I just got an air-purifier from Korea but missed out on the Hz. The specs stated on the back is 220v 60Hz at 15w  but here in Singapore it's 230v 50Hz. So I did some searching online and found out one way is to reduce voltage to compensate for the lower Hz. I measured a 15w bulb in series with another bulb and got a 30w halogen to allow 200v output. But now my concern is the switching power supply if there is one in there would reducing the voltage damage the power supply? Thanks.

Lowering the voltage may be a problm anyway. If there is a motor the wrong Hz lower its torque and increase losses, if the drop the voltage it will lower its torque more - it mignt even fail to start. In which case it will just hang in there buzzing

So when checking, see if the motor is reliably starting and that it is still starting with enough force if you mechanicaly resist to it and then release it

Lamp does not look to me like proper way to do this. Better try some fluorescent choke

But to really drop the V how about using a 230 --> 24v/48v etc transformer and subtracting its secondary from the mains

The SMPS does not care about the Hz and just use it as is. But i would not connect it from the "voltage dropping" arrangrements but only straight from the line - any coupling between the motor and SMPS is harmfull for both the motor (harmonics) and SMPS (back emf)

[Double post]

Reducing the voltage to 200V from the 220 nominal should not do any harm to the SMPS.

But for the compensation of the frequency it depend, what is to be compensated, so what is inside.

I'm pretty sure the series incandescent would do very terrible compensation job - mainly when there is higher current crest factor (e.g. an input of the rectifier in a SMPS). With that the SMPS could see even way lower voltage (as the peak value is, what matter there), even as low as ~160V equivalent and that could be really too low. But mainly the output impedance of such thing would be so high, the SMPS could become unstable (it act as a negative dynamic impedance - higher voltage mean lower current on the input), so it would not be able to supply the required power...

So I'm afraid the frequency should be compensated separately per each feature (separately the fan motor, none for the SMPS,...)



@Ash: With the motor, you have to keep the V*Hz product constant in order to maintain the torque, so the voltage reduction is the correct thing to do there, if the frequency is going to be lower. The only thing is the phase shift capacitor (if used), it's value have to be increased in order to maintain the startup torque. But as the fan itself poses nearly no load when not turning, it is usually not needed either.
But for the most frequently used (for low power fans) "shielded pole" motors the V*Hz product is the only thing, what matter (well, neglecting the winding resistance, but that would be important, when you would reduce the frequency way lower)

15w is pretty low for an ac motor. I suspect it is dc driven from the SMPS. There have been some users who posted using this same product in Singapore they just plugged it in and say it is very quiet. So it may be a dc fan.

15w may well be a shaded pole motor running from the AC. Open the deice and ceck whats in there

[Double post]

If you accidently posted or did a double posting, you can delete you extra post by the "Delete" option.
Edit: Wait: Where is the option for commets deleting? It was there in the past?

I think I will run it at 215v since it is rated at 220v it should work right? Our supply is 238v measured. The incandescent lamp will be shorted at the output by a timer so there would not be inrush voltage/current or high crest factor during switch-on by the time the timer opens the short the lamp is already run at full power.

I think I will run it at 215v since it is rated at 220v it should work right? Our supply is 238v measured.

If it is rated at 220V/60Hz and the most sensitive part is an inductive style load (fan motor,...), the correct voltage at 50Hz for the same magnetic flux would be actually 180V (= 50/60*220)

The incandescent lamp will be shorted at the output by a timer so there would not be inrush voltage/current or high crest factor during switch-on by the time the timer opens the short the lamp is already run at full power.

The high current crest factor come from the rectifier and if it is there, it is there all the time, not only during the startup. But the question is, what part of the total power input came from the rectifier (so with high crest factor) and what part from another load (fan motor,...)

Shorting the bulb out by a timer mean the whole thing get full 240V on power ON, but it is unlikely for that to be a problem.

15w is pretty low for an ac motor. I suspect it is dc driven from the SMPS.

Small fans are as low as few W, on the contrary the 15W would be way too high for a DC fan, these run at max 2..3W and have quite low efficiency.

There have been some users who posted using this same product in Singapore they just plugged it in and say it is very quiet. So it may be a dc fan

If there is an experience with this product running well directly on the 240V/50Hz, then why to bother with any conversions.
And, if it is internally all powered by DC from a SMPS, then it would be the best thing to do...

For the fan to be quiet you need low RPM and well designed aerodynamics of the impeller and the complete ductwork.
The DC fans tend to be very noisy for the output power they provide, their only advantage is the cost at low power levels (3W and mainly below): No need for super-thin wires, they are SELV, so no need for any safety insulation,...
For 5..50W the shielded pole induction "classic" design become the design of choice, for power above 20..30W the phase shift capacitor become the concept of choice, mainly for the losses in the "shielding" loop becoming too large.

