6

I've been looking at my scope's power supply board and I noticed that R7731A "Burst Triple Mode PWM Flyback Controller" in DIP has two NC pins which are not connected to absolutely anything. They've been inserted into holes, but there are no matching pads on the solder side.

On the other hand, in various other devices I've seen components with NC pins which are soldered to a pad, but the pad is not connected to anything else.

So my question is: What's the benefit of not using solder pads for NC pins?

AndrejaKo
  • 23,261
  • 25
  • 110
  • 186

2 Answers2

3

When you solder to a pad -

Pro: -It is mechanically stronger

Con: -You will likely have a very, very minor increase in noise inside of the device, negligible in all but the most extreme situations (which you wouldn't be using a plastic device with NC pins anyway)

-You increase build time slightly and you also increase the risk of a solder short (or other process errors)

Anyway, I usually put a non connected pad there just in case I want to change ICs and wire a jumper wire over or something like that.

scld
  • 2,216
  • 15
  • 17
  • 1
    Build time is not a function of how many pins of a part are soldered down. The solder is placed on the pads on one operation, the parts are placed in single operations, and the board is soldered in one operation. There is no per-pin procedure. – Olin Lathrop Aug 07 '13 at 13:11
  • 1
    If you are wave soldering or using a reflow oven this is true. I was assuming a manual process in a more general manner. – scld Aug 07 '13 at 13:29
  • Manual processes are a miniscule exception, especially for parts that would have unused pins, like ICs. Those are all (to reasonable approximation) reflow soldered in a oven. – Olin Lathrop Aug 07 '13 at 14:56
  • 1
    I can imagine some situations where mechanical stresses could be reduced by leaving some pins unsoldered. Consider, for example, a DIP8 which needs pins 1-4 and 8 and was mounted to a somewhat-flexible board that was subjected to twisting forces. If pins 5-7 can move freely, the board will be able to twist without putting much force on the solder connections (e.g. on a horizontal board, if 1 and 5 move down while 4 and 8 move up, the chip just moves up). If pins 5-7 were soldered, the chip would not be able to move up with pins 4 and 8. – supercat Aug 07 '13 at 14:56
  • This issue can be especially significant with things like LCDs. I've seen a number LCDs which have two rows of pins, but only one of them is connected. If both rows are soldered, any board flex will break the LCD. If the non-connected row only has a single pin soldered, the board will be able to flex without flexing (breaking) the LCD. – supercat Aug 07 '13 at 14:58
3

In the normal case, make sure there is a pad for every pin. If you don't want to connect to the pin, don't connect to the pad.

The reason all pins must have pads is that during the soldering process when the solder paste is heated, the molten solder has surface tension that pulls each pin to the center of its pad. All pins being tugged on together pull the whole part into the correct alignment. With some pins not being pulled, the pull could be unballanced and the part not centered correctly.

For unusual mounting arrangements, like thru hole or where positioning tabs fit into holes on the board, this issue is different. In those cases it's less of a issue to leave off the pad. However, it reduces mechanical strength so defeats the main purpose of using the unusual mounting method in the first place.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915