Relay, Transistor, or Triac? Understanding PLC Output Types with Live Circuit Diagrams

When you pick an output unit for a PLC, the catalog always presents the same three words: relay output, transistor output, and triac output. All three do the same job ── they turn your program’s ON/OFF into a signal for external devices ── yet choosing the wrong one produces strange behavior in the field. A lamp keeps glowing faintly after you switch it OFF. One output stops working after a few years while its neighbors are fine. A transistor output dies the instant a solenoid valve is switched off. None of these are defects or bad luck: every one of them follows directly from what kind of switch each output type has inside.

This article sorts out the three output types from that single viewpoint ── what the switch inside really is. Here is the one-line answer up front: a relay output is a mechanical contact, a transistor output is a DC-only semiconductor with a fixed direction, and a triac output is an AC-only bidirectional semiconductor. Once that line clicks, the catalog differences ── response time, lifespan, supported power ── and all the troubles above can be understood from the same root. And in this site’s tradition, every diagram in this article is interactive: press the buttons and the current in the circuit actually starts and stops, so you can verify each claim with your own eyes as you read.

What this article covers

Here is how common questions map to the sections that answer them.

Question	Section
What actually differs between the three output types?	§1
How does a relay output work, and what is it good and bad at?	§2 (interactive figure)
Why is a transistor output DC-only?	§3 (interactive figure)
What is a triac output for?	§4 (interactive figure)
How do I choose in practice?	§5 (comparison table)
What happens when I switch off a solenoid or relay coil?	§6 (interactive figure)
What is an interposing relay, and why is it used so much?	§7 (interactive figure)
Why won’t the lamp turn off even though the output is OFF?	§8 (interactive figure)
How do I check and isolate problems in the field?	§9 / FAQ

1. The three output types differ in “what the switch inside really is” ── the big picture first

Strip a PLC output circuit down to its essence and it is “a switch operated by the program”. When an output bit (Y0 and so on) turns ON, a switch inside the output unit closes, and current flows through the external load ── a lamp, a relay, a solenoid valve. That much is common to all three types. The only thing that differs is what is used as that switch.

Type	The switch inside	In one phrase
Relay output	Mechanical contact moved by an electromagnet	Metal contacts physically touch and separate
Transistor output	Semiconductor (transistor)	DC-only electronic switch with a fixed direction
Triac output	Semiconductor (triac ── a bidirectional thyristor)	Electronic switch that passes AC’s back-and-forth current

Once the switch is fixed, the properties follow almost automatically. A mechanical contact does not care which way current flows, but it is a moving part ── slow, and it wears out. A transistor is fast and never wears, but its current direction is fixed, so it is DC-only. A triac passes AC in both directions, but as we will see later, it has one quirk: it never switches off completely. The next three sections examine each type with an interactive figure.

A shared rule for reading the figures: everything left of the vertical boundary is external (power supply and load); everything right of it is the PLC side. The terminal labels 24V / AC / Y00 / COM / 0V mark the positions that correspond to a real terminal block. Greenish wires carry positive voltage, gray wires sit near 0V, and the moving dots show the direction and speed of the current.

2. Relay output ── a mechanical contact switches AC and DC alike (Figure 1)

A relay output energizes a small electromagnet (coil) inside the PLC, which physically closes a mechanical contact. The program turns Y0 ON → current flows through the internal coil → the contact snaps shut → the external circuit is completed. In the figure below, the “Logic Y0” button plays the role of the program’s output command. Turn it ON and the contact closes, letting current flow along +24V → external load → terminal Y00 → contact → COM (0V), and the load lamp lights.

A contact is just two pieces of metal touching, so it does not care about the direction or the kind of current. The figure switches DC 24V as a representative example, but the same contact can switch an AC 100–240V lamp (on real hardware, within the contact’s voltage and current ratings). This “AC or DC, and different voltage classes too” flexibility is why relay outputs remain the default choice for general-purpose units. The coil side (inside the PLC) and the contact side (external circuit) are also electrically isolated from each other, which helps protect the equipment.

The weaknesses come from the very fact that something mechanical moves. Because the contact physically travels, response takes roughly 5–15 ms ── nowhere near semiconductor speed. And every operation wears or roughens the contact surfaces, so the lifespan is finite: as a rule of thumb, hundreds of thousands to a few million operations with a resistive load, and orders of magnitude fewer if you switch inductive loads without protection (see §6). At 100 operations a day that lasts many years; drive it several times a second and it can be worn out within months.

