# Why are bulky SCR power controllers used when the same can be achieved with a relay?

This is kinda noob question. I recently got an SCR Power controller (WATLOW DIN-a-mite DB20-24CO-0000 Style B). This is the first time I actually got to test an SCR controller for myself. The input trigger is 4VDC to 32VDC, which I powered with a fully charged Li-ion at 4.1V . AC input:115VAC from 12V inverter and out is a 115V digital panel meter (I wanted to try first with a low powered device) though its meant for a heater.

As per the specs the SCR is rated for 25Amps. Similar or even higher rated current relays are available and much smaller in size.

So,is there any specific reason for using an SCR power controller over a higher current relay/contactor? Also, is an SCF power controller same as an Solid State Relay?

• I could give you a bag of parts and say.. here is your computer, use that. The SSPC is a complete product, not just a component. – Trevor_G Nov 13 '17 at 20:01
• Yes ,Trevor I'm aware that this is a complete package and the SCR is the main component. What bugged me is why this is preferred over a relay/contactor? – The_Vintage_Collector Nov 13 '17 at 20:06
• Because it has a bunch of added value safety features. – Trevor_G Nov 13 '17 at 20:12
• The question you need to ask yourself is, for your application, do you need all that or will a simple relay or SSR be enough. – Trevor_G Nov 13 '17 at 20:14
• I'm not using this for any specific application now maybe in future. I just wanted to know what advantage would this offer if I went with the SSPC. Also I'm slightly confused about the controlling part that Mike has mentioned below. What factor needs to be varied to achieve that. For, now in my testing the Vin was a constant 4.1V input trigger. Will it change the duty cycle if the keep increasing it? – The_Vintage_Collector Nov 13 '17 at 20:23

So,is there any specific reason for using an SCR power controller over a higher current relay/contactor? [From comments: What I meant to ask is if a simple SCR triggered controller (ignoring the above example) is better than a relay?]

• No moving parts.
• Usually can switch very frequently (although the datasheet for the part you chose suggests that these should run a 3 s duty-cycle minimum).
• Zero-cross turn-off. One of the characteristics of SCRs (including triacs) is that once triggered they remain on until current falls below the hold-on value at the next zero-cross.
• The possibility of zero-cross switching at the turn-on point. This type of SSR will, when triggered, wait until the next zero-cross before turning on. Both this and the zero-cross turn-off result in much reduced electro-magnetic interference (EMI).
• The possibility of dimming. See Figure 1.

Figure 1. With non zero-cross SSRs dimming is possible.

• Proportional power is still possible with zero-cross SSRs but on a longer time scale. This is usually more than adequate for heater control where the time constants are long. See Figure 2.

Figure 2. Proportional on-off time control. Note that step size is one half-cycle minimum. This can make the response seem coarse if the repeat time is short.

• SSRs are silent.

Also is an SSR almost same as SCR triggered controller?

simulate this circuit – Schematic created using CircuitLab

Figure 3. A very simple SSR (a), a thyristor (b) and a triac (c).

An SSR (a) will have electrical isolation between the trigger circuit and the actual SCR. An SCR controller might not and, at its simplest, might just be a thyristor or triac. Note that in Figure 3a I have shown a constant current source to represent the internal circuitry that allows the SSR to work over a wide range of input voltages - 4 to 32 V in your example.

The relay will provide "all on" or "all off" but the scr will be able to control the supply through a defined controllable range from 0 to 100%.

• So, for a relay as its powered it passes 100% current. But the SCR gets triggered only at the rated min., here 4VDC and can go all the way uptown 32VDC, below that the output at pin 5&6is only 24VAC. So, did you mean that if I vary my voltage from 4-32V I can control the current? – The_Vintage_Collector Nov 13 '17 at 20:05
• I think that if you read the datasheet you will find that there are several models. The 4-32V models are on/off type at zero-cross only. They don't dim. The 4-20mA models seem to offer dimming. They also have models with triac short-circuit detection. – Transistor Nov 13 '17 at 20:57
• @Transistor In page 5 of the datasheet "input wiring" section it shows the same pins 7 & 8 can be used with mA current too. Does it mean I can vary the output power too? – The_Vintage_Collector Nov 14 '17 at 13:30
• @RahulSalin: I think that applies to specific models only. – Transistor Nov 14 '17 at 13:32
• @Transistor Sorry mate, my mistake. I didn't read carefully. You are right I read it again and it is actually about the current draw at different voltages from 4-32VDC, not about current control. This model doesn't have current control. Also your explanation is orderly and easily understood with figures and waveforms.Thanks:) – The_Vintage_Collector Nov 14 '17 at 18:55

