DēLonghi HVS3032 EX:3


After removing four T20 “security” screws (two on the back near the top, and two on the base) as well as the switch knob, then the front panel can be lifted off (you can just let it pull the switch knob off as it’s removed). Since the controls are mounted to an internal frame (rather than the front), you don’t have to worry about pulling the wires in the process (this of course also serves to keep the internal wires safely away from the steel trim, as they don’t need to have much slack with this design).

The front panel by itself weighs 554g, so obviously the majority of mass is in the back. The trim (including paint) measures 0.77mm thick and the mesh 0.6mm. The front fan shroud (+ grille support) is polycarbonate, and control panel is ABS (which is fine in black, given that the panel doesn’t get hot in operation); the knobs don’t say and I can’t really tell from their physical characteristics either (less shiny than the ABS panel, too-easily shaved to be polycarbonate, harder than the PA66 knobs on a DL2401TF), but whatever plastic they are seems to be of good quality. (My younger sibling – who handles most things roughly – has managed to wear the older unit’s knob rim against the panel; but it’s not broken yet, which is a good sign.)


Here we have generous sizes as well:

Item Colour Style Rated VAC Rated T (°C) Conductor size Loading
Main neutral Light Blue UL AWM 3135 600 200 16AWG (1.31mm2) 10A
Front heating element Red UL AWM 3135 600 200 16AWG (1.31mm2) 3.75A
Rear heating element White UL AWM 3135 600 200 16AWG (1.31mm2) 6.25A
Fan motor Black (fast) UL AWM 1430 (earlier)
UL AWM 1330 (EX:3)
300→600 105→200 22AWG (0.33mm2) ≈0.1A
White (slow)
Blue (common)
Neon indicator White UL AWM 3135 600 200 20AWG (0.52mm2) ≈2mA
Light Blue

One piece in the neutral wiring is even 15AWG (1.65mm2). There is no table entry for a main active wire, as the cord active wire is connected directly to the thermal fuse; this would arguably make a 90°C (H05V2V2‑F) cord worth opting for, but being at the outermost reach of the terminal board it doesn’t bear the brunt of the heat, and the thermally‑similar Kambrook models have a 75°C thermal cut‑out so it’s probably good enough.
The fan motor leads are tucked into a secure channel (on the back of the heater frame) in the earlier version, and in EX:3 are stitched through slots in the (now mica) terminal board:

All of the high-current blade terminals (most latching, except on the largest cores) feel secure. The fan wires have terminals with less contact force, but this is because they’re made of thinner metal, not by lack of quality control (and in any case, those will be pretty-much fine as long as they aren’t flapping in the breeze 😀).
As usual, DēLonghi avoid using any of the free-placed “cup” crimp connectors (distrusted by some people).

Double Insulation

As far as I can see, all live parts and internal wires maintain a safe enough distance (>6mm for live parts, >3mm for single insulated; and in most places, the actual distance is considerably greater again) from the metal trim. Nit-picking, I could suggest that the tip-over switch wires (see below) have supplementary insulation sleeves, but they’re not flailing around and they don’t lean towards the trim with their distance from the switch; so on the whole, I’d rate it safer than even properly-made small SMPS (never mind dodgy models), and indeed safer than Class I on anything less than a TN‑S earthing system (which is uncommon in most countries, Australia included, due to its cost). (Still, DēLonghi’s own HSX3324FTS convector/fan heater has mostly double-insulated internal wiring…)


The rotary power switch is a Merchant ST-902-2 with the E35 sequence, rated 16(3)A per contact for 10,000 cycles or 6(1)A per contact (very close to the 6.25A drawn by the 1500W element) for 50,000 cycles. Since the two fan speeds and 900+1500W elements use up all four switch channels (with none being directly powered on all settings), the neon indicator is unswitched (only affected by the tip-over switch).

The thermostat is a “YWS Forever Great” WK-04, rated for 250V 16A. It is driven from the outer concentric knob by a gear arrangement, with grooves indicating the correct alignment at assembly.

The tip-over switch is a Merchant SM-51D with a stainless-steel lever (28mm long, style 1 according to the datasheet), rated 16(4)A for 10,000 cycles or 10(3)A for 50,000 cycles (as is the case in this application).