3. Transistor output ── a DC-only switch with a fixed direction (Figure 2)

A transistor output replaces the mechanical contact with a semiconductor switch. The figure below shows the NPN (sink) type. Turn Logic Y0 ON and the transistor inside the PLC conducts, pulling output terminal Y00 down toward 0V. Current flows in along +24V → external load → terminal Y00 → inside the PLC → 0V, and the indicator lamp lights. Note what happens at OFF, too: the terminal voltage floats up near the supply voltage (C1 ≈ 24V in the figure).

A transistor conducts in one direction only, so this type is strictly DC. It cannot handle AC, whose current reverses direction constantly. In exchange, with no moving parts, response is around 0.1 ms ── orders of magnitude faster ── and with no contacts to wear, the lifespan is effectively unlimited in operation count, as long as electrical stress (overcurrent, overvoltage, heat) is kept under control. Pulse trains for stepper and servo drives, and anything that switches tens of times per second, are territory where the transistor output has no rival.

Transistor outputs come in two flavors, NPN (sink) and PNP (source), with opposite current directions. All figures in this article use NPN. The sink/source distinction and how to match polarities with the receiving device are covered, with the same style of interactive figures, in the companion article “Understanding PNP and NPN in Electrical Design with Live Circuit Diagrams”. If polarity matching is your question, start there.

4. Triac output ── a semiconductor switch that quietly handles AC (Figure 3)

“I need to switch AC ── and frequently.” For that requirement, a relay runs out of lifespan and a transistor points the wrong way. This is where the triac output comes in. A triac is a semiconductor switch that conducts in both directions, so it handles AC’s alternating current as-is. In the figure below, the AC source is at the top left and the return (COM) at the bottom left. Turn Logic Y0 ON and the path conducts ── watch the current commute back and forth through the load.

The figure simplifies the triac’s internals ── gate drive, holding current ── and presents it as “a switch that opens and closes AC on a logic command”. The indicator lamp is modeled with a resistor rather than an LED: an LED passes current in one direction only and would hide the AC round trip. The “both ways” note in the status line is reading the magnitude of that alternating current.

A useful way to place the triac output is “the AC counterpart of the transistor output”. No mechanical contact means no operation-count lifespan, and it responds faster than a relay. In exchange it is AC-only ── it cannot switch DC ── and it carries one semiconductor quirk of its own: a small current keeps flowing even when OFF, called leakage current. That quirk is a field-trouble classic, so it gets its own section with an interactive figure in §8.

5. How to choose ── the load’s power type and switching frequency decide almost everything

With all three characters introduced, here is the comparison table.

	Relay output	Transistor output	Triac output
The switch inside	Mechanical contact	Semiconductor (for DC)	Semiconductor (for AC)
Supported power	Both AC and DC	DC only	AC only
Response speed	Slow (about 5–15 ms)	Fast (around 0.1 ms)	In between (bound to the AC half-cycle)
Switching lifespan	Finite (hundreds of thousands to millions of ops)	Effectively unlimited	Effectively unlimited
Current per point	Larger (about 2 A)	Smaller (about 0.1–0.5 A)	In between (about 0.6 A)
High-speed pulse output	✗	✓	✗
OFF-state leakage current	None (a true open circuit)	Almost none	Present (a few mA ── beware)
Main caution	Contact wear on inductive loads (§6)	Surge destruction (§6), polarity (§3)	Leakage current (§8), heat

Selection boils down to two questions: “Is the load AC or DC?” and “How often does it switch?”

1. Load is DC with high-speed pulses or frequent switching → transistor output
2. Load is AC with frequent switching → triac output (for small loads, check the leakage countermeasures in §8)
3. Switching is infrequent, AC and DC are mixed, or future loads are undecided → relay output
4. High-current loads, or loads on a different voltage → put an interposing relay (§7) in between, whatever the output type

A policy you will often meet in practice: standardize on transistor outputs, and receive AC or high-current loads through interposing relays. The output unit stays fast and long-lived, while the interposing relays absorb the variety on the load side. Whichever policy you adopt, the justification always comes back to the same two questions.

6. What happens the instant you switch off an inductive load? ── surge and the flyback diode (Figure 4)

Loads that contain a coil ── relay coils, solenoid valves, solenoids ── share a property: their current refuses to stop instantly. The moment the output turns OFF, the current that has lost its path reappears as a high voltage spike at the output terminal ── the surge (back-EMF). Transistors are fragile against this surge. Connect an inductive load with no protection, and one day that single output point dies without warning ── a textbook failure.

