flatheads

Well there ya go, I was going to suggest having the edge beveled at the board house. Can they still plate it after counter-sinking? That would be the way to go. (if you had some conic washer made with a plated through hole could you count on the solder surface tension to center the thing?)

George H.

The usual way it's done is a square area of copper on both sides larger than the bolt/nut/washer with lots of vias dotted about.

Also the usual approach is a header with plastic mounting pillars, it's quicker to assemble and isn't prone to undoing as nuts are. Doubtless you have your reasons.

NT

I have not done this, but have observed some things that require precise lo cating. Ideally I would use two flat head screws, one at each end of a spa cer. And the rest with round head screws with one of the round head screws going thru a elongated hole. So the flat head screws would establish the small board location in one dimension. The screw in the elongated hole woul d establish the location in another dimension. And the rest of the screws w ould not be creating stresses.

But you could probably could get away with using four flat head screws and no elongated holes. Might have some stress, but less then if you used all flat head screws.

Dan

-----------------^^^^^^^^^^^^^^^^^^^^^^^^^^^

Why not just a set of headers (that finely define the registration of the two boards) with a *gross* mechanical support elsewhere.

What happens if a screw is not tightened (or even *installed*)?

IME, people tend to think of hardware in mechanical terms -- with the routine exception of "picking up a chassis ground". So, you'll often find something reassembled with missing or different hardware than originally intended.

I assume you have a brain and these would be Phillips. But then not having been in that industry you might not know the correct term in "countersunk". That means the head is flay on top and though it could have any type of ma ting method broached or cut into it, there is a taper at the bottom of the head, which accomplishes what you apparently want.

Screws that were called flathead, with the slot, have been long obsolete, t hey are simply too hard to use. If you do not have a problem with the insta lled height of the head, the hole in the board is not that critical, howeve r then you make tightening torque critical as over tightening will break th e board. So the workers or robotics must have clutches on their tools. And of course you should be using machine screws, not be cutting the thread dur ing assemble, they need to already be cut.

It is not a bad way to have an automatic way to assure accurate positioning though. You just need accurate positioning of the standoffs or whatever yo u are using. Another thing is that for machine screws the workers or robots should put it to the hole and turn backwards until it drops, though going a full turn would work 99.9 % of the time without having to sense that. But dealing with that is the job of the process engineer. They do get it done one way or the other. Actually it is surprising how much assembly is still done by hand, but there is a reason for that.

Anyway, the main thing is to not overtighten it.

Normally this would be done using a connector for the electrical connections and whatever type of bolts/spacers for the mechanical connections.

The precision alignment can be done using either the connector or the bolts/spacers. To use spacers you just use the two flat head screws or use shoulder bolts with shoulders just a bit shorter than the thickness of the board. Shoulder bolts get around the problems with flat head screws putting side forces on the hole.

I think it would be easier to use a connector with alignment pins which can give you a very precise alignment by fitting in accurate tiny holes. Then the bolt holes can be coarse and not so accurate. There are many to choose from depending on your other requirements. The connectors can be bought in various heights to give you exactly the spacing you want. I have used Conan connectors from FCI before. They have a nice snap action when assembled. I don't think they get near a half inch spacing though. Samtec has lots of connectors that I'm sure will do this well. Heck, a PTH connector will do this.

--

Rick C

Oh, as to that specific question, yes I have seen it done.

The baby board supports a quadrature encoder and a knob, so I need the rig to be stiff and accurately aligned.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

can give you a very precise alignment by fitting in accurate >tiny holes. "

If I got that straight, I agree. Especially now he says there is a knob on it. And really, you only need two locating pins, which are called bosses I think in some types of mechanical engineering. They can be made to very tig ht tolerances as well as tapered to facilitate easier assembly.

What's more, without the taper on the screws a more reliable electrical con nection is likely.

Around here we call this a "flathead" screw

._SX342_.jpg

regardless of the type of screwdriver slot. In fact, I'll probably use Torx.

"Countersunk" refers to the surface that it screws into.

There are zillions still in use.

We'll use electric, torque limited screwdrivers.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Countersunk does not imply flat. Oval head screws have a tapered underside but a rounded top (not as "square" as pan head).

Also, you can buy countersunk screws with varying chamfer angles, 90 and 100 degrees being relatively common. The flatter the angle, the finer the thread pitch to achieve a given precision (in John's app).

Nonsense!

They are particularly common among woodscrews, as well.

No, you just have to use the right (slotted) screwdriver for the screw in question. E.g., a blade that is wider than the screw head at the depth of the slot will interfere with the material into which the countersink is bored.

