Snubber circuit — is it necessary and what values to use?

Hardware & Wiring
1

From a circuit perspective, I have a question about snubber circuits on the rbdimmer modules.

I can see from the PCB that there’s already an RC snubber across the TRIAC. Several tutorials and application notes recommend adding an external snubber — though most of those are written for people building their own DIY TRIAC dimmer circuits from scratch, not for pre-assembled modules like rbdimmer.

My use case: I’m dimming a mix of loads — mostly LED ceiling lights (resistive from the TRIAC’s perspective once through the driver), but I also want to control a 150W universal motor (desk fan) on one channel.

Questions:

  1. What are the exact values of the built-in snubber on the rbdimmer modules?
  2. For which load types is the built-in snubber sufficient?
  3. When would I need an additional external snubber, and what values should I use?

I ask because the BTA16 datasheet specifies a dV/dt limit of 50 V/μs at commutation. If the rate of voltage rise exceeds this, the TRIAC can false-trigger — turning on when it shouldn’t. The snubber’s job is to limit this dV/dt to a safe value.

2

Good question. I’ve been investigating this topic as well because I had false triggering issues with a 500W transformer load on a BTA16-based dimmer.

The key parameter is commutation dV/dt — the rate of voltage rise across the TRIAC when it turns off at current zero-crossing. For a purely resistive load, voltage and current are in phase, so the TRIAC commutates cleanly and dV/dt is moderate. The built-in snubber handles this easily.

For inductive loads it’s different. The current lags the voltage by the phase angle φ. When the current crosses zero (TRIAC turns off), the voltage is already at some non-zero value. The rate of voltage reapplication across the TRIAC is much steeper — potentially exceeding the 50 V/μs limit.

I measured this with my oscilloscope on a 500W transformer load without external snubber. The dV/dt at commutation was around 35 V/μs — still below the BTA16 spec but uncomfortably close. With a 100Ω + 100nF external snubber across the load, it dropped to about 12 V/μs.

For LED loads with electronic drivers, the power factor is typically 0.9+ (nearly resistive), so the built-in snubber is more than sufficient.

3

That matches my theoretical analysis. Let me work through the numbers.

For a resistive load on 230V/50Hz mains, the peak dV/dt at zero-crossing is:

dV/dt = Vpeak × ω = 325V × 314 rad/s ≈ 102 kV/s = 0.1 V/μs

That’s well below the 50 V/μs limit — no snubber needed at all for pure resistive loads. The built-in 39Ω + 100nF just adds extra margin.

For an inductive load with power factor 0.5 (φ = 60°), the voltage at current zero-crossing is:

V = Vpeak × sin(φ) = 325V × sin(60°) ≈ 281V

The reapplied voltage step is much larger and faster, which is where snubber values matter. The snubber RC time constant limits the rate of voltage rise:

τ = R × C = 39Ω × 100nF = 3.9μs (built-in)
τ = 100Ω × 100nF = 10μs (with additional external snubber)

The larger time constant with the external snubber gives a much gentler voltage reapplication curve.

4

Thanks for the official specs. The X2 capacitor requirement is critical — worth emphasizing.

For anyone sourcing components, the typical part numbers for X2 100nF capacitors are:

  • EPCOS B32921C3104K (100nF, 305VAC, X2)
  • Vishay MKP3386B (100nF, 275VAC, X2)

The resistor should be a metal oxide or wirewound type rated for pulse handling — standard carbon film resistors can fail under repetitive surge conditions. A 100Ω 1W metal oxide (like Yageo MO-100R) is a safe choice.

5

Agreed on the component selection. One additional point from my testing:

The snubber placement matters. I tested three positions:

  1. Across the TRIAC (MT1-MT2) — most common, reduces dV/dt directly
  2. Across the load — reduces voltage spikes at the load, slightly less effective at protecting the TRIAC
  3. Both positions — best protection but adds cost and complexity

For my 500W transformer setup, position 1 (across TRIAC) was sufficient. The oscilloscope showed clean commutation with no false triggering at any dimming level.

For motors specifically, I’d recommend across the TRIAC. The motor’s own back-EMF creates additional voltage spikes that the snubber needs to absorb, and those spikes appear at the TRIAC terminals first.

6

That’s a good practical comparison. From a circuit perspective, placing the snubber across the TRIAC is always the first choice because that’s where the dV/dt limit applies.

Let me also calculate the power dissipation in the snubber resistor for sizing purposes:

P = C × V² × f

For the built-in snubber (100nF, 230V, 50Hz):
P = 100×10⁻⁹ × 230² × 50 ≈ 0.26W — handled by the onboard resistor

For the external snubber (100nF, 230V, 50Hz):
P = 100×10⁻⁹ × 230² × 50 ≈ 0.26W — a 0.5W resistor handles this with margin

So the general conclusion:

  • Lamp dimming (LED or incandescent): built-in 39Ω + 100nF is sufficient, no external snubber needed
  • Motors above 200W or transformers: add external 100Ω + 100nF (X2 cap, 0.5W+ resistor) across the TRIAC — but check if the motor already has its own snubber first
7

Well summarized. I’ll add one final note for completeness:

If you’re unsure whether your load needs an external snubber, the simplest test is to set the dimmer to minimum brightness and observe the lamp or motor. If you see random flashes or the motor twitches when it should be off, that’s likely false triggering from excessive dV/dt — and an external snubber will fix it.

No oscilloscope required for the diagnostic — just watch the load at minimum setting. If it behaves cleanly at 1-5% power, your built-in snubber is doing its job.