The standard countermeasure is a flyback diode (freewheeling diode) in parallel with the load. At the OFF instant, the coil’s remaining current circulates through the diode and burns off as heat, suppressing the surge at its source. In the figure, Logic Y0 switches the output, and the second button inserts or removes the flyback diode. Try this sequence:

1. Leave the diode “Off”, turn Logic Y0 ON → OFF, and watch the C1 voltage spike at the OFF instant (a surge warning appears)
2. Set the diode to “On” and repeat ── the spike at the OFF instant is now suppressed

This surge is no stranger to relay outputs either. Break an inductive load with a contact, and the surge becomes a spark (arc) across the gap, accelerating contact wear by orders of magnitude ── this is why §2 said unprotected inductive loads shorten relay life “by orders of magnitude”. The base rule: give DC inductive loads a flyback diode, and give AC inductive loads an RC snubber (surge killer) or a varistor, mounted at the load.

7. The interposing relay ── a transistor drives the coil, the contact switches the AC (Figure 5)

Combine everything so far and one classic field pattern falls out by itself. “A transistor output is fast and long-lived, but DC-only and low-current.” “A relay contact switches AC and heavier currents.” So: let the PLC’s transistor output drive the small DC coil of a relay, and let that relay’s contact switch the AC load. The relay playing this bridging role is called an interposing relay.

In the figure below, the upper row is DC (coil side) and the lower row is AC (contact side). Turn Logic Y0 ON: the transistor energizes the coil, the contact below closes in step, and current flows through the AC indicator lamp. Watch how one command makes the small DC circuit and the AC load circuit move together. Following the lesson of §6, the coil comes with a flyback diode.

The interposing relay buys you more than “now it can switch AC”. Contacts are consumables that eventually wear out ── but with an interposing relay, maintenance means pulling a relay worth a few dollars out of its socket and plugging in a new one, which spares the PLC output unit itself. It also acts as insurance: if the load side suffers a short circuit, the damage tends to stop at the interposing relay. The price is response time ── one relay stage adds roughly 10 ms ── plus more parts and wiring in the cabinet. For AC or high-current loads that do not need speed, the benefits win almost every time.

8. The lamp glows faintly even when OFF ── triac leakage current (Figure 6)

Triac outputs differ from mechanical contacts in one decisive way. A contact’s OFF is a physical break ── current becomes exactly zero. A semiconductor’s OFF is merely “conducting very little”: a small leakage current keeps flowing. It amounts to a few milliamps. Rounding error for a big load ── but not for a small one.

The figure reproduces this with two paths: the main conduction path switched by Logic Y0, and a thin, permanently connected leakage path beside it. Leave Logic Y0 OFF and watch the load current: it refuses to drop below roughly 1 mA, and the status reads “leakage (persists at OFF)”. Turn it ON and the main path takes over, with current orders of magnitude larger.

In real equipment, those few milliamps are known for two symptoms. First, ghost-lit pilot lamps: LED indicators start to glow at anywhere from a few mA to 20 mA, so leakage alone can keep a lamp faintly lit while “OFF”. Second, small relays and solenoid valves that fail to release: if the leakage exceeds the load’s drop-out current, the load can never return fully to OFF after the command is removed.

The standard fix is a bleeder resistor (dummy resistor) in parallel with the load, giving the leakage a harmless path. Typical values are a few kΩ to a few tens of kΩ, rated from 1/2 W up to a few watts on 100–120 VAC circuits, chosen so most of the leakage flows through the resistor. For small loads that still misbehave ── or circuits that genuinely require a true open circuit at OFF ── reroute that load to a relay output (or an interposing relay). That settles it.

9. Checking in the field ── translating the figures into multimeter moves

Finally, here is everything above translated into an isolation procedure for “the output does not work / does not turn off”. Working from the top is fastest.

