Author Topic: "Grounding" according to Analog Devices  (Read 3414 times)

0 Members and 2 Guests are viewing this topic.

Online SiliconWizardTopic starter

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
"Grounding" according to Analog Devices
« on: February 17, 2023, 02:19:03 am »
https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html

(Well, I don't know how the author is affiliated with AD though.)

A lot of bad advice and a few reasonable points. Curious to hear about your thoughts.
 

Offline KaneTW

  • Frequent Contributor
  • **
  • Posts: 811
  • Country: de
Re: "Grounding" according to Analog Devices
« Reply #1 on: February 17, 2023, 02:46:45 am »
I saw this ages ago. IMO it just doesn't hold the test of time anymore.

Keep all your ground well connected. Be aware of your return currents and how/where they flow. Enclose sensitive and noisy signals between ground planes if you can.
 
The following users thanked this post: wraper

Online Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: "Grounding" according to Analog Devices
« Reply #2 on: February 17, 2023, 03:02:48 am »
(Well, I don't know how the author is affiliated with AD though.)
The author's Bio is provided at the bottom of the web page.
Facebook-free life and Rigol-free shack.
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
 
The following users thanked this post: RJSV, AlbertL

Offline sleemanj

  • Super Contributor
  • ***
  • Posts: 3051
  • Country: nz
  • Professional tightwad.
    • The electronics hobby components I sell.
Re: "Grounding" according to Analog Devices
« Reply #4 on: February 17, 2023, 03:34:35 am »
Cant get past the stupid cookie consent.  Why do they do this....

Because it's a legal requirement in europe. 

