TXVs vs. EEVs HVAC

Page 1 named the metering device as one of the four components of the refrigerant cycle and left its insides for later. Page 2 showed that the metering device is what a unit controller leans on to hold superheat at a target — it's not a fixed orifice, it's a feedback element. This page opens the box. Two common designs sit behind the same job description: a thermostatic expansion valve (TXV) does the work mechanically with a sensing bulb on the suction line, and an electronic expansion valve (EEV) does it electronically with a stepper motor and a controller algorithm. Service techs mostly meet TXVs in the field; BMS engineers mostly meet EEVs in the controller. Knowing which one is which — and how each decides how wide-open to be at any moment — is the cleanest way to keep their behavior straight.

What the metering device does

A quick recap. The high-side liquid leaving the condenser is warm and pressurized; the evaporator on the low side needs cold, low-pressure refrigerant to boil and absorb heat. The metering device sits at the boundary and throttles the liquid through a small port, dropping its pressure sharply. Upstream of the port: warm liquid at high pressure. Downstream: a cold, low-pressure mix of liquid and flash vapor heading into the evaporator. So far that's just a restriction in a pipe; a fixed orifice or capillary tube does the same thing. What makes a TXV or an EEV different is that the port isn't fixed. Both kinds of valve modulate how wide-open the port is in response to what's happening downstream, and the target they're modulating against is the evaporator-outlet superheat. Too little feed and the coil starves (SH climbs); too much and liquid escapes the coil (SH falls toward zero, then negative). The valve's job is to keep that number sitting at the target.

The TXV — a mechanical control loop

A TXV is a control loop made of metal. Three pressures push on a diaphragm inside the valve body; the diaphragm position decides how far open the port is. The sensing bulb, strapped tight to the suction line just downstream of the evaporator, is filled with a charge of refrigerant (often the same one the system uses). The bulb is connected to the top side of the diaphragm by a thin capillary tube. Whatever temperature the suction line is at, the bulb settles to the same temperature within seconds, and the saturation pressure of the bulb charge at that temperature pushes down on the diaphragm. On the bottom side of the diaphragm, two forces push up: a stiff spring (adjustable on most valves, set at install for a target superheat) and the evaporator pressure itself, fed to the bottom chamber through a small external equalizer line tapped into the suction line near the bulb.

That's the force balance: bulb pressure on top opens the valve; spring plus evaporator pressure on the bottom closes it. The diaphragm settles wherever the two sides match, and the needle hanging off the diaphragm settles into a matching port position. Now follow what happens when superheat drifts. Say the evaporator load drops — cooler air over the coil on a milder day, or a damper closing in the conditioned space. Less heat into the refrigerant means less boiling per pound of liquid fed; liquid lingers further into the coil; the suction line at the bulb runs cooler. The bulb charge cools with it, its saturation pressure drops, the downward force on the diaphragm drops, the spring and evaporator pressure push the diaphragm up, and the port closes a touch. Less refrigerant feed lets the coil catch up — the lingering liquid finishes boiling, suction T rises, and the system settles at the same target superheat with a smaller mass flow. Load swings the other way and the chain runs in reverse: warmer bulb, more downward force, port opens, more feed, evaporator handles the larger load. The valve does this continuously and silently, no power supply, no controller, no setpoint to enter.

A few practical notes ride along with the picture. The spring is what sets the target superheat — bigger spring force, higher SH target, because the bulb has to get warmer (more downward force) to balance it. Service techs turn an adjustment screw on the bottom of the valve to change that spring tension, in fractions of a turn; the manufacturer's data sheet says how much SH per turn. The external equalizer matters more than it looks: without it, the pressure on the bottom of the diaphragm is whatever the pressure is right at the valve outlet, which on systems with much evaporator pressure drop is meaningfully different from the pressure where the bulb actually lives. The equalizer routes the suction-line pressure (at the bulb) back to the lower chamber so the force balance reflects what's happening at the coil outlet, not what's happening immediately past the valve. Older or very small systems may skip it — internally-equalized TXVs exist — but anything with a multi-circuit distributor between the valve and the coil needs an external equalizer.

The EEV — an electronic control loop

An EEV does the same job — modulate the port to hold superheat — but with a digital control loop replacing the mechanical force balance. The sensing parts are a pressure transducer and a temperature sensor mounted on the suction line, both wired into the unit controller (on a packaged piece of equipment) or sometimes into the BMS itself (on built-up systems and chillers). The controller reads both inputs, looks up the saturation temperature for the measured pressure, subtracts to get superheat, and compares to a target value held in a parameter. Whatever direction the error sits in, the controller commands a stepper motor on top of the valve to drive the needle a few steps wider or a few steps tighter. The stepper holds position when power is off — no power, no movement, last commanded position stays put — and a typical EEV port has somewhere between a few hundred and a few thousand discrete steps from fully closed to fully open. That's the resolution the BMS gets to work with.

The control loop is just PID, or sometimes a simpler proportional-with-deadband scheme; the controller manufacturer picks the algorithm and exposes a few tuning parameters (target SH, gain, sometimes an integral time). From the BMS side you see a point list with a few familiar shapes — superheat actual, superheat setpoint, valve position, sometimes a valve position override — and the trend lines look like any other feedback loop. The EEV's advantages over the TXV are the ones electronics tend to have: a target you can change without a wrench, faster reaction to load steps, the ability to hold a different SH at different operating conditions (say, a lower SH at low load to maximize evaporator use), and a position reading that's visible on a graphic. The disadvantages are the ones electronics tend to have: a sensor that drifts is invisible until something downstream complains, a stuck stepper looks like a flat-line on the trend, and a control algorithm that hunts will hunt with no mechanical damping to save it.

The same load swing that closed the TXV's port closes the EEV's port too, just by a different route. Cooler air over the coil → less boiling → cooler suction line → suction temperature drops while suction pressure changes only slightly → the controller's computed superheat falls below setpoint → the algorithm decrements the port-position command → the stepper drives the needle that many steps closed → mass flow into the coil drops → the lingering liquid in the evaporator finishes boiling → SH climbs back to the setpoint. The controller, the bulb, and the diaphragm are all doing the same kind of work; only the medium changes. On a unit-controller trend graphic you can usually watch the loop in motion — valve position changes step by step, superheat hugs the target with a few degrees of ripple, and the relationship between the two should look orderly. If valve position is hard-rail at 0% or 100% with superheat still off-target, the loop has run out of authority — feed isn't the limit anymore (charge is, or the coil load is wrong, or a sensor is lying).

Same job, two different surfaces

If you spend most of your time in the BMS, the EEV is the surface you'll see — valve position as a point, superheat as a point, a setpoint you can change. The TXV doesn't put any of that on a network; from a BMS that's looking at a packaged unit with a TXV, the only refrigerant-side points are usually suction and discharge pressure, and the metering device is implicit in the readings. If you spend most of your time turning wrenches on equipment, the TXV is the surface you'll see — it's the thing in your hand when a system runs the wrong superheat, and the bulb's location on the suction line is one of the first things to check when something's off (loose strap, missing insulation, lost bulb charge, all common). When a controls tech and a mechanical tech compare notes on the same system, the same valve gets described from two angles, and the names don't always line up. Knowing both surfaces lets you bridge that conversation.

A worked example — holding superheat as load drifts

Bring the same R-410A split system from page 2. Steady state on a typical operating day: suction-side gauge reads 118 psig, suction-line thermometer 50 °F, so superheat sits at 50 − 40 = 10 °F. Liquid-side gauge reads 340 psig, liquid-line thermometer 95 °F, so subcooling sits at 105 − 95 = 10 °F. The metering device — a TXV in one variant, an EEV in the other — is the element holding that 10 °F superheat by feeding the evaporator at whatever rate the coil load demands.

A cloud passes, the outdoor temperature drops a few degrees, and the indoor return air follows it down. Heat into the evaporator coil drops with the load. With the metering device's current port position, more refrigerant is being fed than the coil can boil; liquid lingers further out toward the suction header; the suction-line temperature at the bulb starts to fall. Within a couple of degrees of drift — suction T sliding from 50 °F toward, say, 47 °F at the same 118 psig — superheat has fallen from 10 °F toward 7 °F.

On the TXV system: the bulb cools with the suction line. The bulb charge contracts, its saturation pressure drops, the downward force on the diaphragm drops, the spring and evaporator pressure overpower the new equilibrium, and the diaphragm rises by some small amount. The needle rides up with it and the port tightens — fractionally, a sliver of lift out of the needle's total travel. (The needle slides straight up and down with the diaphragm; nothing rotates — "turns" belong to the adjustment screw.) Mass flow into the coil drops to match the new boiling rate, the lingering liquid finishes its phase change, suction T climbs back, and the bulb settles at whatever temperature gives the spring its target counterbalance. The system is back at 10 °F superheat, with a slightly lower mass flow rate, no controller involved, no setpoint changed.

On the EEV system: the suction-line temperature sensor reports a cooler reading; the pressure transducer reports essentially the same 118 psig (a small drop with the lower load). The controller looks up the dew temperature for the current pressure, subtracts, and computes a superheat of about 7 °F against a 10 °F setpoint. The control algorithm decrements the port-position command by a few steps. The stepper drives the needle in, the port tightens, mass flow drops, the coil empties, suction T climbs, and the computed superheat returns to 10 °F. Same outcome as the TXV; the trend graphic shows it as a small step in valve position and a brief dip-and-recovery in superheat.

The other direction works the same way in reverse. The load picks back up later in the afternoon; the coil boils faster than the current feed can supply; the suction line warms past 50 °F; superheat climbs past 10 °F. The TXV's bulb warms, its charge expands, the diaphragm pushes down, the port opens, more feed, SH returns to target. The EEV's controller sees a higher computed superheat, increments the port command, the stepper drives the needle open, same recovery. Both valves are doing classic closed-loop control on the same process variable; what's different is what they use as their controller (a spring + a sealed gas charge, vs. a CPU + a trended sensor) and what surface that controller exposes to the people working on the system.

What this chapter sets up

That closes the three-page chapter. Page 1 built the cycle and the saturation lock; page 2 turned the lock into superheat and subcooling, the two measurements that prove the cycle is running right; this page opened the metering device that holds superheat in place. With those three concepts in hand, the Refrigerant P-T & Superheat Calculator turns from a black box into a tool whose every input and output is something you can place inside the cycle and read against the saturation curve. Bring a suction pressure, a clamp-on suction-line temperature, and a refrigerant name back to that calculator and you've already done the measurement half of refrigerant troubleshooting; the verdict pill just reads off what your numbers mean.