0
\$\begingroup\$

I'm about to replace a relay with a discrete triac design. The load is a (resistive) heater element and power 1200W @ 240V (RMS current ~5A). I will use zero crossing detection and max switching frequency of 1Hz. Environment temperature < 50C. The triac is fired from an mcu via optocoupler.

I'm looking for advices to create a robust solution from a practical point of view and in particular from field experience.

  1. What rating should the triac have for V and A.
  2. Is it possible/advisable to avoid a heat sink?
  3. Voltage surges are common in the region, how will that effect component selection?
  4. Can a (fast) fuse be effective in protecting the triac from short circuit and power surge? If not, are there known working solutions for circuit protection.

Thanks in advance!

\$\endgroup\$
5
  • \$\begingroup\$ Properly rated relays are magic devices. A TRIAC will dissipate substantial energy due to voltage drop when on. A relay won't. The TRIAC has two series junctions so dissipate about 1.2W/A or more. If you switch very frequently a TRIAC may be the best solution. If you SOMETIMES switch at 1 Hz but on average switch at a lower rate a relay may be the best solution. A 100,000 operation rated device switched at 1 Hz average lasts about 1 day. At 1,000,000 operation it lasts about 2 weeks. But an average of 1 operation per hour of a 100,000 operation device will give about a 10 year lifetime. \$\endgroup\$
    – Russell McMahon ♦
    Oct 30, 2017 at 12:42
  • \$\begingroup\$ Thanks Russel. Yes, reason for the change was the limited life time of relays when switching frequency increases. \$\endgroup\$
    – Mike
    Nov 2, 2017 at 11:25
  • \$\begingroup\$ Yiu say load is pure resistive. If this is the case the following will not help as much but may still be useful. - With reactive loads One "trick" is to use a TRIC in parallel with relay contacts. Switch TRIAC on, then relay, then turn TRIAC off. TRIAC sees lower to much lower dissipation depending on on/off times and relay has surge current removed. Better is probably a FET based bipolar switch (FET in bridge or 2 FETS) which entails more complexity but gives more flexible control of what happens when. ... \$\endgroup\$
    – Russell McMahon ♦
    Nov 2, 2017 at 23:32
  • \$\begingroup\$ ... Having solid state switch handle turnoff is also liable to be a major contributor to lifetime improvement with reactive loads. \$\endgroup\$
    – Russell McMahon ♦
    Nov 2, 2017 at 23:32
  • \$\begingroup\$ Thanks Russel, never thought of that. I'll take note of that and keep it for future projects! \$\endgroup\$
    – Mike
    Nov 12, 2017 at 6:27

1 Answer 1

Reset to default
2
\$\begingroup\$

  1. As high as you can afford. 25A or 40A would be good. No triac you can afford will be as rugged (or efficient) as a relay contact.

  2. Not a hope. Dissipation will be in the 5W range- read the datasheets and do some calculations. Do not run the triac junction too hot. It won't last long and you need the delta between operating temperature and absolute maximum to absorb the temperature surge from a short or other abuse.

  3. Use at least a 600V triac, preferably 800V, and consider an MOV.

  4. Yes, a fast fuse can protect a generously overrated triac from a short circuit. The more demands you place on the fuse, the more expensive and hard-to-find it becomes, and the more likely your customer will replace it with a penny or whatever.

Keep in mind that triacs are much more likely to fail 'on' than properly rated relays or contactors are.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks Spehro. You confirmed my suspicions, in particular the risk of failing on, which is a headache. \$\endgroup\$
    – Mike
    Nov 2, 2017 at 21:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.