1. Check the program’s output indicator ── is the output bit really turning ON/OFF? If not, the problem is in the program and interlock conditions, not the wiring
2. Check the output unit’s status LED ── does it follow the command? If the LED responds but the load does not, the problem is downstream: wiring, load, or a type mismatch
3. Match the output type against the load’s power type ── AC on a transistor output, or DC on a triac output? Verify with the physical part number, not the drawing
4. Measure the terminal voltage ── does it move with the command? An NPN transistor output drops near 0V at ON (sink/source reading guide in the companion article)
5. If the symptom is “won’t turn off”, suspect leakage ── a triac/SSR with a small load points to §8: add a bleeder resistor, or move the load to a relay output
6. If the symptom is “that one point keeps dying”, suspect surge ── is that load inductive? Check the §6 protection parts, and their polarity
7. On relay outputs, count the operations ── an old cabinet with one flaky output suggests contact wear: replace the unit, or convert the point to an interposing relay

Summary ── know the switch inside, and both selection and trouble become readable

• The three types differ in what the switch inside really is: relay = mechanical contact, transistor = DC semiconductor, triac = AC semiconductor.
• Relay outputs take AC and DC alike with generous current ratings, but they are slow, and their life expires in operations.
• Transistor outputs are fast and long-lived but DC-only and low-current ── and remember the sink/source polarity.
• Triac outputs excel at frequent AC switching ── but leakage current remains at OFF.
• Selection reduces to two questions: is the load AC or DC, and how often does it switch?
• Inductive loads get protection parts: a flyback diode for DC, a snubber for AC. This saves transistors from destruction and relay contacts from erosion alike.
• For high current, AC, or easy replacement, receive the load through an interposing relay ── the field standard.
• Symptoms map back to the quirks: “won’t turn off” → leakage, “that point keeps dying” → surge, “flaky after years of service” → contact wear.

If the catalog words “relay / transistor / triac” now read as three kinds of switch ── mechanical contact / DC semiconductor / AC semiconductor ── this article has done its job. Whenever a selection decision or a strange field symptom comes up, the interactive figures here will be waiting.

FAQ

Q1. The difference between relay, transistor, and triac outputs in one sentence?

A. The switch inside is a mechanical contact (relay), a DC-only semiconductor (transistor), or an AC-only semiconductor (triac). Every difference in speed, lifespan, and supported power derives from that.

Q2. Which one should I pick if I am unsure?

A. Narrow it down by the load’s power type (AC or DC) and switching frequency. DC with fast or frequent switching → transistor; AC with frequent switching → triac; infrequent switching with mixed or undecided loads → relay. When extending an existing cabinet, matching the existing type and common wiring takes priority.

Q3. How long does a relay output actually last?

A. As a rule of thumb, hundreds of thousands to a few million operations on resistive loads ── and orders of magnitude fewer on unprotected inductive loads. The life is counted in operations, not time, so estimate it as operations per day × working days.

Q4. Can a transistor output switch an AC load?

A. No. A transistor is a DC-only switch with a fixed current direction. For AC loads, use a triac or relay output, or the transistor-plus-interposing-relay arrangement of §7.

Q5. Can a triac output switch a DC load?

A. No. A triac turns off by exploiting the instant the current crosses zero (the AC zero cross). On DC the current never reaches zero, so once triggered ON, it cannot turn off. Handle DC with a transistor or relay output.

Q6. Do I need a flyback diode with a relay output too?

A. Not strictly for protecting the output device, but it greatly extends contact life. The base rule: a flyback diode for DC inductive loads, an RC snubber or varistor for AC inductive loads, fitted at the load. Arc erosion of the contacts drops by orders of magnitude.

Q7. The lamp stays faintly lit after switching OFF. Is the unit broken?

A. If a small lamp hangs on a triac output or AC solid-state relay, it is almost certainly leakage current (§8) ── a spec, not a fault. Add a bleeder resistor of a few kΩ to a few tens of kΩ in parallel with the load, or move that load to a relay output.

Q8. Are there downsides to interposing relays?

A. Response gets slower by one relay stage (roughly 10 ms), and the cabinet gains parts, wiring, and space usage. In exchange, for AC or high-current loads that do not need speed, easy contact replacement and protection of the PLC itself outweigh the costs in most cases.

Q9. How do NPN and PNP relate to transistor outputs?

A. NPN (sink) and PNP (source) are the two current directions available within the transistor output type. If the polarity does not match the receiving device’s common, two perfectly healthy devices will not work together. The companion article “Understanding PNP and NPN in Electrical Design with Live Circuit Diagrams” covers this with the same style of interactive figures.

Q10. How do these interactive figures differ from real hardware?

A. Real hardware adds wiring resistance, contact bounce, the triac’s gate drive and holding current, protection circuits, and temperature effects; the figures strip these away to show the skeleton of the principle. Use the figures to understand the principle ── and manufacturer datasheets and standards for real design numbers.