Powering the SMPS and motor together through input impedance in series with the line is bad thing since they will couple to each other - The motor make surges, the SMPS make harmonics, both will be unhapy with each other

220v rated SMPS can be connected directly to 240V no problem - just detach it and plug directly. Well it might need bigger electrolitic in th einput for 50 Hz but i am pretty sure the existing one is made with more than sufficient margin so dont bother

The motor i'd see how hot it gets / how reliably it starts. If there is need to drop the voltage, i'd use a fluorescent choke instead of incandescent lamp - that's morereliable, less breakable

Runing SMPS in series with a choke (but usually of way lower value than a lamp ballast) is actually common as PFC, that would improve the crest factor.... But again do not run the SMPS and the motor on the same choke

It's a nice purifier by Samsung so I don't think I want to open it up since it's factory assembled (may  not be able to put it back). I think what I'll do is run the thing through the bulb/ballasting device giving out 220v on our 238v hopefully the motor won't heat up as much.


@Medved- I meant shorting the output from the bulb to N so the bulb runs at 100% for a bit then the timer releases the short and the bulb 'gives out' x voltage.

@Medved- I meant shorting the output from the bulb to N so the bulb runs at 100% for a bit then the timer releases the short and the bulb 'gives out' x voltage.

I see what you mean. But that is exactly, what may end up in startup problems with SMPS (except special ones, designed for operation on high impedance source)...
But don§t forget the bulb does there quite terrible job: Seen by the SMPS, the voltage is reduced to way below 200V equivalent, but seen by the fan motor it could be still 220..230V (so no significant change).
The reason is, than the bulb resistance limit mainly the peak value (the voltage get clamped by the SMPS rectifier input), so exactly the opposite from what you would really need.
So I'm really still convinced operating it directly would be of less problems for the SMPS, while for the fan (what is, what need the voltage to be reduced) the voltage virtually haven't changed at all.

Don't forget the lower power motors are way more tolerant to voltage fluctuations (they are designed with larger ohmic drop in the winding), so most fans 30W and below accept both 50 and 60Hz (the 60Hz yield to larger rotor slip, so moving losses there, the 50Hz operation yield lower slip, but higher winding losses, what on low wattage motors about compensate each other)

The supply I measured just now is 240v so wouldn't it be overdriving the 220v SMPS and motor if plugged in directly.

The supply I measured just now is 240v so wouldn't it be overdriving the 220v SMPS and motor if plugged in directly.

The problem with the SMPS is not overdriving, but more a risk of underdriving. So when the impedance of the "mains" feeding the SMPS exceed some limit, a system dead-lock may occur:

The SMPS maintain the output voltage constant, regardless of the actual input (assume the input is within it's operational range).
It does that with high efficiency, so the input power is about the same as the output power.
Because the output voltage is constant and the output load is all the time the same, regardless of the input voltage (it may depend on the SMPS output, but that is held constant, so the output power have no reason to change)
As the output power is constant, the SMPS input power would be constant as well - the regulation loop adopt the input current so, the transferred power match the output load. So when it detect the output voltage going down, it increase the input current, assuming than it would lead to increased input power - with low impedance voltage source, so direct mains, it does that, restoring the balance.

But when the supply before the SMPS have high output impedance (e.g. series bulb,...), the SMPS input voltage would drop as soon as the SMPS tries to draw higher current (to recover the power balance). But as the input voltage drop, the actual power does not increase, but decrease. That mean the power imbalance is made even worse, yielding the SMPS drawing even more current, till it reaches it's limit (usually the duty ratio of the switching activity). But at the same time the input voltage would be so low, the power available at the input won't suffice for the load, so the system get stuck in such state. And exactly in that state the system may endup before it even starts up: To charge the output capacitors, the SMPS temporarily transfer way higher power than with normal operation. And if the input is not able to provide that, the system get stuck.

Now it is not impossible to design a SMPS in such way, it does not get stuck at these conditions. But it mean complications and usually higher cost (larger tank capacitors, delayed startup,...) and degraded performance (output undervoltages,...), so it is not implemented when the system won't need it (e.g. when a series high inductance choke is supposed to be used as HV overvoltage protection,...).
And as the unit is designed to be supplied directly from the mains, it's SMPS for sure won't be designed to overcome those high mains impedance dead-locks.

Of course, when the power handled by the SMPS is very small (1..2 watts only for the microcontroller,...) and main power goes directly to the load (fan, UV lamp,...), the mains could be still considered as "low impedance" even when there is the bulb in series.

But as your observations indicated the SMPS is used to transfer the majority of the 15W, the series bulb may cause the impedance to be too high and so mean the risk of dead-locks.

So if it is OK or not to use such voltage reduction, you may say only after analyzing how the power is handled within the unit.
Don't forget, than with your experiment the system may be "just on the edge", so it could collapse at any time the conditions worsen a bit more (the bulb increase it's resistance due to wear, mains voltage drop,...)...
And could happen, than the operation on the "voltage collapse" regime could mean the primary side of the SMPS get (mainly thermally) overloaded and so could fail after some time when it collapse (I doubt you would recognize that happening in the real life soon enough to turn it OFF before it get damaged).

Really the only safe thing is:
- Either use the good experience of others (run it directly on the 240V)
- (Technically the best) analyze the inner guts and base all adoptions on that (but that require the unit to be opened for an inspection)
Blindly reducing the voltage, mainly by something like high series impedance, is really asking for troubles.

Deleted