Author Topic: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?  (Read 3205 times)

0 Members and 1 Guest are viewing this topic.

Offline veryevilTopic starter

  • Supporter
  • ****
  • Posts: 221
  • Country: gb
ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« on: June 25, 2018, 11:38:24 am »
Hi, was about to add some a couple of unidirectional TVSs to a few  I/O lines on my lastest project and discovered that you cannot get TVSs for 1V8!

How do I protect the lines from ESD and transients? The are simple 1V8 high or low digital signals on exposed pins which could be touched.

I havent come across this and would love to know as 1V8 and lower isn't uncommon these days so I assume there is standard methods of doing it but my google fu has let me down on this onw.

Cheers Guys!
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #1 on: June 25, 2018, 03:17:34 pm »
The usual way is with diodes connected to the supply rails, as ordinary silicon diodes have too higher voltage drop, use Schottky.

The 1.8V supply can be protected with four silicon diodes in series.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #2 on: June 25, 2018, 03:53:52 pm »
Although if these are high speed connections, you will still want to use junction diodes -- less capacitance.  A packaged TVS array would be fine, it'll just have a superfluous TVS part inside (typically 5V).  Connect it to supply and put a bypass cap directly beside.

There will be some let-through which may still be a concern, especially on low voltage IOs.  Add a series resistor, between TVS and IO pin, as large as you can afford (10s of ohms to a few k?).

The better question is, what the bloody hell are you doing, bringing CPU internals out to a connector?!  Level-shift those up to 3.3V or more, if at all possible.  Use LVDS if it's fast.

I know there are some MCUs and low power devices down there, which might be expected to see ESD, while also not being very fast.  But still, even just the noise margin there is pitiful.  You'll need coax to bring those signals any distance over a cable!

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline CopperCone

  • Super Contributor
  • ***
  • Posts: 1415
  • Country: us
  • *knock knock*
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #3 on: June 25, 2018, 04:19:55 pm »
I dont see a good reason to have 1.8v thats not shifted or nothing unless its really phase clean or if someone has a problem with local regulators. They make ldo 1.8v with sense connection so someone probobly can put it off board in some kind of power supply card or module. Might be done for easy testing on psu integrity and easy sequencing with out complex wiring... Screams of cheap design unless its weird
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 26757
  • Country: nl
    • NCT Developments
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #4 on: June 25, 2018, 04:24:11 pm »
The usual way is with diodes connected to the supply rails, as ordinary silicon diodes have too higher voltage drop, use Schottky.
Never do that! You'll lift up the entire power rail and you'll capacitive couple noise on the power rails to an external connector.
I like to use varistors and series resistors if the signals are slow. The capacitance of the varistor absorbs narrow pulses anyway.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 19345
  • Country: gb
  • 0999
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #5 on: June 25, 2018, 05:00:12 pm »
The usual way is with diodes connected to the supply rails, as ordinary silicon diodes have too higher voltage drop, use Schottky.
Never do that! You'll lift up the entire power rail and you'll capacitive couple noise on the power rails to an external connector.
I like to use varistors and series resistors if the signals are slow. The capacitance of the varistor absorbs narrow pulses anyway.
That's what decouping capacitors are for: to prevent the supply rail from being lifted up, by these transients. If they don't, then they're too small or it's being badly abused! I did suggest four silicon diodes, in parallel with the supply, to clamp at <3V, as a last resort, but they won't do anything most of the time.

Good point about the series resistors, I forgot that. They can be low values, for fast signals, but they'll provide less protection.
 

Online tszaboo

  • Super Contributor
  • ***
  • Posts: 7313
  • Country: nl
  • Current job: ATEX product design
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #6 on: June 25, 2018, 06:14:19 pm »
Here is one for 1.2V katalog.we-online.com/en/pbs/WE-TVS-SS
They dont seem to have 1.8 though. I would just protect these with 3.3V TVS, since the voltage always going to be higher than the clamping voltage during an ESD event anyway.
 

Offline SiliconWizard

  • Super Contributor
  • ***
  • Posts: 14309
  • Country: fr
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #7 on: June 25, 2018, 07:15:55 pm »
TVS are supposed to protect from fast high-voltage transients, a lot less to protect from general overvoltage. For instance, most TVS used for protecting 3.3V lines (such as USB) have 5V or even 8V working voltage and clamp at higher voltages, often up to 12V. So by themselves they are meant to protect from ESDs but are limited for other uses. So you can use a 5V TVS for instance to protect your 1.8V lines. They will absorb most of the energy of an ESD, which is the whole point, but it's not always enough.

A more complete protection is often the following: place the TVS as close to the connectors as you can, then put a series resistor to the I/O. Additional protection may consist of putting two clamping diodes (to 1.8V and GND) after the series resistor, followed by yet another series resistor. Of course you have to be careful about the resulting low-pass filtering. Note that some modern ICs, especially interface chips, already have this kind of protection scheme built-in for their I/Os, so an external TVS may be enough.

You may also use buffers/level shifter or any kind of phy interface to your signals for an additional layer of protection (we'd have to know what you intend to connect those signals to...)

A few tips from On Semi: https://www.onsemi.com/pub/Collateral/AND8230-D.PDF
 

Offline veryevilTopic starter

  • Supporter
  • ****
  • Posts: 221
  • Country: gb
Re: ESD Protection on Low Voltage I/O lines (1V8, 1V2)?
« Reply #8 on: June 26, 2018, 06:51:11 am »
That's a lot of different ideas, thanks. I'm surprised there is no standard way.

There are two pins and they are simple enable type signals so very low speed.

The are on a pogo type connector on a docking port so when not docked are exposed.

Currently thinking of 3v3 esd, 10k series resistor then BAV99 clamping to 1v8 and ground. making sure there is some local capacitance on the 1V8 rail.

Thanks 
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf