Separating mechanical noise from magnetic noise in solenoids
Change the coil voltage by 20% and listen again. If the noise changes, it is magnetic and it is coming from the force itself. If it sounds the same, it is mechanical and no electrical fix will help. On AC units check the shading ring first; on DC units check PWM frequency and supply ripple; on both, check whether the plunger is hitting a hard stop.
Why this happens
Two completely different mechanisms produce a sound that a listener describes the same way, and the repairs share nothing. Separating them is the first task, and it is easier than most people expect.
Magnetic noise comes from the force itself varying with time. On an AC solenoid this is unavoidable in principle: single-phase current passes through zero twice per cycle, and force, which depends on the square of flux, collapses at each zero crossing. On a 50 Hz supply that is a pulsation at 100 Hz, which is squarely in the range where a housing will radiate it as a hum. The shading ring is the standard countermeasure, producing a phase-shifted flux that keeps attractive force from reaching zero. A damaged or wrongly seated ring removes that protection and the hum comes back.
On DC, force is steady, so magnetic noise has to be introduced by the drive. A PWM hold creates ripple at the switching frequency. An undersized reservoir capacitor on the supply creates ripple at twice mains frequency. Both modulate the force and both become audible if the frequency or a harmonic lands where the structure radiates efficiently.
Mechanical noise comes from contact and from movement: the plunger striking a hard stop, a housing that is not fully clamped, a return spring that rattles at some point in its travel, a core that is loose in its bore. These are indifferent to how much force the coil produces, as long as it is enough to operate, which is exactly what makes the voltage test so effective.
Check these in order
1. Voltage sweep, listening for change. Raise the coil voltage 20%, then lower it 20%. Magnetic hum responds to force, so its level and character change with voltage. Mechanical rattle does not, beyond whether the unit operates at all. This test costs a minute and eliminates half the possibilities.
2. Power off, shake and push by hand. With no current, displace the plunger through its travel and tap the unit gently. Any noise you can reproduce by hand is mechanical by definition. If you cannot reproduce it, the noise needs the unit operating to appear, which points at the force or the drive.
3. AC: inspect the shading ring. Look for a crack, a gap, corrosion under the ring, or a ring that has been installed with a burr lifting one edge. This is the single highest-yield check on an AC solenoid that hums, and it is visual.
4. DC: change the PWM frequency and see whether the pitch follows. If the tone moves with the switching frequency, the drive is the source. If it does not, look at supply ripple or at something that only appears when the unit is fully energised.
5. Check whether the plunger is hitting a hard stop. A solenoid that slams into its end position radiates a broadband click that can excite a resonance in the housing and sound like a continuous buzz after it settles. Adding a compliant stop or reducing the seating velocity often changes the sound more than any electrical adjustment.
6. Look for a resonance rather than a source. If the noise appears at one specific operating condition and not others, you are probably exciting a structural mode. Tap the assembly with power off and listen — a ringing tone tells you the structure has a mode you can excite.
What actually to change
| Finding | What to change | Why not the other thing |
|---|---|---|
| AC hum, ring damaged | Replace or reseat the shading ring | Rewinding does not address a missing phase-shifted flux |
| AC hum, ring fine, gap large | Reduce residual gap, check pole face contact | Force pulsation is worse when the circuit is poorly closed |
| DC buzz follows PWM frequency | Move the switching frequency out of the audible and resonant range | A larger coil still buzzes at the same frequency |
| Buzz follows supply ripple | Increase reservoir capacitance, add regulation | Coil changes do nothing to supply ripple |
| Rattle from a loose part | Clamp, preload, or change the fit | No magnetic change silences a rattle |
| Ringing at one condition only | Change stiffness or mass to move the mode | Chasing the source alone will not fix a resonance |
When it IS the harder problem
A resonance that is excited by a harmonic, not by the fundamental. If your PWM runs at 20 kHz and the unit buzzes audibly, the switching frequency is not the problem — something in the structure is being excited by a harmonic or by the ripple envelope. Blanket changes to coil or drive do not help. The approach that works is to identify the mode by tapping the assembly with power off, then move its frequency: stiffen the mounting, change the mass, or add damping. Moving the drive frequency also works, but only if you know which harmonic is doing the exciting.
Noise that appears only when the gap closes. As the plunger approaches the core, the air gap shrinks and the magnetic circuit becomes much more efficient. Force rises steeply, and any small variation in the drive or in the fit is amplified into a larger force variation. This is why a unit can be quiet at mid-stroke and loud when seated. It also explains buzz that appears only after the unit has been running for a while: the clearances have closed, the gap is smaller, and the forces are higher.
A noise complaint that is a force complaint wearing a disguise. Customers report buzzing, and what they mean is that the force is marginal and the unit is chattering at the point where the force curve and the load curve cross. The audible symptom sends people to the drive and the mechanism when the actual deficiency is force margin at end of stroke. When the noise test tells you it is magnetic, and the voltage sweep changes it a lot, measure the force at end of stroke before you spend time on the drive.
A note on what this page is
This is a personal notebook, not a product page. I write down the checks that actually decide the outcome and the order to run them in, including the ones I got wrong first.
solenoid noisesolenoid buzzingsolenoid hummingsolenoid shading ringsolenoid rattle
Frequently asked
- What is the fastest way to tell magnetic noise from mechanical noise?
- Change the voltage and listen. Raise it 20% and then lower it 20%. Magnetic noise comes from the force, so it responds to the force — the character changes with voltage. Mechanical noise from contact, rattle, or a loose part does not care how strong the magnet is beyond whether it holds. It takes under a minute and it splits the problem in half.
- The unit is AC and it hums loudly. Where do I look first?
- The shading ring. A single-phase AC solenoid produces zero force at every current zero crossing, which is 100 Hz on a 50 Hz supply, and the shading ring exists to keep the force from collapsing there. A cracked, missing, or incorrectly seated ring gives you exactly the buzz you are hearing, and it is a visual check.
- It is DC, so where does buzzing come from?
- Almost always the drive rather than the coil. A PWM hold introduces ripple at the switching frequency, and if that frequency is audible you hear it as a tone. Supply ripple from an undersized reservoir capacitor does the same thing at a different frequency. Change the PWM frequency and see whether the pitch moves with it — that identifies the source in one test.
- The noise only appears sometimes. What does that mean?
- Intermittent noise usually means either a resonance that is only excited under certain conditions, or a mechanical fit that changes with temperature. Resonance appears at specific drive frequencies or duty cycles and vanishes at others. A temperature-dependent rattle points at clearance that opens or closes as parts warm, which is a different fix entirely.