Your SSR Is voltage controlled power by adjust phase of the TRiac which is a low cost way of proportional control.

However heaters have long time constants and can also be controlled on-off with small hysteresis.

The linear controllers prevent the hystereies when used with a setpoint thermistor. You can also adjust gain so this can improve regulation. Too much sensor drift will cause it go full on off with high gain and too little gain will have more overshoot when a door closes or more lag when a cold door opens.

Unlike just a ZCS Triac or Relay, this unit allows external PID compensation for ideal temperature control under with . disturbances. A relay has much lower life cycle than SSRs depending on load rating fros and quality of relay about 50k cycles MTBF... How many cycles per hour you expect depends on the hysteresis. ( 0.5 to 1 'C )

However I assumed proportional control in above remark. THe best controllers include PID signal conditional on temp feedback vs setpoint error.

Solid State Power Controllers may provide controlled rise and fall times (trapezoidal control to help reduce radiated emissions), digital processing features, measurement features, higher speed short-circuit protection, plus specifiable specific action (melting) integral (often just written as $I^2\:t$) protection. This last one helps protect the wiring, loads (and of course the controller itself) from overheating. Some support cycle by cycle control, or phase angle control, so that you can use them as a kind of "dimmer" of sorts (if the load is appropriate for that kind of control.)

A Solid State Relay is just the "guts" so to speak of the SSPC, which is the actual switching part inside without all the fancy features. It requires external control. But if all you want is a replacement for a relay, and you don't need all the fancy features of a SSPC, then these are an alternative to a relay.

Relays have contact arcing and oxidation and erosion and welding issues, plus everything that goes together with physical, moving parts -- longer delays and contact bounce, etc.

Relays can have far less dissipation than either the SSR or SSPC because they are just a controllable mechanical switch. So very low losses in the switch itself. Their main loss is the coil. Which is probably under $2\:\textrm{W}$ (and sometimes significantly under that.) Unlike the SSR or SSPC, relay losses are fixed (coil power) and are not dependent on the load they are designed to operate under.

It's possible arrange a hybrid of both a relay and an SSR, where you activate both. The SSR will be faster and will not have arcing problems. Then the relay will engage, but because the SSR is already active there almost won't be any arcing on the contacts and bounce is irrelevant, as well. So that's nice. Once the relay has finally engaged, it takes over and the SSR no longer dissipates energy. So heat sinking for the SSR isn't nearly so much a concern and can be a lot smaller. I doubt SSMCs can use a hybrid approach inside themselves and still maintain many of their other features.

In short, there are options. And that's a nice thing, I suppose. It's just a matter of deciding what's better for the situation.

Why avoid relays? Usually because of cost/size versus limited life and failure mode.

SCRs are small and inexpensive, but are vulnerable to line surges and spikes. Protection circuitry and CPU-monitoring can be included.

A small, inexpensive relay will have limited life, as the contacts are slowly eroded by arcing during opening, and by surface pitting/vaporization during closure. Eventually the relay will become erratic, either refusing to close (oxide fragments in the way,) or when the corroded contacts spot-weld together. Until this occurs, an inexpensive relay may be more reliable than an SCR, so its added protection/monitoring isn't necessary.

The relay life can be greatly extended by using more-expensive, physically large relays (having high closure force and thick, wide contacts.) Will such a device be too large to fit inside your controller? Is your controller handling multiple outputs, or just one?

Small separate issue: relays have unsyncronized AC timing and inductive kick, and they create electrical noise, but SCRs can use low-noise, zero-crossing switching. Some sensitive applications (research labs, etc.,) may preclude use of relay-based controllers.