Mega-minireview: Vornado 633 "whole room air circulator"

[LongRunner]

On the 633, I count 32 flow-focussing vanes; the intake has 64 slats in the outer section and 36 in the inner part, and all three numbers are multiples of 4. (Removing the grille also gets rid of the fourth harmonic.) The 533 (which doesn't produce a noticeable fourth harmonic) has 25 flow-focussing vanes, 51 slats in the outer section of the intake, and 28 in the inner part; none of which share a common factor (other than 1).

If anyone reading this happens to have other Vornado models, it would be nice to respond with the number of vanes/slats (or photos in which all are visible) in them, and the audibility of any harmonic of the rotation (beside that expected from the blade count).

The DC models have made their way to Australia in the meantime, so I'll notify you if I get one.

[LongRunner]

I was tossing up among which of the DC models to get; ultimately I got a 6303DC, the latest generation (also includes the smaller 5303DC and pedestal 6803DC).

Operation
Tonality is even better-suppressed than before; both the rear and front grilles now have 38 vanes, so no fourth harmonic instability.
Even at maximum speed (99), it now sounds quite smooth.

Minimum speed is lower than on the 633, but could be lower still.

Motor
It's 8-pole, 3-phase for smooth running (like HDD spindle motors) with PWM control, made by Delta with dual NMB 608ZZ ball bearings. Two lines of ID:
MO3-0217-07
MH08824PD-S3006
There are 5 wires; one white (unused) and the others as usual for Delta (black ground, red +24V, blue frequency output and yellow speed control).

One annoyance is that although the manual says the blades can be removed for cleaning, here they've put red threadlocker on the nut .
It is for this reason that I couldn't examine the motor driving circuitry.

PSU
While the 6803DC (pedestal) still uses an external power-brick, the 5303DC and 6303DC build it into the unit.
It's a bog-standard flyback SMPS, driven by a TOPSwitch-JX (going by the package and pinout; can't reach the clamp screw to reveal which variant) with the +24V output rectified by an MBR20200CT with insulated tab. No PFC, no MOV, though at least the NTC thermistor is present (but unsleeved).

Could be worse I suppose, but it's quite bad enough (especially being before the fuse).

Soldering mess.jpg (266.23 KiB) Viewed 17382 times

Mess 2.jpg (181.03 KiB) Viewed 17382 times

There's quite a mess of flux where the cord wires are soldered in (and a bit at the junction of the two Y1 capacitors), so I cleaned that up after taking these.

On that note, being somewhat heavier, the cord is H05VVH2-F rather than H03VVH2-F. The air accelerator now doubles as a storage place for slack.

Controller
An LCM08F18GT20 is used, powered through a 78M05. Oddly, the PWM signal is low-pass filtered into a continuous voltage.

Capacitors
No joy; the main output caps are Xunda TM series (replace with Chemi-Con KYB), the rest are "SWC" KM series (probably general-purpose, I suggest replacing with something better for longer endurance). At least the output capacitors have a sufficient ripple current rating for the task (≈1.6A RMS for about 2A output at most, and actually doubled given that they are two in parallel with no coil between them), so it's not outright suicidal.

General build quality
About the same as before (fairly solid but with the occasional over-tightened screw), although at least with no knob to break.

Conclusion
The design in general is nice, but for the price only quality capacitors should be used.
As such, it comes close but ultimately falls a bit short of winning an award; good caps and clean soldering would raise it to Bronze.

UPDATE 2022-05-05: I recapped it using KYB on the PSU output, a KXG primary (which I put neutral-cure silicone under as with the original), Rubycon YXM for the 47/50 in both the PSU and control board (so all PSU caps now have 10k-hour endurance), a Panasonic FR for the 470 on the control board (although uncritical being after the 78M05), and ceramics for the 47/16 (although only operated at +5V) on the display/± button board.
I even replaced the Tenta X2 cap with an Epcos/TDK one for good measure.

Grades

• Airflow: A. (Perhaps an incremental improvement from before, but nothing radical)
• Noise: A−. Almost perfectly broadband (a superb result with uniformly sized and spaced blades), but I'd like minimum speed to match or undercut the sound level at least of a Barracuda ATA IV (it's only been 20 years since then ); as it stands, it's still much louder than that.
  (The motor itself is actually perfectly capable of going much slower; the limit being imposed by the microcontroller program.)
• Appearance: B−. It's glossy, but at least being white mitigates that; and unlike some manufacturers, they actually remembered to peel the protective film off the LED display inside (either that, or it wasn't applied in the first place).
• Build quality: B− again. Not bad overall, but for AU$349 I want to see nothing short of excellent. At least the warranty is decently long (10 years in their native USA, 5 years in Australia although that should rather be the minimum for a premium appliance like this).

