacitances in an unshielded circuit, this is not the reason for ground bouncing.

Ground bouncing (and other oddities) can be avoided by applying the usual rules for good circuit layout whenever high frequencies (= fast edges) are present: a multilayer PCB with (at least) one uninterrupted(!) ground plane.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 21225
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #6 on: June 21, 2023, 07:48:39 am »
Clock frequency (repetition rate) of 500kHz makes no difference.. That chip can use 100 MHz SPI clock.. Which means it will have sub 1ns edges... Those are seriously fast edges..
Like Grandchuck say, use very short probe wires (use direct spring connection if you can).

Title is wrong.. What you see is not crosstalk between channels but ground bounce...

Crosstalk between channels is what you get if you connect a signal to one ch, and nothing at all to other, and then try to see if you see anything where there is nothing connected..

Ground bounce is artefact of PCB layout and probing technique...

So I would suggest to change misleading (wrong) title..

Yes, pure and simple. In addition your breadboard is probably part of the problem.

For the basic physics and examples, see https://entertaininghacks.wordpress.com/2015/04/23/scope-probe-accessory-improves-signal-fidelity/ and https://entertaininghacks.wordpress.com/2018/05/08/digital-signal-integrity-and-bandwidth-signals-risetime-is-important-period-is-irrelevant/

Solderless breadboards inherently have serious limitations. For other prototyping techniques and products with fewer limitations, see https://entertaininghacks.wordpress.com/2020/07/22/prototyping-circuits-easy-cheap-fast-reliable-techniques/
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
 
The following users thanked this post: 2N3055

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7463
  • Country: hr
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #7 on: June 21, 2023, 08:18:20 am »
Thanks for the reply much appreciated. So is this an artifact of measurement or is it inherently in the circuit itself? If it's only showing up on the ossciloscope because of leads and it's not actually causing havoc in the circuit then I don't mind.
I just need the circuit to be as noise free as possible.

The ADC will be on an analog board recieving a differential signal from a femto bias current op amp thats measuring tunneling current of a Scanning tunneling microscope so having this bounce is def going to slow the scanning speed down somewhat if it is actually something inherent to the circuit. Is this ADC not suitable because it has such fast rise and fall time?

It probably is a bit of both. You need to have very meticulous probing to minimize it's effect to be able to see if there is real ground bounce in a circuit itself..

Performa01 is right and also tggzzz. Make sure to go to tggzzz web site and read the articles. He has some excellent content there...

And yes, breadboards are evil....  :-DD
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 21225
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #8 on: June 21, 2023, 10:07:57 am »
And yes, breadboards are evil....  :-DD

I will quote you on that :)
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
 
The following users thanked this post: 2N3055

Offline tv84

  • Super Contributor
  • ***
  • Posts: 3345
  • Country: pt
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #9 on: June 21, 2023, 10:16:38 am »
Solderless breadboards inherently have serious limitations. For other prototyping techniques and products with fewer limitations, see https://entertaininghacks.wordpress.com/2020/07/22/prototyping-circuits-easy-cheap-fast-reliable-techniques/

Curiously I don't see wire-wrapping in that page.  :-// Used it a lot in college some decades ago.
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 21225
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #10 on: June 21, 2023, 10:31:12 am »
Solderless breadboards inherently have serious limitations. For other prototyping techniques and products with fewer limitations, see https://entertaininghacks.wordpress.com/2020/07/22/prototyping-circuits-easy-cheap-fast-reliable-techniques/

Curiously I don't see wire-wrapping in that page.  :-// Used it a lot in college some decades ago.

I'll argue it was - and is[1] - a production method. No, that doesn't convince me either, since I have the wire and sockets and hand tool and even an example I threw together specifically to show the consequences of wires plus slow (LSTTL!) di/dt :)

It is useless for modern logic since every connection comes complete with a selection of stub filters :)

There's also the IDC version, which minimises the stub problem. It consists of a multilayer PCB fully populated with two-level IDC pins on 0.1" centres. Wirewrap wire is used to make the connections. I even used that once, 30 years ago. Hellish expensive PCBs, only affordable by companies.

I really ought to add them :)

[1] You can still find it buried inside some modern light bulbs, and possibly some equipment where you need to make reliable connections in the field.
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 tv84

  • Super Contributor
  • ***
  • Posts: 3345
  • Country: pt
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #11 on: June 21, 2023, 10:48:53 am »
I really ought to add them :)

 :-+ Just for "memory lane's sake".
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 7463
  • Country: hr
Re: Siglent SDS1202X-E Crosstalk between channels
« Reply #12 on: June 21, 2023, 11:51:08 am »
It is useless for modern logic since every connection comes complete with a selection of stub filters :)

LOL   :-+
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 

Offline BillyO

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: ca
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #13 on: June 21, 2023, 02:10:55 pm »
In my opinion the best approach to prototyping, beyond something dead simple (**), these days is just have PCBs built.  Even if it takes 2 or 3 iterations to get it "perfect" the cost in terms of money and effort is the lowest and if you get the hang of the technical side of it for the kind of circuit you are designing can (will) produce the best results.

