# How does a TRIAC work?

I’m trying to understand the concept of a TRIAC. I’ve heard that a TRIAC is made with 2 SCRs (thyristors) connected in anti-parallel.

But in the diagram I looked at, it looks like two SCRs were connected together at the gates, but one was a P-type gate and one was an N-type gate. If the SCRs are identical but simply anti-parallel, why would a P-type gate be connected to an N-type gate? I read online that thyristors only have a P-type gate. So how could one of them have an N-type gate? I added a picture to show you what I mean.

So to clarify: I’ve heard that SCRs only have a P-type gate, yet clearly the P-type gate of the first thyristor is connected to one of the N-channels of the second one in the picture. So what am I missing?

• Using two SCRs is better (and that means different) than using one TRIAC. Don't imagine there is no difference.
– jonk
Nov 7, 2022 at 8:32
• thyristor made of 4 layers. triac has 5. Nov 7, 2022 at 9:45
• Two anti-parallel thyristors have symmetrical switching behaviour. TRIACs only exist to try and emulate this while reducing cost by using a single slab of silicon and have asymmetrical behaviour as a result. That's why you find TRIACs in cheap consumer goods and you find thyristors in industrial products. Nov 7, 2022 at 15:21

The two-SCR model is not very helpful in understanding how a triac works. It's correct in that there are two parallel inverse PNPN structures that carry the current, but the triggering mechanisms (plural because there are several) are not explained.

Here is a typical triac structure (from the GE manual):

The mechanisms are as follows (voltages wrt MT1):

QI (MT2+, G+) conventional thyristor, p1-n1-p2-n2 main current flow

QII (MT2+, G-) junction gate thyristor with n3 acting as the junction gate. p1-n1-p2-n2 main current flow

QIII (MT2-, G-) "remote gate" p2-n3 injecting electrons collected by p2-n1. p2-n1-p1-n4 main current flow

QIV (MT2-, G+) "remote gate" p2-n2 injecting electrons collected by p2-n1, more forward biasing p2-n1 and triggering the p2-n1-p1-n4 main current flow (this is the dodgiest quadrant, the triac is typically much less sensitive in this quadrant, and sometimes the minimum trigger current is not specified by the manufacturer). For this reason, typically we either use the same polarity on the gate as on MT2 (eg. phase control dimmer) or drive the gate with a negative current regardless of the MT2 polarity.

You'll also note that, unlike the ideal SCR model, there is resistive conduction from gate - MT1. That's not necessarily a bad thing, you don't really want it to be too sensitive.

The combined structure also has more issues with commutation than two discrete SCRs, which can be a problem with inductive loads. There are modifications to the triac design to get closer to the two-SCR commutation ability, marketed as "alternistors". I believe they've tweaked the construction and/or used methods such as gold doping to reduce carrier lifetime.

• Hi. Thanks for your reply. Is a TRIAC literally made of two SCRs? Or is it similar to the transistor explanation(like how people say a transistor is made of two Diodes even though it’s really not)? If it’s literally made of two SCRs, I can’t understand why one is a P-type gate and one is an N-type gate(I mean I get why for positive and negative cycles, but not why the second one has an N-type gate rather than a P-type gate). I read online that SCRs only have a P-type gate. Thanks again for your reply. I’d appreciate any more help, ideally if you would please dumb it down for me. I’m a noob. Nov 7, 2022 at 22:52
• It's a single monolithic block of silicon as in the sketch. P and N regions are created by adding impurities to the silicon. The dashed areas represent metal contacts on the silicon surfaces. The MT2 (bottom, mostly covered with metal) is generally attached to the lead frame. I don't think you need to know how exactly a triac works to apply them in a professional manner, but if you need to know how they work, maybe start with how diodes and transistors work, then look at some of my keywords such as "remote gate". Nov 7, 2022 at 23:03

So if I understood correctly, you are confused about the SCR model of TRIAC, and specifically why the right-hand side SCR has its gate from N material while the left-hand side does from P material, as in SCRs.

SCR itself allows one-direction current flow and accepts positive gate-cathode voltage only. However, the TRIAC allows the current to flow in both directions. That's why the anodes (or Main Terminals) are taken from both P and N materials.

And the gates are taken from both P and N materials so that the TRIAC can be triggered with either polarity gate-cathode voltages or either polarity gate currents.

• Thanks for your reply! Yes, you’re correct in understanding what I expressed. What I’m confused about is that I read that an SCR only has a P-type gate. So how can the one on the right have an N-type gate? I underhand the need for it(for positive and negative voltage cycles), but I don’t understand how the right SCR has an N-type gate. Are some SCRs N- type gate controlled? Or is the TRIAC not actually 2 SCRs(sort of how people say a transistor is made of 2 diodes even though it’s actually not: they use that example for simplicity. Thanks again for your help! Could you dumb it down please? Nov 7, 2022 at 23:01
• @theguineapigking well, that's the confusing part of two-SCR model actually. Because although it makes the explanation easier, the construction is different. Now think about the transistor equivalent of an SCR. The P-gate is taken from the base of the NPN transistor. This is the case for ordinary SCRs and therefore for the left-hand side SCR. But for the right hand side SCR, the N-gate is actually taken from PNP's base (i.e. the junction of PNP's base and NPN's collector). Remember that this is not the case for an ordinary SCR but for the model this is what it is. With this configuration ... Nov 8, 2022 at 11:08
• ... the TRIAC can be triggered with negative gate pulses because PNP wants negative Vbe voltage to turn on. Now the right-hand side is vertically mirrored. Remember the transistor equivalent again. For the right-hand side SCR, the emitter of the NPN is connected to MT2, and the emitter of the PNP is connected to MT1. A positive gate voltage (w.r.t. MT2) will turn the left-hand side SCR on, current will flow from MT1 to MT2. A negative gate voltage (w.r.t. MT1) will turn the right-hand side SCR on, current will flow from MT2 to MT1. Check Sphero's answer for MT1/MT2 polarity and current flow. Nov 8, 2022 at 11:12