Yes, condensation on the capacitors is real, and it is the main reason a phase-converter installation sheet sets both the control panel and the idler between 10 and 40 C with relative humidity under 85%. A water film forms on a run-capacitor terminal whenever that terminal is colder than the dew point of the air around it — about 8 C at 10 C and 85% RH — and 230 to 480 V across a wet, dusty terminal tracks and flashes over. The number that decides everything is the gap between the coldest surface and that dew point; 5 C of margin is the working rule. A 100 W PTC heater at a 15 C setpoint holds a 20 x 16 x 8 in stainless box on the safe side of that line through a -20 C night.
Why does a phase converter manual call for 10 to 40 C and under 85% RH?
That band protects two separate things at once: capacitor and winding insulation that must stay above the dew point, and idler bearings whose grease stiffens as it cools, with 40 C the usual ceiling for industrial control gear. A typical converter installation sheet sets the same band for both the control panel and the idler because they fail for different reasons. The panel carries run capacitors, contactors and the control board; its insulation is what water attacks. The idler is a motor, and its weak point in the cold is lubrication, not voltage. The 40 C top end matches the ambient limit most drive and control datasheets carry. If you are laying out the whole conversion at once, rotary phase converter enclosure size and layout covers the spacing side.
Can condensation really arc across a capacitor terminal?
Yes, whenever that terminal is colder than the dew point of the air touching it, a water film forms, and 230 to 480 V on a run capacitor with a little dust turns the film into a leakage path that tracks and flashes over. The classic sequence is the morning after a cold night. The box and everything in it cold-soaks to -5 C overnight, then the building warms to 18 C at 70% RH by mid-morning, which puts the dew point at 12 C. The capacitor ends lag hours behind the air and sit near 5 C, about 7 K below the dew point, so water condenses on them the way it does on a cold can. The film mixes with dust into a leakage path, the leakage carbonizes into a track, and the track ends in flashover. The condensation guide walks the same physics for any sealed box. One safety note: run capacitors hold a charge after shutdown, so discharge them and lock out before you touch a terminal.
How do I check the dew-point margin in my panel?
Measure air temperature and humidity beside the box, work out the dew point, then read the coldest surface with an IR gun and keep it at least 5 C above that dew point; the bottom wall and gland plate are usually the coldest. Work out the dew point from those two readings with the Magnus approximation, which stays within about half a degree across this range. The 5 C margin covers sensor error and the 2 to 3 K by which an outside wall runs colder than the air inside it.
| Air temperature in the box | Dew point at 85% RH | Coldest surface must stay above |
|---|---|---|
| 5 C | 3 C | 8 C |
| 10 C | 8 C | 13 C |
| 15 C | 13 C | 18 C |
| 20 C | 17 C | 22 C |
Read the second row against the manual's floor. At 10 C air and the full 85% RH, the dew point is 8 C, and a bottom wall 2 to 3 K colder than the air is already sitting on it. So in a damp building you hold the box at 13 to 15 C instead of the bare 10 C floor, and the margin comes back. When an outdoor enclosure needs an anti-condensation heater works the threshold logic in more depth.
Is the real winter problem the idler's bearing grease?
Sometimes, and the two leave different evidence: grease thickening shows up as slow acceleration and high starting current with clean terminals, while moisture leaves tracking marks, corrosion or a film on the capacitor ends. About 80% of industrial bearings run on grease, and grease is 80 to 90% base oil in a soap thickener, per a bearing maker's low-temperature lubrication guide. The oil's viscosity climbs steeply as it cools, so a frozen idler needs more torque to turn and more current to reach speed. Typical general-purpose mineral greases work down to about -20 C, and low-temperature synthetics reach -40 C. A panel held at 15 C never troubles its grease; an idler parked at -25 C in an unheated shed does. That is why the band names both assemblies: heat the panel box, and give the idler the same room or a heated space of its own. Cold-temperature electronics enclosure planning covers the rest of the winter checklist.
How many watts of heater does the enclosure need?
Size it from the box's heat loss at the coldest expected ambient: a 20 x 16 x 8 in stainless box holding 15 C inside at -20 C outside loses 3.7 W/m²K x 0.58 m² x 35 K = 75 W, so fit a 100 W PTC heater. The area behind the number: a 20 x 16 x 8 in 316L wall-mount box measures 0.508 x 0.406 x 0.203 m, and its five wall-exposed faces total 0.58 m². Multiply by this site's stainless figure of 3.7 W/m²K and the 35 K difference: 3.7 x 0.58 x 35 = 75 W. A 100 W PTC heater covers it with a third in reserve, and a PTC element self-limits so it cannot cook the panel.
Two things move the number. A milder site that only sees -10 C needs 3.7 x 0.58 x 25 = 54 W, where a 75 W unit fits. Painted steel walls lose faster, at 5.5 W/m²K, so the same box needs 5.5 x 0.58 x 35 = 112 W and the 125 W size. For the warm-up transient rather than steady state, enclosure heater sizing for a one-minute warm-up has that arithmetic, and heat loss from a heated body in ambient air shows the same method on a different shape.
Where does the heater go inside the enclosure, and at what temperature?