Heater assembly

Here are the heating element parameters:

Identified Measured Calculated (power at 240V)
Alloy Wire diameter Cold resistance Cold power (estimated hot power) Length Surface load
Front FeCrAl 0.46mm (0.166mm2) 63.5Ω 907W 889W 7.59m 8.1W/cm2
Rear 0.6mm (0.283mm2) 38.1Ω 1512W 1482W 7.75m 10.14W/cm2
(combined) 23.8Ω 2419W 2371W

Hot power estimation is based on a 2% resistance increase, as specified for Kanthal AF or Kanthal D wires at 400°C.
Adding about 20W for the fan motor, overall power of the heater is pretty much bang-on to the nominal.

I’ve calculated the lengths for Kanthal AF wire; if Kanthal D, then it will be about 3% longer.
End terminals appear to be the usual nickel-plated steel (with its usual weakness to moist dust).

The different surface loads are why fan speed is only lowered on 900W mode (and then only to a modest extent), as the 1500W element is calculated to work with the full airspeed (so would glow if used with the lower speed). (This could of course be overcome by using a thicker wire, 0.65mm or so, to get its surface load down to about 8W/cm2 again; but that would require more turns and a finer pitch on the former to accommodate the length, and in any case there’s a compromise between quietness and heat distribution, so economy isn’t the only justification for full speed at 1500W.)

The thermal cut-out is a Klixon TH11CA091, which trips at 90±5°C and includes a self-hold feature (via internal heating resistor) so won’t reset until after switching the power off. The thermal fuse is an AUPO BF121X (TF=121°C, can handle 250V 16A loads), so pretty much everything in the full-current path is rated for 16A even though only 10A is actually drawn. Nice 😉 The rivet connecting it to the input does go through the frame for mounting, but the hole has been reinforced with a metal grommet beforehand; elsewhere DēLonghi have applied the same caution as Kambrook (dedicating each rivet to connection or mounting, not both). While it may seem odd that the thermal fuse’s casing is connected to the active (and hence live even if it blows), this is done to avoid conducting the Klixon’s self-heat to its body (as would occur if it was installed the other way around).

The EX:3 revision changes the terminal board from plastic to thick laminated mica (as you could already see above), for additional heat resistance (and fireproofing at the terminals and rivets). Some mica heat shields have also been strategically added between the heating elements and housing (and on the underside of the heater frame), to protect it from the temperature surge if the fan fails (even though PA66 already has a higher melting temperature than the polypropylene used on cheaper fan heaters).


Another 2-blade impeller (about 160mm diameter) like the Kambrook’s, but this time made of 0.7mm aluminium for faster spin-up and no rusting; driven by a two-speed shaded-pole ABJ Motor, with visible felt rings holding extra oil. Its winding is rated for 155°C (class F), significantly better than the usual 120°C (class E) on cheaper motors.
Unlike in many other heaters, the casing design keeps the back bearing well-shielded from dust.

The impeller has the usual black (powder?) coating on the front, but white on the back; I can only guess this is to avoid absorbing the heat radiated from the elements upon fan failure (it wouldn’t make any discernible difference in normal operation), or possibly to make it more obvious which side is which (although it won’t fit on back-to-front anyway).

I would prefer a 3-blade impeller for harmonic stability (the fan heaters I’ve heard with 3 blades do sound noticeably smoother than 2-blade models), but these fans still aren’t actually so loud for their combination of speed and diameter.

Neon indicator

It was a bit hard to get to (and I was slightly fearful of breaking the bezel tabs, since there was significant friction even after unlatching them), but the neon bulb is 4.8mm diameter by 11.3mm long with a 100kΩ (98.5kΩ measured) series resistor; that’s passing more current than I’d really like through that size of neon bulb (especially being unswitched), but a good‑quality bulb might still survive a reasonable time. (So far the older unit’s bulb is still bright enough, but shimmers noticeably.)
Even so, it seems a bit out of place with DēLonghi’s otherwise-conservative engineering.

Pages: 1 2 3 4