It's IMHO possibly the stupidest legal requirement that Europe has ever made, because it's just an annoyance and everybody just clicks ok anyway without even considering reading it.  All it does is train people to click the obvious button to see the content.
~~~
EEVBlog Members - get yourself 10% discount off all my electronic components for sale just use the Buy Direct links and use Coupon Code "eevblog" during checkout.  Shipping from New Zealand, international orders welcome :-)
 
The following users thanked this post: tom66, bookaboo, Siwastaja, CatalinaWOW, tooki, newbrain, AlbertL

Online Psi

  • Super Contributor
  • ***
  • Posts: 10385
  • Country: nz
Re: "Grounding" according to Analog Devices
« Reply #5 on: February 17, 2023, 03:42:48 am »
https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html

Cant get past the stupid cookie consent. 

Did you try actually accepting the cookie?
Or are you doing a 'homer' and holding onto the can in the vending machine while complaining you're arm is stuck?
Greek letter 'Psi' (not Pounds per Square Inch)
 

Online SiliconWizardTopic starter

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: "Grounding" according to Analog Devices
« Reply #6 on: February 17, 2023, 03:50:01 am »
I saw this ages ago. IMO it just doesn't hold the test of time anymore.

Keep all your ground well connected. Be aware of your return currents and how/where they flow. Enclose sensitive and noisy signals between ground planes if you can.

Yep. The slits in ground planes really were too much. :(
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7681
  • Country: ca
Re: "Grounding" according to Analog Devices
« Reply #7 on: February 17, 2023, 03:56:14 am »
There are no simple one sentence answers about grounding. It's not that simple and takes many years of practical experience to know what you are doing.
Entire books are written about the subject. I don't say much to anyone who asks about it (here) because I'm not offering a course on it. There is a lot of wisdom needed and every year digital circuitry goes up in speed.

The AD article looks at a few specific cases, gives orders on what to do... I don't agree with all of it but at least some equivalent circuits are presented. He ignores transmission line effects which I think is a mistake. I say 'proof is in the pudding' - look at Analog Devices eval boards, how they do the PCB layout. Not that they are always well done but apparently they achieve the IC's specifications. You'll soon see lately the approach is just to use more PCB layers and panic ground everything.

Because there is a disconnect between the PCB layout person and the IC engineering team and PCB layout is always a last minute rush at the end of that kind of project, it is not always done the best. When it fails EMC then people pay attention lol. But AD rarely makes product designs that far, so how can their people even know their approach is bad?
 
The following users thanked this post: Someone

Online Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: "Grounding" according to Analog Devices
« Reply #8 on: February 17, 2023, 04:22:49 am »
10 years ago when I was speaking to an AD Lab engineer he said they use solid ground plane for characterization of their ICs and design of evaluation boards.
Facebook-free life and Rigol-free shack.
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 11340
  • Country: us
  • $
Re: "Grounding" according to Analog Devices
« Reply #9 on: February 17, 2023, 05:13:43 am »
there is ground advice

then you run into a real life USB implementation.


does anyone not care anymore? Like its not worth the stomach ache because you can... respin the board if there is a problem. I don't care about these insane grand unified theories of construction anymore.


Here is a visualization exersize. Imagine for thermal and cost reasons you have to put a IC clamped to the side of a chassis that has hand soldered hook up wire attaching it to the PCB. You did it, and it got through the test lab. There are no real problems, and someone told the stickler to shut up because it made enough money. It won't be getting re designed, and the customer is happy.
« Last Edit: February 17, 2023, 05:22:22 am by coppercone2 »
 

Offline mulcahla

  • Newbie
  • Posts: 1
  • Country: us
Re: "Grounding" according to Analog Devices
« Reply #10 on: February 17, 2023, 02:57:42 pm »
https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html

(Well, I don't know how the author is affiliated with AD though.)

A lot of bad advice and a few reasonable points. Curious to hear about your thoughts.

Which sections are bad advice in your opinion? I have been designing 40 input, 1.0 uV-stable ADC circuits for thermocouple measurements with their IC's (AD7793) and following their design guidelines. Still looking for ways to improve this circuit if you could offer better advice.
 

Offline shapirus

  • Super Contributor
  • ***
  • Posts: 1721
  • Country: ua
Re: "Grounding" according to Analog Devices
« Reply #11 on: February 17, 2023, 03:52:37 pm »
I saw this ages ago. IMO it just doesn't hold the test of time anymore.
What would be a good article to read on this topic these days?
 

Online PlainName

  • Super Contributor
  • ***
  • Posts: 7509
  • Country: va
Re: "Grounding" according to Analog Devices
« Reply #12 on: February 17, 2023, 04:08:47 pm »
Quote
because you can... respin the board if there is a problem

You can respin if you notice a problem. What about if it's a bit iffy but works fine on your bench - is it still going to work under different circumstances somewhere else?
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 11340
  • Country: us
  • $
Re: "Grounding" according to Analog Devices
« Reply #13 on: February 17, 2023, 04:49:36 pm »
https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html

(Well, I don't know how the author is affiliated with AD though.)

A lot of bad advice and a few reasonable points. Curious to hear about your thoughts.

Which sections are bad advice in your opinion? I have been designing 40 input, 1.0 uV-stable ADC circuits for thermocouple measurements with their IC's (AD7793) and following their design guidelines. Still looking for ways to improve this circuit if you could offer better advice.

I suggest a 1/4 size prototype for testing to see if you have the right track.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: "Grounding" according to Analog Devices
« Reply #14 on: February 17, 2023, 04:51:07 pm »
So... what's wrong with the article?
 

Offline Infraviolet

  • Super Contributor
  • ***
  • Posts: 1185
  • Country: gb