Bottom third of the box, at least 2 in [50 mm] clear of plastic duct and capacitor ends, thermostat at 10 to 15 C, with the humidistat option holding relative humidity under about 60%. Mounted low, the heater's convection sweeps the walls, and the bottom wall is the surface you are really protecting. Put the sensing element on the wall farthest from the element and out of its radiant path, with the clearances the heater datasheet prints for wiring and duct. Set the thermostat to 10 C in a dry building and 15 C in a damp one, which matches the margin rule above. The humidistat variant runs only when the air approaches saturation, so it wastes little in spring and autumn. Once the idler is spinning, its own losses keep the room warm; the heater's shift is the 12 stopped overnight hours.
Does a heater change the enclosure's NEMA 4X rating?
Only through the hole you make for it: a heater built for sealed enclosures, with its thermostat wired through a gland of the same rating, keeps the box at its own rating, and anything weaker — an IP54 cable gland, an unsealed knockout — becomes the assembly's rating. The enclosures we build are rated NEMA 4X / IP66 by the manufacturer and not UL listed; NEMA writes the type definitions and tests nothing itself. In practice the rating is lost at field-cut holes, so bring the thermostat cable through a gland rated IP66 in a hole from the factory drawing. Manage the water with heat and leave weep holes to Type 3R gear: NEC 314.15 permits drainage openings of 3 mm [1/8 in] to 6 mm [1/4 in] in boxes and conduit bodies listed for damp or wet locations, and an open drain hole on the wall ends the 4X claim there.
What happens at the 40 C end of the band in summer?
The same sealed box needs its summer check: at 30 C ambient, the example panel's 20 W of control gear raises the 20 x 16 x 8 in box 20 W ÷ 2.1 W/K = 9.5 K, to just under the 40 C ceiling. The 2.1 W/K is the same stainless wall figure, 3.7 x 0.58, now carrying heat out instead of holding it in. That gives this box a ceiling of about 21 W of internal gear at 30 C ambient (10 K of allowed rise x 2.1 W/K), and the 20 W example fits under it. Past 21 W, order one size up or move heat out with an enclosure heat exchanger, and count the idler's losses when it shares the box. How to check control cabinet temperature limits and finding the real heat gain of an outdoor enclosure carry the summer side.
What should the order specify?
A 316L body in damp or salt-air buildings, the size from the panel layout plus heater space, manufacturer-rated NEMA 4X / IP66, and the heater with its control: for the example, a 100 W PTC unit at a 15 C thermostat.
| Order line | The example |
|---|---|
| Body | 316L stainless, 20 x 16 x 8 in outside, 19.88 x 15.88 x 7.88 in inside |
| Rating | manufacturer-rated NEMA 4X / IP66 |
| Heater | 100 W PTC with 15 C thermostat, mounted low |
| Entries | bottom-wall glands, each rated IP66 |
| Cutouts | gland holes and heater mounting on the factory drawing |
Set the body to the building: 316L where washdown, brine or salt air reach it; both stainless grades come made to your dimensions, and the 316L wall-mount range lists this size and larger bodies. The enclosure selector sizes a body by inside height and DIN count before you commit. You set the material, width, height, depth and every cutout on one drawing; delivery is 10 business days anywhere in the world, counted from your order, or from your approval of the drawing for custom work.
The decision in three lines: dry building, thermostat at 10 C; damp building, 15 C and a humidistat; idler outdoors in a cold region, move it into the heated space or grease it for -40 C.
Enclosures in this article
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 20 × 16 × 8 in | 316L stainless | 20 x 16 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Does the converter's own heat keep the box above the dew point while it runs?
Usually, yes. A 5 hp idler on light load still sheds a few hundred watts, and 75 W held the example box 35 K above a -20 C night. The risk is the stopped hours, when the box cold-soaks to building temperature and the first warm, humid morning puts the terminals below the dew point. The thermostat holds that floor for you.
Is 85% relative humidity the number to watch?
Watch surfaces, not the meter. Relative humidity moves with temperature while the dew point stays put: the same 85% RH air has a dew point near 8 C at 10 C and near 17 C at 20 C. Keep every terminal, bottom wall and gland plate at least 5 C above the dew point, and the humidity number takes care of itself.
Can a 40 W lamp replace a proper enclosure heater?
Not for this size of box. A 40 W lamp against a 75 W loss at -20 C leaves the panel below setpoint all night, and an uncontrolled lamp can overshoot toward the 40 C ceiling on a mild day. A 100 W PTC heater with a thermostat covers both ends.
What if the building stays unheated over a weekend?
Leave the heater on its thermostat. Holding the example box at 15 C through two days at -20 C costs about 75 W x 48 h = 3.6 kWh, and the box walks into Monday's warm, humid morning already above the dew point, which is exactly when an unheated box condenses on its capacitor terminals.
How cold before bearing grease, not condensation, is the limit?
Typical general-purpose mineral grease works to about -20 C and low-temperature synthetic to -40 C, per grease makers' selection tables. Grease is 80 to 90% base oil, and its viscosity climbs steeply as it cools, so a hard start below those temperatures points at the bearings. Swapping to synthetic grease, or moving the idler into the heated room, fixes it.
