Simple VFD for Low Power Single Phase Induction Motors

[zundk]

hi

for an art exhibition we need to control the speeds of a bunch of domestic floor fans with an arduino which will have some interactivities with the environement

each AC motor run at 220v/50hz/55w and seem to be of permanent split capacitor type

they will be driven each by separate "channesl" of the arduino


after desperatelly scrutining the web during the last weeks for a simple and quick solution we have found that there is aparently none apart possibly from VFDs.


could anyone here please help with some simple to build and not too much expensive circuit that could be driven by 5v from an arduino?


we are thinking maybe to buy several "cheap" VFDs (if that exists...) on ebay but we don't know what type could work for us and we need some advice from experts

would maybe anyone here have some other solution ?

we are are in a real hurry to get a definitive solution for this, our time and budget are running out quickly

thankyou very much

zundk & ninonpas

[ignator]

VFDs are ONLY for 3 phase motors.  I'm not aware of any method to speed control an induction motor.  However a triac speed controller may work, as this could duty cycle the AC sine wave form.
Try an off the shelf light dimmer and see if this can control the desired speed range of the fans.
If this works, then some circuit research/lab work could produce a circuit that takes 0-5DC and controls the triac phase delay that controls the duty cycle at the 50Hz power source.
You may have some startup problems of the fans, and initially require 100% duty cycle to get them spinning, then back off to a control speed.
If you have a budget for VFDs, maybe you could find fractional HP (or less then 100 watt) 3 phase motors to replace the existing fan motors. 
Most domestic fans here in north america are a shaded pole design with 3 windings for speed control (2/4/6 pole fast/medium/slow).
ignator

[Kremmen]

Take a look at this appnote from Freescale http://cache.freescale.com/files/microcontrollers/doc/app_note/AN3471.pdf. It documents one solution using their processor, but any other one will do as well. The key thing is to control the motor voltage using a simple pwm scheme. The control element is a simple triac and the control circuit is optically isolated so nobody dies while using it. An Arduino should be capable of controlling several channels with a bit of programming.

[NiHaoMike]

VFDs are ONLY for 3 phase motors.  I'm not aware of any method to speed control an induction motor.  However a triac speed controller may work, as this could duty cycle the AC sine wave form.
Try an off the shelf light dimmer and see if this can control the desired speed range of the fans.
If this works, then some circuit research/lab work could produce a circuit that takes 0-5DC and controls the triac phase delay that controls the duty cycle at the 50Hz power source.
You may have some startup problems of the fans, and initially require 100% duty cycle to get them spinning, then back off to a control speed.
If you have a budget for VFDs, maybe you could find fractional HP (or less then 100 watt) 3 phase motors to replace the existing fan motors. 
Most domestic fans here in north america are a shaded pole design with 3 windings for speed control (2/4/6 pole fast/medium/slow).
ignator

It's possible to drive a "single phase" PSC motor as a 2 phase motor.

[John_L]

VFD with ONLY single phase input can drive capacitor start/run single phase motors. VFD operating from single phase have output voltage on each phase equal to input voltage. Higher capacity single phase motors with separate high torque/high current  starting winding operated by centrifugal switch can not be used with VFD.

[NiHaoMike]

VFD with ONLY single phase input can drive capacitor start/run single phase motors. VFD operating from single phase have output voltage on each phase equal to input voltage. Higher capacity single phase motors with separate high torque/high current  starting winding operated by centrifugal switch can not be used with VFD.

Every split phase motor is really an asymmetrical 2 phase motor where a capacitor and/or inductance difference provides the second phase. You can just disconnect the windings from the existing starting circuit and operate them independently from a 2 phase inverter.

[Kremmen]

While all of the 2 phase stuff is essentially so, it will hardly be useful for the OP. A fan can be successfully controlled by the primitive method of phase angle control outlined in the app note i included. That is due to the nature of the load and the torque curve of a typical induction motor. In this scheme there is a 1 to 1 correspondence with the phase angle and resulting fan speed. While the curve is anything but linear, it probably doesn't matter in this application. An it is nearly trivial in Arduino to compensate using a linearization curve if needed.
The simple triac controller is entirely doable for this kind of project. A 2 phase VFD? Forget it.

[flolic]

Here at work, we have whole bunch of industrial single phase induction motor fans, and simple triac regulators for them. They work just fine. Another option for regulation is 5-steps autotransformer.

[zundk]

thank you all for your responses

ok, so we are back to our original idea to use optocouplers, triacs and a snubber then...

it is very confusing to read such divergent opinions all around the net concerning the control of speed of such motors

well, it's time to test by ourselves now.

if it wasn't about 220v i wouldn't be so cautious
untill now I've allways managed to stay away from the mains
but one day one has to confront the beast...

i hope nothing will explode

[zundk]

flolic, this idea of autotransformer seem an interesting option if it works
i am reading this: http://www.allaboutcircuits.com/vol_2/chpt_9/5.html

[fcb]

The OP could very simply wire up a cheap 'dimmer-switch' to test the degree of control that they might be able to achieve with the phase angle control method.

If they are happy with that range of speed, simply execute multiple channels of phase control in a microcontroller (pretty straightforward).  If the OP finds that phase-angle doesn't do it, they can report back.

[fcb]

One other thing.

If the motor is a 2-phase type with a phase-split capacitor, the capacitors purpose tends to be to create an approximate 90 degree phase shift for the 2nd phase winding.  This would imply that changing the frequency going in would change the phase shift also, resulting in some loss of efficiency - perhaps small variations wouldn't be significant - but running the thing on 10Hz might not work.