Re: "Grounding" according to Analog Devices
« Reply #15 on: February 17, 2023, 08:17:19 pm »
If you're still having a cookie popup problem, try NoScript or another script blocker, page loads fine for me using those, no pop-ups in the way anywhere.
 
The following users thanked this post: coppercone2

Online SiliconWizardTopic starter

  • Super Contributor
  • ***
  • Posts: 15800
  • Country: fr
Re: "Grounding" according to Analog Devices
« Reply #16 on: February 17, 2023, 10:21:50 pm »
10 years ago when I was speaking to an AD Lab engineer he said they use solid ground plane for characterization of their ICs and design of evaluation boards.

Not surprised. Anyone sane does that. You don't cut your planes with slits or worse (in most cases), separate ground planes and connect them in a single point.
That's a case of: do what I do, not what I say.
 

Offline EPAIII

  • Super Contributor
  • ***
  • Posts: 1165
  • Country: us
Re: "Grounding" according to Analog Devices
« Reply #17 on: February 18, 2023, 07:13:01 am »
They do it because the people who write code are tyrants. Their way or the highway! Period. Ever try to get one to do a simple project? No way Jose! Gotta have all the latest bells and whistles when all you want it to do is ring a doorbell.

Been there! They won! Overnight project took 1.5 years.



https://www.analog.com/en/analog-dialogue/articles/staying-well-grounded.html
Cant get past the stupid cookie consent.  Why do they do this....
Paul A.  -   SE Texas
And if you look REAL close at an analog signal,
You will find that it has discrete steps.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: "Grounding" according to Analog Devices
« Reply #18 on: February 18, 2023, 07:56:31 am »
Ground splitting is a sure way how to fail EMI tests and have random operation failures. In some cases causing IC latch-up, personally seen that in a commercial device that brand new cost like $100k. Replacing ADC or even swapping the whole board did not fix the issue of ADC latching up and heating like hell like 1-3 times in a year (sat on both analog and digital GND connected in other place on PCB). The only fix that worked was connecting grounds under this IC. It more or less worked in the past when digital part was slow. But now with very fast digital signals making slits in GND is intentionally making good antennas. I'm not saying is should never be done, however it only makes things worse in 99% of cases. Digital/analog ground splitting is not the worst though. The most popular thing seems to be splitting GND at USB shield. Many app notes suggest that too. The funny thing is that some suggest adding a cap in series, others a ferrite bead, making them do exactly opposite thing  :palm:. And some even add a cap in parallel with inductor making a nice LC circuit. Just look on this abomination from Arduino UNO. A nice way how to divert ESD from USB shield into data lines through devices which are supposed to protect from it.
« Last Edit: February 18, 2023, 08:18:18 am by wraper »
 

Offline JohanH

  • Frequent Contributor
  • **
  • Posts: 669
  • Country: fi
Re: "Grounding" according to Analog Devices
« Reply #19 on: February 18, 2023, 08:16:47 am »
"How to Achieve Proper Grounding" - Rick Hartley

Definitely recommended to watch this training. This is an expert with 50 years of experience of signal integrity and EMI and he assists companies in these matters.


 
The following users thanked this post: wraper

Offline JohanH

  • Frequent Contributor
  • **
  • Posts: 669
  • Country: fi
Re: "Grounding" according to Analog Devices
« Reply #20 on: February 18, 2023, 08:25:19 am »
Here's a list of literature that Rick Hartley recommends:

https://resources.altium.com/p/reading-list-emi-noise-control



 

Offline JohanH

  • Frequent Contributor
  • **
  • Posts: 669
  • Country: fi
Re: "Grounding" according to Analog Devices
« Reply #21 on: February 18, 2023, 08:27:13 am »
To be noted that @RobertFeranec on Youtube has many sessions and interviews with Rick Hartley.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: "Grounding" according to Analog Devices
« Reply #22 on: February 18, 2023, 09:15:59 am »
I quickly skimmed through the article and I don't see horrors that you guys suggest. Their example of adding slits makes perfect sense to me: to prevent high currents going through sensitive analog part where space is at premium and relocating components away from current paths is not an option ("due to the realities of the mechanical design").

I guess what we are talking here is that modern best practice is to have an uninterrupted ground plane, use of 4+ layers pcb, route digital signals on adjacent to gnd layer, and minimize interference between components by carefully placing them so their "return" paths do not overlap.
 

Online wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: "Grounding" according to Analog Devices
« Reply #23 on: February 18, 2023, 09:27:46 am »
I quickly skimmed through the article and I don't see horrors that you guys suggest. Their example of adding slits makes perfect sense to me: to prevent high currents going through sensitive analog part where space is at premium and relocating components away from current paths is not an option ("due to the realities of the mechanical design").

I guess what we are talking here is that modern best practice is to have an uninterrupted ground plane, use of 4+ layers pcb, route digital signals on adjacent to gnd layer, and minimize interference between components by carefully placing them so their "return" paths do not overlap.
The thing is that return path at high frequencies go only right under the traces if there is a solid ground plane, so they will not travel to somewhere far away into the analog part. And if you have traces which go over a slit in the ground plane, you create more problems than you solve as energy somehow needs to travel in the return path and it won't go over remote connection between analog and digital grounds. Note that frequency is not a frequency of digital signals but a rise time.



« Last Edit: February 18, 2023, 10:30:05 am by wraper »
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: "Grounding" according to Analog Devices
« Reply #24 on: February 18, 2023, 10:27:54 am »
The thing is that return path at high frequencies go only right under the traces if there is a solid ground plane, so they will not travel to somewhere far away into the analog part.

For digital signals yes, but that doesn't apply to low frequency analog signals. So no digital traces should go between analog parts, right?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf