Author Topic: Disconnecting power jack and ESD  (Read 783 times)

0 Members and 1 Guest are viewing this topic.

Offline lawrence11Topic starter

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: ca
Disconnecting power jack and ESD
« on: October 14, 2019, 04:52:51 pm »
Am I just paranoid?

Or should I add a zener clamp between Jack+5v and GND?
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12851
Re: Disconnecting power jack and ESD
« Reply #1 on: October 14, 2019, 05:14:49 pm »
That won't happen.  The input capacitance prevents a significant voltage rise on disconnect., or even on HBM ESD discharge to the jack center pin.

However when you reconnect the jack, if the supply voltage at the moment of connection exceeds half of the LDO's max input voltage rating, all hell can break loose.   The problem is the transient due to the supply lead inductance ringing with the input capacitance, which if undamped can reach double what the supply voltage was at the moment of connection.   If you use ceramic input capacitor(s) its compounded by their reduction of capacitance with increasing voltage.  Pop goes the LDO and everything on its output rail that can't tolerate the full input voltage.

Therefore Zener clamping (with an upstream fuse or polyfuse)  is *STRONGLY* recommended if the nominal supply voltage exceeds around 40% of the LDO's max rating.

Alternatively, use a regular electrolytic of 10x the ceramic input capacitor value in parallel with it so its ESR provides damping, or if you need to eliminate 'wet' caps,  use an extra ceramic of 10x the value with a 1R resistor in series to provide a similar ESR.
« Last Edit: October 14, 2019, 05:20:00 pm by Ian.M »
 

Offline lawrence11Topic starter

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: ca
Re: Disconnecting power jack and ESD
« Reply #2 on: October 14, 2019, 05:18:37 pm »
That seems like a pretty random situation,  but it could happen. You just never know between debouncing and just bad handling.

So it would be wise to put a through hole zener on the bottom side of the pcb  right underneath the plastic jack?

From jack+5 to GND?

The flat like jack connector has a half moon of space, for the through hole lead to be inserted.
« Last Edit: October 14, 2019, 05:26:59 pm by lawrence11 »
 

Online ejeffrey

  • Super Contributor
  • ***
  • Posts: 3709
  • Country: us
Re: Disconnecting power jack and ESD
« Reply #3 on: October 14, 2019, 09:53:05 pm »
Its not that weird or unusual and it doesn't have anything to do with debouncing or bad handling.  A lot of LDOs have very low absolute maximums.  For instance, I have used a 3.3 V LDO with an absolute maximum of 6V.  If you hotplug that to a 5 V supply (like a USB port) with no clamping or damping you are going to easily exceed the max rating.  Even with a minimum supply voltage like 3.6 that can go over the 6V rating.

 

Offline pigrew

  • Frequent Contributor
  • **
  • Posts: 680
  • Country: us
Re: Disconnecting power jack and ESD
« Reply #4 on: October 15, 2019, 02:56:09 am »
Thanks for the reminder of the LC ringing issue. I need to test for it in future designs I do...

I often add a soft-start/OVP IC to the input, which solves the issue by limiting the capacitance presented to the line when first plugged in.

I really do like ceramics, but this is a good reason to add tantalum or electrolytic capacitors.

I also found a nice website discussing the issue.
« Last Edit: October 15, 2019, 02:46:41 pm by pigrew »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf