Why Control Power Fails Differently Than Load Power

Why this matters

Load power tells you when it is gone. A conductor is open or it is not, a device is tripped or it is not, and the machine either runs or sits there. Control power has a wide middle region where the voltage is present and the energy is not, and inside that region the equipment does something far worse than stop: it reports a false story about itself. A contactor that will not pull in looks like a bad contactor. A sequence that stalls at step four looks like a bad board. The meter says everything is fine because the meter is asking the wrong question, and the shop sells three parts before anybody measures the circuit under the condition that actually fails.

The two failure shapes, side by side

Every row of that table comes from the same root: a control circuit carries so little current that the ordinary tools and instincts developed on load circuits stop resolving anything.

Why a small resistance matters here and nowhere else

Voltage lost across a connection is the current through it times its resistance. On a load circuit carrying tens of amperes, a connection with meaningful resistance gets hot and announces itself by discoloring, smelling or opening. On a control circuit drawing a fraction of an ampere, that same resistance dissipates almost nothing, never gets warm, never discolors, and drops a voltage nobody would notice.

Then the coil picks up. A magnet coil presents a much larger burden with its armature open than with it closed, because the open air gap leaves coil inductance at its lowest, and the sibling reference on what a control transformer is sized for carries that mechanism and the sizing method that follows from it. The consequence here is the point: the same connection that dropped a fraction of a volt at sealed current drops several volts at pickup current, and it does so for exactly as long as the coil is trying to close.

A coil needs a minimum voltage to pull in, commonly around 85 percent of its rated coil voltage, with the specific figure on the device data sheet. Above that it snaps closed and the burden collapses. Below it the armature hangs and the burden stays high. There is no partial pull, which is why the symptom is a buzz, a chatter or nothing at all rather than a slow, weak close.

Why your meter agrees that everything is fine

A digital multimeter has an input impedance in the megohm range, so it draws microamperes. Across a series resistance of a few ohms, microamperes develop a drop of microvolts, which is to say nothing. The meter therefore reports the potential available at that point with no load on it, which on a control circuit is close to the source voltage no matter how bad the connection between them is.

That is the same effect people call phantom or ghost voltage when it shows up on a disconnected conductor, and it is the same instrument behaviour causing both. The remedy is to make the measurement draw current: a low-impedance meter setting, or a reading taken with the actual load connected and energized. Use a low-impedance setting deliberately and not by habit, because it draws real current through whatever it is across, which is inappropriate on a sensitive electronic input and on any circuit whose protection you do not want to operate.

The call

A packaged unit that intermittently fails to start. On some calls the compressor contactor pulls in normally and the unit runs its whole cycle without a complaint. On others the contactor does nothing, the unit sits, and the next call an hour later works.

The first tech measured 24 V at the control transformer's secondary, read it as nominal, and replaced the contactor. The unit failed the same way four days later.

Hypothesis one, the source. The transformer reads nominal at its own terminals. That does not clear it, because the reading was taken at the source with nothing calling. What it does establish is that the transformer's no-load output is correct, which removes a shorted turn or a wrong tap from consideration.

Hypothesis two, a safety control. The obvious next suspect is a limit or pressure switch opening the control circuit. This is where the temptation to jumper arrives and where it gets refused: a jumper across a safety control removes the protection the machine was built with, and the correct move is to establish whether the device opened by reading what it monitors. Here, the sequence fails with the unit cold and idle, in a state where nothing the safety controls watch is anywhere near a trip point, so the safety controls are eliminated on evidence rather than on a bypass.

Hypothesis three, control power at the pickup instant. The measurement that matters is taken at the coil terminals, with the thermostat calling, at the moment of the attempt. Doing that is energized work: 29 CFR 1910.333(a)(1) permits it only where the employer can demonstrate that de-energizing introduces additional or increased hazards or is infeasible due to equipment design or operational limitations, which a reading that only exists during an energized attempt genuinely does. The low voltage of the control circuit does not lower the hazard of the enclosure, because the line side of the transformer is in the same box, so boundaries and PPE come from the risk assessments at NFPA 70E-2021, 130.5 and 130.7, in the edition your employer's electrical safety program has adopted.

The numbers that named it

From the device data: the coil's sealed burden is 10 VA, its inrush burden is 60 VA, and it requires 85 percent of its rated 24 V to pick up, which is 20.4 V. A corroded terminal in the control run measures 2 ohms.

At sealed burden, coil current is 10 VA divided by 24 V, which is 0.42 A. Across 2 ohms that is a drop of 0.83 V, so the coil sees 23.2 V while the unit runs. Nothing about a running unit will ever show this fault.

At inrush burden, coil current is 60 VA divided by 24 V, which is 2.5 A. Across the same 2 ohms that is a drop of 5.0 V, so the coil sees 19.0 V at the instant it is asked to close. Pickup needs 20.4 V. It fails, by 1.4 V.

The intermittency falls straight out of that margin. Contact resistance at a corroded terminal is not a constant; it moves with temperature, with vibration and with the oxide layer being disturbed. At 1.2 ohms the drop is 3.0 V, the coil sees 21.0 V, and it picks up normally. At 2 ohms it sees 19.0 V and it does not. The fault is a resistance sitting either side of a threshold, and the machine's behaviour is a direct read-out of which side it is on that hour.

Once it picks up, everything is fine, and that is the trap. The burden collapses from 60 VA to 10 VA, the drop falls from 5.0 V to 0.83 V, and a tech who arrives while it is running measures a healthy circuit at every point. The condition that fails does not exist while anyone is looking at it unless they force the attempt and measure during it.

What would have made this a different fault. If the coil had seen 19.0 V continuously rather than only at pickup, the machine would have failed every time rather than intermittently, and the first tech's contactor swap would have been diagnosed as ineffective on day one. And if the shortfall were on the drop-out side rather than the pickup side, the unit would start and then release mid-cycle, which points at the source or at a competing load on the same transformer rather than at a series connection, because dropout voltage sits well below pickup and a series resistance that only acts at inrush cannot reach it.

The other thing control power does that load power does not

A ground fault on a control circuit does not simply open something. It can close something.

  transformer coil switch in the
   secondary                        GROUNDED leg
      +----------------[  ]-------------/ --------+
      |                        |                  |
      |                    a fault to             |
      |                    ground here            |
      |                    completes the          |
      |                    circuit around         |
      |                    the open switch        |
      +--- grounded conductor --------------------+
           and pulls the coil in with nothing
           commanding it

Where one side of the control circuit is intentionally grounded, the dangerous arrangement is a switching contact sitting in the GROUNDED conductor rather than the ungrounded one. A fault to ground on the conductor between the coil and that contact ties it straight to the grounded side of the source, completing the circuit around the open contact and energizing the coil regardless of what the contact is doing. The symptom is a device that operates when nothing commanded it, which is a failure mode load power does not have, and which is the reason control circuit grounding is prescribed rather than optional.

Note the contrast, because it tells you which half of the circuit to look in. A ground fault on the UNGROUNDED side, between an open contact and the coil, does the opposite: it puts the grounded potential on the coil's line terminal, so the coil cannot pick up at all, and if the contact then closes the fault shorts the coil out and clears the control fuse. Same fault, opposite half of the circuit, opposite symptom. Those requirements live in NEC Article 250 for grounding generally and in Article 430 Part VI for motor control circuits, both in the edition your authority having jurisdiction has adopted.

An ungrounded control circuit inverts the problem. The first ground fault does nothing observable at all, so it sits there undetected until a second fault on the other conductor completes a path and something operates or a fuse clears. An ungrounded control circuit is not a circuit without ground faults; it is a circuit that hides the first one.

Shared sources make one branch's fault another branch's symptom

Where several control branches share a transformer, the transformer's regulation couples them. A branch drawing heavy inrush pulls the secondary down for every other branch at the same instant, so a relay elsewhere in the panel can drop out or fail to pick up at the exact moment an unrelated device energizes. That correlation is the diagnostic: if the misbehaving device always misbehaves at the same step of the sequence rather than at random, look at what else is energized at that step rather than at the device that complained.

How to verify you got this right

The finding is only proved by measuring under the failing condition, so the verification is a reproduction rather than a re-reading. Force the attempt and capture the voltage at the coil terminals during it, then compare that number against the coil's pickup voltage from the data sheet. A reading above pickup during a failed attempt means the coil or its mechanics are at fault after all; a reading below it means the circuit is, and the difference between the two settles which part gets replaced.

Then close the loop on the connection itself. After remaking the suspect termination, repeat the same measurement under the same forced attempt and confirm the drop has fallen, rather than confirming that the unit started once. A single successful start proves the resistance was on the good side of the threshold at that moment, which is what it was doing on half the calls before anybody touched it.

References

• 29 CFR 1910.333(a)(1) energized-work gate and (b)(2) electrical lockout; 29 CFR 1910.147(a)(1)(ii)(C) excludes electric utilization installations; 29 CFR 1926.417 construction counterpart
• NFPA 70E-2021, 130.5, 130.7 and 120.5, in the edition adopted by your employer's electrical safety program
• NEC Article 250 and Article 430 Part VI for control circuit grounding and protection, in the edition adopted by your authority having jurisdiction
• Device data sheets for coil pickup voltage, sealed VA and inrush VA
• See related: What a Control Transformer Is Sized For; How Control Voltage Differs From Line Voltage