My screw is not connecting to my NPTH

[Saimoun]

Hi

I have heard using Plated Through Hole for screws is a bad idea since tiny bits of the platting my come off and go on the board between two pins (like the GND and VDD pin  )

So I switched to NPTH (non-plated). So far so good. I just noticed though, my screw (and the metal case) is not electrically connected to the pad (=GND) anymore.

My questions:
1) Is PTH that bad? My smallest pitch is 0.5mm in a QFP or QFN package.
2) Am I doing the NPTH right (see pictures)?
3) Should I change the screws?

Thank you

EDIT/PS: to check connectivity I used a multimeter in "beeping" mode, putting on end on a ground pin and other end on the screw head. On the boards with the PTH it beeps, with the NPTH is does not.

[Saimoun]

Ok actually if I really tighten the screw (like a lot  ), then I do have electrical connection.

[sokoloff]

Split the problem in half.
Is the top surface of the NPTH grounded?
That will tell you if it’s a board problem or a screw problem (a surface finish likely).

[wraper]

Unless you are using screws which are too large for the holes, there should be nothing flying around. In computer hardware using both NPTH and PTH is a common industry practice as much as I'm aware. If copper around holes on both sides are connected with vias, it should make no difference IMHO.
What's on the bottom side of your PCB? Is copper around the hole exposed or under solder mask?

[wraper]

I noticed a diode or TVS connected to GND and with a ferrite bead in series. And no via in between. Looks weird.

[Saimoun]

@sokoloff: yes the top surface is grounded.

@wraper: Nice to know about PTH
My screws are M3, the hole is based on the M3 footprint from Kicad: hole is 3.2mm, copper is 0.25mm away from the hole, and solder mask is 1.9mm from the hole (more than enough for the screws as you can see on the pictures).

Bottom side is identical to top side (ground copper pour and yes the copper is exposed just like the top).

[Saimoun]

I noticed a diode or TVS connected to GND and with a ferrite bead in series. And no via in between. Looks weird.

Unrelated to the discussion but I'd like to hear more about that 
Why is that weird? That trace is the USB 5V power. TVS is to protect from ESD and ferrite to reduce noise/EMI in or out.

[wraper]

I noticed a diode or TVS connected to GND and with a ferrite bead in series. And no via in between. Looks weird.

Unrelated to the discussion but I'd like to hear more about that 
Why is that weird? That trace is the USB 5V power. TVS is to protect from ESD and ferrite to reduce noise/EMI in or out.

It's just that it does not look like ferrite bead is connected in series with 5V VBUS. But GND-TVS-FB-VBUS.

[gcewing]

A star washer might improve contact between the screw head and the trace.

[Saimoun]

@wraper: Ah ok No it is connected to the VBus See picture.

@gcewing: great idea!

[T3sl4co1l]

What type of hardware?

I suppose stainless might have trouble connecting, though I can't say I've seen that happen.

It looks to be steel, either bright plain, or stainless, or plated.  Zinc plating should be fine too.  Nickel I guess might not be so great for a similar reason as above, but isn't as common anyway.

I think I see a star washer under there, so, give or take if it's properly clamped under the head as the head looks fairly well rounded.  Maybe an internal tooth type gives better results?

What's on the bottom, metal standoff? Another toothed washer? (Shouldn't be needed, but just for completeness..)  Guessing it's tied in to more metalwork hence the point of grounding.


Also, what the heck is that, it's like a trace that's not poured over, but everything else is poured around, between the screw and another pin?  Why didn't that pour?

Tim

[Saimoun]

Hi Tim 

Not sure what the screw is, but I guess it is standard plain steel.

There is no washer, not sure where you saw that But I might use one in the future.

On the bottom it is a metal standoff indeed, directly clinching to the case, the PCB is resting on it, nothing else in between.

And about that trace, you mean the large trace from the pin underneath to the screw head? Long story short I have a pot made of metal, but it does not quite touch the metal enclosure. So I wanted to ground the pot structure, but in case it touched the enclosure I was worried it would create a path: ground plane -> metal pot -> enclosure -> screw -> ground plane, so instead of connecting the pot directly to the plane I connected it to a trace which then connected to the screw. Anyways, it does not seem the pot touches so I changed that to simply connect the pot structure to the ground plane directly.

And to finish about the screw hole, I decided to go back to plated, like wraper said it should be fine The other screw holes are still non plated.

[HwAoRrDk]

You might want to consider using a flanged head screw instead.



Your current screws look like the underside of the head is too rounded to make proper contact.

[T3sl4co1l]

Oh those are just the vias aren't they, not a star washer peeking out.