Or, you select a screw material that isn't as rigid: In John's case, the requirement that the screw also carry current rules this choice out.

M3 countersunk head screws in a 4mm hole auto-centre nicely. I did this recently. Do this and have large pads top & bottom stitched with vias.

Yeah, I've done that for self-centering in loose holes. It's not the best for electrical connection, though: that would be a largeish pad, with a star washer, that makes gas-tight metal/metal contact when tightened. Round head or pan head screws are fine (or Allen screws, or brass standoffs). After coming up with a work instruction on how tight was tight enough, it had to pass ground-bond testing (30 amps).

1. Use dowel pins:
   formatting link

Press them into the bottom board, using a slightly undersize hole. Use a slightly oversize hole in the top board, so it slides on, without slop.

Better still, press them into a metal frame that both boards are secured to, so they aren't at the mercy of floppy, cold-flowing FR4.

(You'll still need at least one screw to anchor the board in the axial direction, or maybe some snaps or something.)

1. Some PEM nuts[1] have a collar or shoulder, with a threaded inside or outside. The top board could be made to fit snugly over the shoulder; the screw(s) would only serve to keep the board anchored down against the studs.

[1] Nearly a genericized brand name. They still make the originals, but a huge variety is available from wherever. They can be pressed, soldered or swaged into boards and sheet metal.

1. Use through-hole board-to-board stacking headers. (SMT headers will be subject to placement and soldering alignment problems. Good, but not low-thous good.) Snug fitting holes and connectors will align closely, and the bulkier type connectors (the ones with alignment pegs or anchor clips) will stand up straight without bending or flexing.

2. Use more rigid PCBs, mounted sideways. Solder the monstrosity together. It's like building with copper clad, except the pieces are as-machined so they fit square and precise.

1. Insert zanier ideas here if you still aren't satisfied. ;-)

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: http://seventransistorlabs.com 

"John Larkin"  wrote in message  
news:2m4htbto0r6egqgfobf1pcp9961gjatp28@4ax.com... 
> 
> 
> We want to mount a small PC board maybe half an inch above a bigger 
> board, with four threaded standoffs. The standoffs would be both 
> mechanical and four electrical connections. We want fairly precise 
> positioning, so we were thinking about using eight flathead screws. 
> They would center in the holes nicely. 
> 
> Has anybody done this? Any problems? 
> 
> I can imagine maybe fracturing a plated-through hole, but we can make 
> the required connections redundantly, on both sides of both boards. 
> 
> I tried it on a scrap board and it did distort the edges of the PTH a 
> bit, but it looked OK. 
> 
> I have actually had boards made with plated countersunk holes, but the 
> board houses consider that to be weird. 
> 
> 
> 
> 
> --  
> 
> John Larkin         Highland Technology, Inc 
> picosecond timing   precision measurement 
> 
> jlarkin att highlandtechnology dott com 
> http://www.highlandtechnology.com 
>

Oh, there are also precision ground shoulder bolts, used for the inverse method as the studs/PEM nuts I mentioned.

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

ng been in that industry you might not know the correct term in "countersun k". That means the head is flay on top and though it could have any type of mating method broached or cut into it, there is a taper at the bottom of t he head, which accomplishes what you apparently want.

that's a countersunk screw.

NT

I've done this, but used crinkle washers against both PCBs top and bottom. It worked fine, and their springiness gave me a nice warm feeling that it would continue to work fine.

Like this, just in case you call them something else over there...

Cheers

--
Syd

I've never heard countersunk screws referred to as 'flathead'. Mine were normal screws.

Cheers

--
Syd

On Sep 13, 2016, John Larkin wrote (in article):

It?s not a good idea to use a machine screw through a PW board to make electrical contact to the copper foil because of the large Z-axis coefficient of thermal expansion of the PWB - the screw will surely become loose. The only remedy is a spring of some kind that can maintain contact even as the board thickness varies. Belleville spring washers are traditional here.

Also beware that in the X and Y axes, the thermal coefficient is roughly that of copper, so the copper foil won?t pop off. So, if one rigidly attaches a pub to a aluminum or steel plate, the assembly will flex (and/or tear itself apart) due to temperature cycling. Think bimetal thermostat.

So, for precise and stable location, implement some kind of kinematic mount.

And for electrical connection, use pin and socket connector of some kind, one with sufficient mechanical flexibility that it will simply bend slightly.

And beware fretting corrosion in the pin and socket connector, as vibration and/or thermal cycling can interrupt the connection. Note that tin contact surfaces in a connector are particularly vulnerable to fretting corrosion - gold may be required.

Joe Gwinn