[SeanB]

If you are worried you can get a complete optoisolated power controller that is basically plug mains one side and 0-5V the other for 0-100% control. Works for resistive and inductive loads,and is easy to use.

[Kremmen]

[...]
it is very confusing to read such divergent opinions all around the net concerning the control of speed of such motors
[...]

Its because lots of people are confused as to how an induction motor speed can and should be controlled. Those 2 are not always the same thing in every case.

In a nutshell, all AC motors, of which the induction motor is one, create a rotating magnetic field in the stator (the non-rotating winding in the motor housing). The rotor follows this rotating field more or less accurately. If the rotor is independently magnetized say with permanent magnets, it follows the rotating field exactly. This would then be a "BLDC" or "permanent magnet synchronous" motor.
Your regular induction motor however does not contain a magnet, but what is called a squirrel cage winding in the rotor. The rotating magnetic stator field induces currents in this cage, pulling the rotor along, causing it to rotate. The torque is generated by magnetic induction, hence the name of the motor. The induction machine rotor never follows the stator field exactly, there is always a certain amount of slip. The slip is the difference in rotation speed of the stator field and rotor, and it is necessary for induction to happen.
When you first connect the motor to an AC supply, the stator field starts rotating at the supply frequency while the rotor is stationary. There is maximum slip and due to effects i am not going to go into, the generated torque is relatively low. When the rotor starts turning, the slip will decrease and torque increases until it reaches maximun with a small(ish) slip. The torque and thus actual rotation speed is a complex function of the load, supply voltage and frequency.

Proper speed control means that you need to vary both the motor voltage and frequency to control slip and keep it in a sweet zone in relation to the actual rotation speed. This is what VFDs or Variable Frequency Drives do. A fly in the ointment however is that for single phase induction motors this idea does not really work. You noted that i wrote earlier that the stator windings create a rotating magnetic field. Only you won't get rotation with just one phase, you get a pulsing voltage only. For this reason single phase motors artificially create a second phase to create the rotation. This second phase can be created in many ways using auxiliary windings and phase shift capacitors or coils. Combining a VFD and a single phase motor is at best a horrible kludge and usually does not even begin to work. All proper VFD controlled induction motors are 3 phase, as are the outputs of the VFDs. 3 phase shifted voltages is the minimum required to create a rotating field.

Back to single phase motors. In certain kinds of application, you can skip VFDs and simply control the motor by controlling the effective stator voltage using a triac or thyristor or whatever. This does nothing to the supply frequency, which is the standard 50/60 Hz all the time. So what is controlled is the amount of slip, by crude indirect torque control by stator voltage. If the load curve is monotonously increasing and sufficiently steep, the rotation speed will follow the stator voltage. Not linearly and not with world class dynamics, but it will follow. Now it happens that a fan is typically this kind of load and for that reason this simple scheme works quite OK for most fans.

[NiHaoMike]

I once tried using a ceiling fan speed control to slow down a reciprocating saw (which tended to burn up expensive carbide blades at an alarming rate when cutting cement board) and while it worked, the saw was easy to stall after slowing it down to a usable speed, making the cutting very slow.

A split phase motor will run as a 2 phase motor because that's what it is. When doing speed control, it won't work as well as a proper 3 phase motor, but the fact that small split phase motors are common and cheap makes the idea really attractive for low cost applications. That's particularly true of refrigeration and small A/C compressors where the 3 phase versions are very difficult to find.

[fcb]

Kremmen: Good & valid explanation.

I'm assuming the OP isn't expecting the fans to go faster than nominal and the torque required at a lower speed will be lower, so the motor can probably afford to be operating with less efficiency.

I haven't really thought through the physics - but as a rough stab, I'm guessing the square-law fits in somewhere and some fluid theory....  Either way, wiring a fan to a dimmer switch might be a quick way to prove it.

 

[David_AVD]

A quick Google seems to suggest that leading edge dimmers can control ceiling fans, but you need to derate them considerably.  One model has a 600W rating (for lights) but only a 75W rating for use with ceiling fans.

Some of the manufacturers such as Clipsal and HPM seem to have 3 mechanism types.  Leading, trailing for lights and "fan controller" for fans, but they don't detail what the technical difference is.

[zundk]

Thank you very much Kremmen for your clear and concise explanation
it resumes in one single bit of text all what we have been strugling to learn from scratch here and there on the net during the last weeks

in the next few days we will concentrate in physically testing several variations of moc/triac +  RC snubber circuits

we will try diferent ways of controlling the optocouplers with the arduino, with as well as without zero-crossing syncronisation and delay for phase angle control...

after our tests i will report here our success or failures
If i say nothing after next weekend it will mean we've been electrocuted and burried 

If the motor is a 2-phase type with a phase-split capacitor, the capacitors purpose tends to be to create an approximate 90 degree phase shift for the 2nd phase winding.  This would imply that changing the frequency going in would change the phase shift also, resulting in some loss of efficiency - perhaps small variations wouldn't be significant - but running the thing on 10Hz might not work.

Ok, I understand that now, thanks

I'm assuming the OP isn't expecting the fans to go faster than nominal and the torque required at a lower speed will be lower, so the motor can probably afford to be operating with less efficiency.

no, we just expect to be able to go from 0 up to a certain medium speed, i don't think we need to go very fast.
we dont care about the torque nor eficiency and we don't even need any high precision control

we just want the arduino to "influence" the fans to go a bit faster or slower
it is important though that we can go down to completelly stop the fans and start them again at any moment

[nctnico]

I don't know what the budget is but for similar situations I just bought an AC servo motor + servo controller. The controller will usually take 0..10V as an input for speed and / or torque. Unless you are into designing an AC motor drive system buying a complete solution is probably cheaper and safer.