3

I'm designing a device for use in a non-climate-controlled environment (garage, shed, etc.) If I understand it correctly, this puts me directly into pollution group 3 unless I can also engineer it to be inside a sealed enclosure, which is probably not practical. It is, however, going to be in an enclosed metal box with basic dust protectors over the air intakes.

I would like to use components with standard 2.54 mm pitch, but as I understand it (not helped by the standards being unavailable), PG3 means I need 1.05mm creepage (for 12 V; 1mm for 5V) which is unattainable at this pitch with solder pads of any reasonable size.

I might be able to substitute some components, but not the microcontroller. Never mind the microcontroller board probably isn't built for PG3 either.

"Pot everything" isn't really an option, as a) I'm assembling this myself, and b) I'm dealing with rather large boards, up to 125 mm in some cases.

Is there anything I can do? Do I really need to be following PG3 rules (this isn't entirely clear), or is there something I can feasibly do that would let me get away with only PG2 creepages? Is there a way to increase creepage, keeping in mind I need to put parts on top of things, and I don't have 1mm between pads to add cuts?

p.s. I'm dealing with entirely THT design, no SMD components. (Also, if it makes a difference, with basic two-layer boards.)

Matthew
  • 481
  • 2
  • 12
  • What voltages are you dealing with? The voltages around a microcontroler shouldn't be high enough that creepage is a problem. – JRE Feb 02 '23 at 18:20
  • @JRE, 5VDC and 12VDC, which have creepage requirements (according to https://pcbdesign.smps.us/creepage.html, anyway) of 1mm and 1.05mm. And sometimes mains (120VAC), but *there* I'm generally not having trouble maintaining adequate creepage. – Matthew Feb 02 '23 at 18:30
  • 2
    I don't recognise those requirements being in the UK, but would some form of conformal coating do it? – Peter Jennings Feb 02 '23 at 18:32
  • By those rules, nothing low voltage can be built at all. – JRE Feb 02 '23 at 18:34
  • _"unless I can also engineer it to be inside a sealed enclosure, which is probably not practical"_ - why not? – Bruce Abbott Feb 02 '23 at 18:35
  • @JRE, huh? 1mm is perfectly attainable *if* you don't use 2.54mm pitch components. More than half my components are actually fine. Also, keep in mind the issue here is the pollution group; PG2 requirements are *much* more lax. – Matthew Feb 02 '23 at 18:36
  • @BruceAbbott, for one, I don't know *how* to make a "protected ingress" enclosure. Note that it needs to be ventilated; I can't just throw it in a water-tight box. – Matthew Feb 02 '23 at 18:37
  • @PeterJennings... maybe? That sounds like an Answer candidate. (I would hope an answer would include how to apply such a coating given that protection is needed underneath mounted components.) – Matthew Feb 02 '23 at 18:38
  • What microcontroller has pin spacing larger than 2.54 millimeters? What IC has pin spacing larger than 2.54 millimeters? – JRE Feb 02 '23 at 18:41
  • _"Note that it needs to be ventilated; I can't just throw it in a water-tight box"_ - perhaps you should consider a heat sinking system that avoids the need to ventilate the box. – Bruce Abbott Feb 02 '23 at 19:12
  • @BruceAbbott, that would be interesting, as the components I'm most worried about aren't intended to *have* heat sinks attached. Um... submerge the whole thing in mineral oil? That would probably *work*, but... – Matthew Feb 02 '23 at 19:58
  • @Matthew Thanks, but unfortunately I don't know enough about conformal coatings to write such an answer. Maybe this will give you a start https://en.wikipedia.org/wiki/Conformal_coating – Peter Jennings Feb 02 '23 at 22:18
  • _"the components I'm most worried about aren't intended to have heat sinks attached"_ - what are these components? – Bruce Abbott Feb 04 '23 at 22:32

2 Answers2

2

If you can't maintain enough creepage distance for a particular component, you really only have three options:

  1. Use a different component. (Not necessarily possible)
  2. Mechanically increase the creepage distance through the use of slots in the PCB if possible. (Not useful if the component itself doesn't have a high enough creepage distance).
  3. Change the pollution degree to a lower level in order to allow for reduced creepage distances.

The first two options are probably not all that useful to you. The third option is your best bet. But how?

If we look at some standards that define creepage distances, such as IEC 62109-1, they will provide information about how the creepage distances and pollution degrees are selected. By changing the environmental factors or enclosures it is possible to move to a different level.

As purely an example, take IEC 62109-1, it provides this table of possible remedies for moving to a lower degree. If you are currently at 3, then the red box shows how to move down to 2 or 1.

Pollution Degree Reduction

Essentially this boils down to either using an IP rated enclosure of suitably high protection, or applying a conformal coating. An IP5X or IPX7 enclosure would reduce the pollution degree to level 2. This would bring down your creepage requirements 0.04mm for PCBs or 0.5mm for other insulators. This is good enough for most components.

Conformal coatings are essentailly like a lacquer applied over the board. These act as water and/or dust barriers which insulate any exposed metal from the outside conditions. With suitable coating you can bring down the pollution degree to level 1, giving creepage of 0.025mm on PCBs or 0.08mm for other insulators.

Note that the above applies specifically to that standard. Other standards may have different requirements, so you should check what standard you are trying to conform to. The online calculator you posted in the comments appears to use UL60950-1 or IEC 62368-1 for example.

Tom Carpenter
  • 63,168
  • 3
  • 139
  • 196
  • I'm almost certain I can't use an IP-rated enclosure, as I have exposed components. *Meeting* IP5x might work, though (won't be certified). IOW, keep it clean, monitor for dust and/or condensation, and don't worry so much? For conformal coatings, I guess getting *under* components isn't an issue, because the idea is to prevent *contaminants* from getting in? – Matthew Feb 02 '23 at 20:06
  • 1
    @Matthew yes, for conformal you don't need to worry about getting under things, just coating the exposed metal. Anything under the coating is effectively inside is own local environment with lower pollution degree (same as anything under solder mask) – Tom Carpenter Feb 02 '23 at 20:21
0

I think you may be misunderstanding the point of creepage distance. It doesn't mean that every component must be spaced that much. It only applies to the space between two conductors at a significantly different voltage. I suspect that the pins of your IC are not experiencing a large voltage difference, so creepage distance is not an issue there.

If your product is connected to the AC power, then creepage distance is only an issue in the circuits connected directly to the AC power, and between those circuits and the low voltage circuit.

If your circuit is completely low voltage, creepage distance is not an issue.

Davide Andrea
  • 16,164
  • 4
  • 33
  • 62
  • "It only applies to the space between two conductors at a significantly different voltage." Besides that I believe I *do* have 12V differences in some of the relevant places... 1mm is the *minimum* creepage to meet spec, at *any* voltage, even 1V. – Matthew Feb 02 '23 at 20:23