[Behemot]

Been asked about ceiling fans for quite large hall with quite high ceiling (pretty much the internal side of a large saddle roof). They are obviously especially in winter heating mostly just the roof part where nobody is, so it would make sense to blow the hot air from the upper part of the hall down towards the floor.

Customer had two offers for single incredibly large (3.6m diameter) and equally expensive ceiling fan, or four smaller ones but with immenselly expensive realisation works because of how high it is (and I wonder how much air would actually just 90cm wide fans so high actually move, since ordinary fan creates very turbulent airflow).

So I remembered these you reviewed, that they provide quite laminar flow, which could be of use if mounted at some 8 meters, that they could still suck enough warmer air and direct it towards the floor, maybe even diagonally towards the farther wall. Do they provide anything wall-mounted, or everything is just floor type?

[LongRunner]

None of the Vornado range is designed for mounting (apart from the Portal and Transom window models) as far as I can see, but since they can be aimed at any angle from straight forward to straight up (when sitting on the floor), they could work on the wall if you improvised a mounting arrangement.
The 6303DC (or a few of them) may just do it if you're prepared to drill holes in the base (so you can glue nuts and washers on the inside to bolt an interposer into), and upgrade the remote control for longer range (hack the battery compartment to accommodate a CR2450 rather than CR2032, and swap the original single 3mm infrared LED and 3.9Ω resistor for two series infrared LEDs and a 1.5Ω resistor*); since the battery compartment has (just) enough room to spare and the infrared window is full‑width, it won't be an ugly bodge (provided you open it up properly in the first place, using a plastic scraper or spudger).
*If that isn't enough, then use a PNP transistor to buffer the drive signal, and put as low of a resistance as you please in the output.
You may also try narrower-angle infrared LEDs, for longer range at the cost of requiring more precise aiming; if all else fails, set up a repeater…
If you're going to void the warranty on a DC model, of course, you might as well recap like I did
The 7503DC would be better if you can get it, but I can't find it on Vornado's current sites (whether in their native USA, Australia, or Germany where you'd presumably import them from) so I guess it's discontinued; anyway if my idea sounds good, obtain as many 6303DC units as the customer can afford, mount them one‑by‑one until they have enough airflow, and then they can keep the remainder for use elsewhere It's one thing to put a 22cm (medium‑size) Vornado against the 40cm Dimplex desk fan (since discarded); a 90cm ceiling fan, even with its crude design, is a more formidable opponent just by way of having 16× the area. So the Vornados might manage, but will need to be on at full speed (fairly loud) to have a chance; still, I suppose it's worth a try…

Well actually bladeless (or more correctly hidden-blade) fans like Dysons provide the most-laminar airflow of all; just not enough of it to help Some turbulence is useful for exchanging heat with objects in front of the fan, and I think Vornado's engineering has arrived at a happy balance between the two extremes (enough to be effective for personal fanning if desired, but not so much as to cause adverse dispersion like the cruder fans do). Or perhaps they just move so much air that it doesn't need to be turbulent to fan nearby items

[Behemot]

something like 293 could do I guess, but they only make it for 120V/60Hz system

of those from european catalog, those EXO models could be most easily mounted on the wall, but they are all tiny, at just 20 cm, would need to use at least like 11+ of them

another option I can think of is some 30cm tube fans, those should also be able to force the air reasonably down, when mounted on the aprox. 45° angled halfs of the roof, facing down - would suck the hottest air from the middle (where they have furnace, assisted with ACs in heating role on the sidewalls, but those could also be directed to blow the air towards the midle of the hall) as per the cutaway image I drawn; could use some like these https://www.ventilatory.cz/kovovy-potru ... -mm-x12741 which are still reasonably priced, could be speed-regulated and they provide simple metalic mountings struts for them too (as they are often used as simple blowers, without actually being in any piping)

[LongRunner]

Yeah, those tube fans look like a decent option for the circumstances.

In the worst case, if they still have a far corner or crevice that's colder than the rest, it could then be filled in with a portable heater (depending on the size of the cold spot, even one of those precious little 500W types may suffice; if you want better quality, DēLonghi also have 360W models with a thermostat)…

[Behemot]

I guess its not like its actually cold, but it takes a lot of energy to get it comfotable near the floor when they are heating like 70 % of the total volume above their heads; I guess I made the walls taller than they actually are relative to the roof, its like 12 meters in the middle where it's tallest, the side walls are barely above 2 meters