You might want to consider using a flanged head screw instead.



Your current screws look like the underside of the head is too rounded to make proper contact.

Can even be the kind with the rough (ratcheting) under surface, which should help with wiping contact.

Or captive star washer screws, pretty easy to get, McMaster has them for example.

Tim

[TimFox]

I have been spoiled by McMaster-Carr:  I have ordered stuff around 5:30 PM and received it the next day (UPS) at my residence.  I was disappointed to find out that they do not solicit international business.  (The original poster is in Denmark.)

[T3sl4co1l]

Yeah, I meant to add: YMMV for international customers...

Tim

[CatalinaWOW]

Just an off the wall thought.  Is your screw bottoming in a blind hole or other physical obstruction before the head applies significant pressure to the board.  Might explain your symptoms.

[TimFox]

There are two types of "star washers":  internal and external tooth.  Depending on the pad surrounding your hole, you should choose the type that contacts more metal.

[Kean]

One resaon for concerns with screws in undersized plated holes is with multi-layer PCBs without sufficient clearance on inner layers.
If the screw bites into the side of the plated hole it can cause a short to power planes on those inner layers.

[T3sl4co1l]

Ironically that might cut the other way.  But it depends on settings of course.

By default, well -- by default default, it's the same pad on all layers, so it clears at about the head diameter (whatever pad OD is set to, hopefully bigger than this!) on all layers.  The pad is physically ripped out of the entire board before other copper is affected, give or take any torn traces or fractured components from the incredible stress you'd have to apply to do that!

If you're using a custom padstack with reduced inner pads, that's less, but still fine -- the screw would have to chew up the hole quite significantly to cause shorts (say 20 mil annulus + 10 mil clearance?).

By default, unconnected inner pads are removed (and if you don't do it in the source, your fab will likely recommend such -- or for the low cost protos, they may simply do it automatically without notification!), leaving both the annulus and clearance as empty area -- again a significant clearance around the hole, even if custom.

Unless you've removed the pads entirely (typically, set inner pads to 0 or hole size), in which case it will be simply the inner layer copper-hole clearance, whatever that's set to.  Which may be rather small, and this can be a hazard for fab (drilling + registration tolerance) as well as damage.

FYI, I've used this trick (reduced inner pads that are optimized out of the design anyway) in some special circumstances, like tight thermal via patterns while allowing inner planes (not connected to the vias) to pour between them (whereas they would be broken by default via padstacks).  Just reduce the inner pads a modest amount, don't set them to zero.  No problem.

Tim

[Saimoun]

Thank you all for your replies 

As I said I already switched to a PTH. But Very good point about switching the screw to flanged headed, that will definitely do the trick 
And yes of course the inner layers would be away from the hole - in my case I have none (2 layer board).

And @CatalinaWOW: I did not think about that, my standoffs are blinded indeed. I do not think the end of the screw gets anywhere near the bottom, but it's definitely something I will keep in mind for the future - a trap for young players like someone would say

[Ian.M]

Its worth noting that FR4 PCB material does tend to cold flow under pressure so you'll always need some sort of spring washer for long-term reliability of a bolted connection to it.   Probably the best option to maintain grounding in this case would be a wave washer, as its less likely to break through the copper layer under load, than alternative types of spring washer.

[CatalinaWOW]

Bottoming in a blind hole is not caused solely by the design depth but also by any debris left by manufacture and/or caused by thread engagement.  It's easy to check, just run a screw into the hole without the board in place look.  My personal standard is that if the head isn't touching the post there isn't enough clearance, but you could cut it closer if needed.

[T3sl4co1l]

Yeah, cold flow can be a problem.

Hmm, seems like I've taken apart plenty of things that seemed to still be nice and sturdy with the traditional toothed washer and ground wire.  It might not be that big of an issue, more of an occurrence thing than a "just wait for it" thing?  Or, it matters more for higher operating temperature or wide temperature swings, such as you'd expect from rough service, or high amperage connections.

Belleville (conical spring) washers, or stacks thereof (same-way nested to increase stiffness, opposite-way-stacked to increase displacement range), may give better results; wave washers tend to be pretty weak, so keep that in mind.

Tim

[Ian.M]

I was working under the assumption that Simon simply needed EMI grounding so as its not a high current connection, the low force from a wave washer would be sufficient, with vastly reduced risk of PCB damage.

Bellville washers bring their own problems with line loading.  If you don't put a large flat washer under them to spread out the load and they are cup down, its possible over time for them to punch out a nice disk of PCB round the bolt hole!  Its pretty rare, but I've seen a couple of boards that had bolted terminals that broke loose suspiciously neatly.