Triac Dimmer

Hi all, I was wondering if I could get some critique on my first PCB layout. The schematic is a direct copy from here.

It is for a 240V Dimmer 10A, and I am after comments on my PCB design.

For clarity, here are the parameters I applied.

  • Grid Size: 100 mils
  • Tracks Digital - 10 mils
  • Tracks AC Sense - 20 mils
  • Tracks AC Power - 150 mils
  • Triac Middle Pin staggered - 150 mils

The plan would then be to get this made at a fab house with 2 oz copper.

  • \$\begingroup\$ are you planning to make this at home or have a fab house make it for you? \$\endgroup\$ – Wesley Lee Jul 27 '16 at 10:04
  • \$\begingroup\$ Oops, forget to add some of that detail. Thanks and fixed. \$\endgroup\$ – Nicholas Tiong Jul 27 '16 at 10:13
  • \$\begingroup\$ how are you considering to cool your triac? at 10A you'll get possibly 15W of heat, maybe I'd make some space for a heatsink (keeping trace-trace distance in mind, more than you put in your current PCB) \$\endgroup\$ – Wesley Lee Jul 27 '16 at 10:59
  • \$\begingroup\$ The triac will be cooled by a heatsink that is about 10x15x40mm in the space behind it (to the left). At this stage I have not calculated if that will be sufficient. \$\endgroup\$ – Nicholas Tiong Jul 27 '16 at 11:07
  • \$\begingroup\$ I'm unsure on what you mean by trace-trace distance however in relation to the heatsink. Was that in reference in ensuring that track clearance is met? The triac is of a isolated tab design. Will additional clearance be required? \$\endgroup\$ – Nicholas Tiong Jul 27 '16 at 11:13

The main issue I see here is that the U1-U2 GND trace infringes on the isolation barrier in the design. I would shift the P1/P2/R1/R2 complex closer to U1/U2 -- this leaves room to reroute the ground trace to U1 between P2 and the edge of the board, then have it pop out between the pins at the bottom of P1/P2, so it doesn't compromise the isolation barrier U1 and U2 provide.

The other issue with this design is an annoying mechanical matter -- the BTA12-600CW has an isolated tab, but it's only rated to 2500V isolation, which isn't sufficient for a reinforced barrier in 240V land AFAIK, and the heatsink infringes on the isolation barrier otherwise. I'd change it to wrap around the side in an L-shape, with the fins exiting the board envelope altogether -- this increases the mechanical envelope of the design, but keeps the isolation area clear and also allows you to increase creepage and clearance distances from the HS to the TO-220's pins. Either that or find a TO-220(iso) triac with a 5kV isolation rating on its tab and suitable specs otherwise (600V/min 10A). Of course, if you're happy with 2.5kV of isolation here, you simply should make sure the HS has sufficient clearance/creepage to the TO-220 pins.

Another option, depending on how much of an isolation stickler you are, would be to shift the optoisolators left by a couple hundred mils. This would place the existing heatsink firmly on the mains side of the isolation barrier. You may wish to swap the optos for 400mil packaged ones while you're at it, too, as these provide improved creepage, to the full 8mm needed for 240VAC reinforced insulation (surge rating of 5kV).


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.