** I recently used a solderless breadboard to throw together an astable multivirbator circuit working @ 1Hz to demonstrate you don't need an entire computer system and software development environment to flash an LED.  It was quick and got the job done with no worries  :-+
Bill  (Currently a Siglent fanboy)
--------------------------------------------------
 

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 21225
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #14 on: June 21, 2023, 02:28:56 pm »
In my opinion the best approach to prototyping, beyond something dead simple (**), these days is just have PCBs built.  Even if it takes 2 or 3 iterations to get it "perfect" the cost in terms of money and effort is the lowest and if you get the hang of the technical side of it for the kind of circuit you are designing can (will) produce the best results.

** I recently used a solderless breadboard to throw together an astable multivirbator circuit working @ 1Hz to demonstrate you don't need an entire computer system and software development environment to flash an LED.  It was quick and got the job done with no worries  :-+

There's truth to that, but I'll add a couple of counterpoints...

The cost with PCBs is measured in money and latency. Iterating can take a boringly long time.

Any technique can be misused, resulting in crap. Arguably a beginner is likely to produce a crap PCB due to inexperience. (Inexperience can be cured, unlike stupidity :) )

IMNSHO manhattan offers a good set of compromises between cost, time, performance, tweakability. There are other valuable techniques; use a good combination suited to the problem at hand.
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 BillyO

  • Super Contributor
  • ***
  • Posts: 1848
  • Country: ca
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #15 on: June 21, 2023, 02:39:41 pm »
IMNSHO manhattan offers a good set of compromises between cost, time, performance, tweakability. There are other valuable techniques; use a good combination suited to the problem at hand.
Manhattan produces good results if you lay things out properly - also a learning curve here. But I really don't have the patience to cut all those little pieces of clad and route out islands.  Plus, for some totally unknow reason, copper clad board has become unreasonably priced (6x6 single sided board from digi-key = $11).  Yeah, iterations in using the PCB method can take a couple of weeks, but you can use the time between to do other things.
Bill  (Currently a Siglent fanboy)
--------------------------------------------------
 

Offline Sighound36

  • Frequent Contributor
  • **
  • Posts: 549
  • Country: gb
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #16 on: June 22, 2023, 07:55:58 am »
Arh yes, spurious images or WTF I am getting this reading? measurement certainty & repeatability, probing techniques and effective bandwidth of the total measurement system even correct VTB & HTB all basic stuff, but at one time or another I suspect it has caught all of us out to a greater or lesser degree. Myself on more than one occasion and it's easy to overlook if your just starting out with electronics with your first scope. Been there done that, have the T shirt, video rights etc  :-DD

Just be aware that the flying lead on a basic divide by 10 probe isn't always your friend :o and yet it is possibly the most used probe in the scope accessory box. It’s the most practical and easy to use as well.

Just make sure you have calibrated the scope via the on board cal ports and get that square wave looking well square lol, check both scope & probe bandwidth remember have a 500Meg scope and 200Meg probes may not be ideal for looking @ 275Meg signals.

As the chaps have pointed out you do have the 'benefit' of a free small loop antenna with your passive probes, great for reaching to those reference points no obvious but farther away from the  desired real world measurement result. If possible, use the spring clip on the probe tip it will reduce errors and improve repeatability of the measurement. Its a simple tip but its imho an essential one.

I'm lucky we have many types and ranges of probes so yes, I can obtain a more accurate repeatable result, however each of those probes has their own set of rules and calibration that if not followed will deliver false results to.

I built a couple of test boards for the apprentices we have now working with us, like a scope demo board but with more going on. One of the exercises is measuring a specific rise time and voltage with three different probes plus one specific exercise with and without the flying lead attached to a PP023 passive.  The number of questions asked as to why is it so different?

It’s just awareness of the measurement situation you find yourself in, it’s just experience sometimes give you that edge. Otherwise sharing your results with a likeminded colleague who you can bounce ideas off is also a great way to crack that nut.

Ground bounce, thats another can of worms at the start, another day topic I feel  :)


« Last Edit: June 22, 2023, 11:26:20 am by Sighound36 »
Seeking quality measurement equipment at realistic cost with proper service backup. If you pay peanuts you employ monkeys.
 
The following users thanked this post: 2N3055

Offline tggzzz

  • Super Contributor
  • ***
  • Posts: 21225
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Re: Siglent SDS1202X-E ground bounce between channels
« Reply #17 on: June 22, 2023, 08:42:24 am »
It’s just awareness of the measurement situation you find yourself in, it’s just experience sometimes give you that edge. Otherwise sharing your results with a likeminded colleague who you can bounce ideas off is also a great way to crack that nut.

Ground bounce, thats another can of worms at the start, another day topic I feel  :)

Yup.

Those of us that grew up using 20kohm/V voltmeters instinctively know that the measuring tool changes the signal being measured.

As for ground bounce, the "equivalent" from the previous decade was flip-flop metastability. While metastability is irrelevant to this topic, ground bounce is very relevant to anything built using a solderless breadboard.
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
 
The following users thanked this post: 2N3055


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf