4
\$\begingroup\$

When looking through datasheets I can't find a single one that shows the inner circuit of a triac driver optocoupler, all I get is this common diagram:

What that zero crossing circuit looks like? how is it connected to the "triac"? I'm specifically looking for details of this part MOC3063M, but any description is appreciated.

\$\endgroup\$
7
  • 2
    \$\begingroup\$ lednique.com/opto-isolators-2/opto-triacs-zero-cross-work \$\endgroup\$
    – G36
    Dec 3, 2023 at 16:00
  • \$\begingroup\$ I have tested MOC3043M and it stops triggering the triac when V > 16V. \$\endgroup\$
    – G36
    Dec 3, 2023 at 16:17
  • \$\begingroup\$ @G36 So it does have a snubber inside? \$\endgroup\$ Dec 3, 2023 at 17:39
  • \$\begingroup\$ No, it does not have any snubber inside. This is just a example circuit. \$\endgroup\$
    – G36
    Dec 3, 2023 at 17:47
  • 1
    \$\begingroup\$ Because this schematic does not show the internal circuit of a MOC3063. Also in IC large capacitors/resistors take a lot of space therefore they are avoided. \$\endgroup\$
    – G36
    Dec 3, 2023 at 18:03

2 Answers 2

7
\$\begingroup\$

Here is a simple possibility that looks somewhat similar to @Tim's die shot.

enter image description here

T1, T2, T1', T2' are the opto triac. aand T3/T3' shunt the gate current when the voltage is more than the zero crossing voltage.

enter image description here

Images are from US 4,396,932 1983.

Obviously you cannot duplicate this circuit with ordinary discrete components because they could not withstand the required voltages (in particular Vgs(max) on T3/T3').

\$\endgroup\$
6
  • \$\begingroup\$ Isn't this better called anti-parallel SCRs, not a triac? \$\endgroup\$
    – bobuhito
    Dec 11, 2023 at 22:01
  • \$\begingroup\$ And, why aren't other companies doing this? The patent is expired. \$\endgroup\$
    – bobuhito
    Dec 11, 2023 at 22:05
  • \$\begingroup\$ There are a number of companies making the ‘3041 type zero crossing optos. The ‘3063 is enhanced and may have some tricks that are still under patent. If you look at patents which reference the above one you should see various improvements which may or may not have made it to market. Whether the IC structure is portrayed for the user as a diac/triac or SBS, dual SCRs etc. is kind of a moot point. There is no gate, after all. \$\endgroup\$ Dec 12, 2023 at 5:18
  • \$\begingroup\$ Really? I can only find MOC3041 from Fairchild, which was later acquired by ON Semiconductor. If it's beyond these two brand names, can you list some more parts that you believe use this patent? \$\endgroup\$
    – bobuhito
    Dec 12, 2023 at 7:38
  • \$\begingroup\$ Isocom MOC3041X, Everlight (Taiwan EL3041), Toshiba TLP266J, Sharp, QT, Panasonic, Vishay, Lite-on, and then there are the Chinese makers such as UMW, Cosmo, CT Micro (who have 800V rated parts such as CT3083). There are probably more. \$\endgroup\$ Dec 12, 2023 at 8:09
7
\$\begingroup\$

The zero-crossing circuit looks like this:

MOC3061 - Zero-cross optoisolated triac driver

Source: MOC3061 - Zero-cross optoisolated triac driver : weekend die-shot | Zeptobars
License: Creative Commons Attribution 3.0 Unported

It appears MOC3061-3 is a family, differing mainly in photosensitivity, which will depend on die and packaging variation, so likely they are identical parts (as fabbed), and labeled after testing; i.e. it isn't material that I'm answering a MOC3063 question with a -1.

As for what this means, that is a good question. It's not clear to me where the bonding pads even should be, and there's no good reference for which layers are which, other than the shiny metal layer on top of course. I would assume the substrate is n- doped, with epitaxial p- on top, then traditional microcircuitry on top of that.

An equivalent circuit comparable to that linked by G36, Opto-triacs, solid-state relays (SSR), zero-cross and how they work, certainly seems likely.

\$\endgroup\$
2
  • \$\begingroup\$ That's both the zero crossing circuit and the high voltage opto-triac, of course. \$\endgroup\$ Dec 3, 2023 at 19:49
  • 1
    \$\begingroup\$ Yeah, I'm guessing something like, whereas a traditional TRIAC is three layers stacked plus diffusions on both sides, this is unfolded like opening a book, so, a two-layer substrate, and some circuitry on top. Circuitry which might not be at all functional while the thing is conducting, but, I mean, not like it has to be, hehe. Maybe it's even just a light diffusion, as seems likely given the example patent, and the apparent lack of mesa etch here. \$\endgroup\$ Dec 3, 2023 at 20:31

Your Answer

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

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