Solid-State Relays with Arno Holschuh
/By Arno Holschuh
I distinctly remember the first time I saw a solid state relay. It was the early 2000s, and my employer had just had their machine converted to digital temperature control. At the time, this was extremely cutting edge, sort of in the same ballpark as clones and flying cars.
We gazed into the machine, looking at blinking lights, wires, and the temperature probes. I noticed the black box that had been glued to the frame.
“What’s that?” I asked.
“Not sure,” my coworker said. “Let’s not touch it.”
And so we didn’t.
It did its thing, silently switching on and off constantly to keep our brewing temperatures on an even keel. It wasn’t until years later that I thought back to that magic black box.
Why solid-state?
First, let’s take a second to consider the lowly relay. A relay is basically a little switch that closes a big switch. What happens in a traditional relay is that a little switch is used to power an electromagnet; the big switch is closed when that electromagnet pulls the contacts together, usually with a satisfying, audible >click<, or >thunk< if it’s large enough.
It seems like a lot of fuss over switch size, but switch size (and wire size, etc) matter a lot. Consider that the kind of switch that could safely control a car’s starter motor would probably take two hands to comfortably operate. Without relays, our lives would have a lot fewer tiny buttons and a lot more Frankenstein-laboratory knife switches. And relays were an essential building block in creating digital logic. Banks of old-fashioned relays were used to run roaster control cabinets into the 70s. (Opening one of these cabinets sounded like a battalion of power knitters clicking their way to a world record scarf.)
But traditional relays present problems. The inherent weakness is that they have moving parts. All components with moving parts will eventually fail, as the forces of metal fatigue, dirt and corrosion take their toll. The contacts on traditional relays also tend to get fouled with carbon due to arcing (as is the case with the closely related traditional pressurestat).
The solid-state relay, on the other hand, does away with all that motion, and is therefore much more reliable. There are no actual contacts to foul, no electromagnet that can wear out. And solid-state relays can switch incredible quickly, which makes them ideal for systems where a digital controller is trying to allow many tiny little bursts of electricity -- like the PID controllers on our espresso machines. The primary downside to a solid-state relay is that when they break, they tend to break in the closed position -- that is, in the position where they are delivering electricity. But it happens so infrequently, and modern machines have such robust safety designs, that this is acceptable.
A wrapped triac
Now that you’ve joined me as an enthusiastic booster of solid-state relays, let’s consider the underlying question: What the heck are they? The following is kind of technical, but bear with me.
Most of the solid-state relays we as coffee technicians encounter are designed for alternating current, so that’s what we are going to consider here. These relays are designed around a triac, which is itself a component that uses a small amount of power to switch a large amount of power. But in order to be useful in today’s equipment, the triac usually has to be “wrapped” with components that make it compatible with low-voltage, direct current circuitry.
You take the high side, I take the low side
Think about a modern espresso machine. You have high-voltage, alternating current circuits that power the heating elements, pump(s), and solenoids. But the brain itself is running off of low-voltage direct current; logic circuits always do. And the temperature sensors and flowmeters also use low-voltage direct current to do their work. Basically, every espresso machine made today has two “sides” to its electronic design -- a high-voltage AC side, and a low-voltage DC side.
The solid-state relay plays a huge role in translating between these two. It will be activated directly by the brain through application of low-voltage DC power, but the power it delivers is high-voltage AC. One could feed a triac DC straight from the brain, but in doing so, the AC circuitry would feed back a tremendous amount of electrical “noise” into the logic circuits. Without dragging us any further down this technical rabbit hole, let me just say that such “noise” will cause every one of your sensors to read incorrectly. Not good.
Better logic through isolation
So the triac is “wrapped” in optical isolation. When DC power is applied to the relay, it is used to light up a little LED. That light is sensed by a phototransistor that then applies AC power to the triac, which allows all that sweet, sweet AC to flow to the heating elements. (This directly results in millions of lattes for customers and personal fame and fortune for all involved.) No wire connects the low-voltage system and the high-voltage system, and that is the key.
This scheme has other major advantages for logic stability and reliability -- zero switching, anyone? But I kind of feel like I’ve already written a forbiddingly technical article. So once again, the main takeaways:
Relays are like switches... for switches!
Solid state is vital for reliability and speed.
Espresso machines actually have two electrical systems, high- and low-voltage.
Noise bad, isolation good!