6
\$\begingroup\$

I am using Nylon plastic screws as PCB mounts on my 2 layer PCB. The PCB has a very ESD sensitive and expensive device mounted on.

Since the Nylon material can develop positive static charge, is it a good idea to have 4 nylon screws on the PCB which houses a very ESD senstive device?

The PCB is to be screwed onto an ESD conductive box.

(The reason why I am not preferring Standard metal screws is that there a chance of an external charge can travel via the screws to the board/device and damage it.)

\$\endgroup\$
6
  • 3
    \$\begingroup\$ Have you asked the manufacturer what they recommend for your circumstances? (I'd expect at that cost, they'd spend a fair effort in helping you figure things out.) \$\endgroup\$
    – jonk
    Sep 16 at 5:32
  • \$\begingroup\$ Why are you using nylon screws? They’re quite weak and not dimensionally unstable with humidity. \$\endgroup\$
    – Spehro Pefhany
    Sep 16 at 6:01
  • \$\begingroup\$ @jonk I will speak with the device manufacturer to figure out their default option for screwing ESD sensitive boards. \$\endgroup\$
    – abunickabhi
    Sep 16 at 7:08
  • \$\begingroup\$ @Spehro the humidity is controlled in the experiment's environment. I saw the datasheet, the nylon screws are good enough for a 150g populated PCB. \$\endgroup\$
    – abunickabhi
    Sep 16 at 7:08
  • 1
    \$\begingroup\$ Use one SS that has ground connection and the GND plane connection of PCB, the others Nylon. You do avoid ground loops, since you have only one connection point. Or you may use all SS, one on plated hole, others on isolated holes. \$\endgroup\$
    – Marko Buršič
    Sep 16 at 7:23
4
\$\begingroup\$

No, they are not. Any plastic like that has the potential to build up charge over time from small movements or vibration (triboelectric charging). I would not use any untreated plastic anywhere near a sensor like that - in the mounting solution or the enclosure.

If you require electrical isolation, you should consider something like a dissipative mica washer paired with regular machine screws instead. This is a mica laminate with a thin conductive internal layer, which gives you both ESD dissipation and modest electrical isolation (depending on thickness).

But absolutely do not use nylon screws. If you can't afford to destroy a $10K sensor, then you won't save enough by using nylon screws anyway.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks for the dissipative mica washer + SST machine screws idea. Is a resin based option better than nylon, if I just want to stick with the idea of polymer based materials for mounting. \$\endgroup\$
    – abunickabhi
    Sep 16 at 7:20
  • \$\begingroup\$ Out of curiosity, what about screws for an M2 drive? If they're made of metal, charge wouldn't stay on would it? Unless these types of screws are treated so they don't hold a charge or they're connected to a chassis/GND so they would just discharge to ground? I would think that M2 drives are crucial, so I'm curious behind the engineering in this scenario given the OP question. \$\endgroup\$
    – BestQualityVacuum
    Sep 16 at 16:17
  • 3
    \$\begingroup\$ @BestQualityVacuum Huh, what are you talking about? Do you mean M2 as in the small form factor for SATA/NVMe SSDs and other computer expansion cards? If so, there's no magic involved at all, those are bog standard PCBs held down by a bog standard steel screw. I don't think the screw serves any electrical role, it's probably grounded just to be sure. \$\endgroup\$
    – TooTea
    Sep 16 at 19:54
  • \$\begingroup\$ @TooTea The OP mentioned how his PCB had an ESD sensitive device so he wanted to avoid std screws. I was assuming that M2 SSDs were also thought to be ESD sensitive, so I was wondering why standard screws would be okay in that case. \$\endgroup\$
    – BestQualityVacuum
    Sep 17 at 17:27
3
\$\begingroup\$

I do not see an issue with nylon screws (or plastic in general) if you create explicit grounding path that you control. The choice of mounting hardware depends on many aspects such as EMI requirements as well as grounding path from external(?) ESD sources.

I assume you may have a specific reasons to use nylon hardware. If not I would stay with SST. Then you better control the ground path. The very common way is to have mounting PTH holes on the PCB FAB connected to your system ground with parallel RC combination. Capacitor to address HF EMI issues in parallel with a high resistance resistor to alleviate ESD concerns while staying away from ground loops in the system.

\$\endgroup\$
2
  • \$\begingroup\$ Yes standard stainless screws sounds like a better option. I do not have a specific reason to use nylon screws, its just a design choice. I will explore the option of connecting system ground with parallel RC combination. \$\endgroup\$
    – abunickabhi
    Sep 16 at 7:13
  • 3
    \$\begingroup\$ @abunickabhi I do think the answer you have choose is the correct one, but the devil is in the details. Preventing an ESD discharge by isolating the board would imply only a higher breakdown voltage. Rather connect them together by using 1Meg resistor. My recommendation is to have just one connection point, so there will be very little discharge current - it has plenty of the time to equalize the charge, keeping the voltage low, so it won't breakdown in one other point. Having multiple points may cause a potential difference, so IMO isolate all of them but keep one PTH island and 1M resist. \$\endgroup\$
    – Marko Buršič
    Sep 16 at 7:46
1
\$\begingroup\$

A metal enclosure functions as a Faraday cage, so absent other connections to the outside world, ESD applied to the enclosure cannot affect the internal circuit, whether it's insulated from the enclosure or not.

The problem is with external connections. ESD applied to the enclosure can drive a current through the board to an external connection. But you have that problem anyway if you have more than a single external connection, since ESD to one connection can return through the others.

If you're concerned about ESD, it is important to put protection on each external connection. Shunt the ESD current to board ground. If your interface components have internal ESD protection that should suffice. Otherwise you have more work to do.

Then, it's generally safer to connect board ground to the enclosure. The enclosure is effectively just another external connection.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Connecting signal ground directly to chassis ground is an easy way out from the ESD perspective. However, much depends on the system design details. If there are multiple PCBAs and OEM components within the system then you do not have the luxury of grounding all of them since you would be creating a ground loops. That's where the ESD dissipative plastic or resistors come to play. All enclosure parts and external accessible connective parts (screws as an example) need to have a low impedance ESD current path to chassis ground bypassing sensitive circuitry. \$\endgroup\$
    – Martin Vana
    Sep 16 at 17:15
  • 1
    \$\begingroup\$ @MartinVana Indeed. If an honest electromagnetic analysis of the system shows that a ground loop involving the enclosure is problematic, then a leakage path is better than a hard connection (been there, done that). However, I'm strongly in the "let ground abound" camp. Usually, unless you have no ground loops at all, adding a ground loop reduces crosstalk problems by helping return currents find the easiest path. Since you usually can't avoid all ground loops, you're better off embracing them. \$\endgroup\$
    – John Doty
    Sep 16 at 17:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.