Fan Regulator Using Arduino

Check out this website. They have the projects also, but may have what your problem is.

http://www.instructables.com

Thank you.

I have searched a lot but i could not get the answer for my question.

It needs to be interfaced.

I have interfaced and the circuit works well with a lamp. But if its replaced with a fan it is not starting. I also kept the led side of opto triac continuously on so that the main triac keeps conducting continuously (zero firing angle delay) so that full voltage will be available across the load as i thought due to firing angle delay sufficient voltage was not available and so the fan didn't start. But even then the fan didn't start and when i measured the voltage across load supply voltage was available and the drop across triac was zero. I don't where the mistake is, is it possible that current through triac is not sufficient. I am using BT136 triac.

Kindly help me out.

Are you starting at full power? ...or trying to ramp up from low?....also what type of motor are you using, a DC stepper?

Sorry, but using the Arduino PWM output to drive an AC motor is a non-starter (pun intended). The PWM output is a chopped DC, at a frequency of either 490Hz or 980 Hz, depending on which pin and which Arduino board. This is wholly inappropriate for driving a 50/60Hz motor.

If you want to vary the rotational speed of an AC motor, you need a variable-frequency drive, starting with a signal generator with an output frequency range of 0-50Hz or 0-60Hz, ideally a sine wave. It would be an interesting but quite practical exercise to program an Arduino to generate a sine wave. However, this would be a 5V low power sine wave, and you would need a power amplifier (similar to an audio amplifier but working at low frequencies) to generate the 110/240V you need, possibly in conjunction with a step-up transformer.

Alternatively you use the Arduino to generate square waves of the same frequency range, and use that to switch a high-voltage DC source. AC motors prefer sine wave to square wave supplies, so filtering the output to get rid not the high-frequency components could be a good idea.

^{i am not using pwm. I am using zero crossing detector to sync with mains. After zero crossing i give a delay of alpha milli seconds i.e. firing angle delay and then i turn the triac on. The ckt in that link makes sure that triac is fired in 1st nd 3rd quadrant so that no firing failure occurs.}

I'm afraid that you are still missing the point. Your firing angle delay after a zero crossing point is in principle exactly the same as pulse width modulation. You are generating pulses of a variable width but constant frequency (the mains frequency). All that does is reduce the power. What you need is a variable frequency and full power.

I get your point. You are saying speed control by V/f method in which you change frequency as a result fundamental frequency varies so synchronous speed varies and as a result motor speed varies.

But this is not done in a light dimmer or fan speed regulator. It implements single phase voltage control using triac, just like what happens in a controlled rectifier except for a difference that the output of this converter is still AC.

I hope I am clearly saying. Kindly help me out.

Thank you for your time.

What may be affecting the outcome is that a lightbulb is almost pure resistance and a motor has a lot of inductance.....This changes things! Significantly!

Current and voltage are no longer in phase with each other when either capacitance or inductance are in the mixture........most mains light controllers, sold for in the home, cannot handle inductance, there is a warning on the packet to tell you so....your circuit is the simplest I have seen, so that may be part or even all f the problems.

I do not know how I can put this any more simply.

1. Your post #6 states that it is an induction motor.

2. The rotational speed of an induction motor is related to the frequency of the applied AC, see here

3. Chopping an AC waveform does not alter its frequency.

Either your motor is not an induction motor or your commercial fan speed regulator does not function in the way you think it does.

With all due respect, I totally understand what you say. If I correctly understand, you are saying speed control of single phase IM is possible "ONLY" by frequency control. I am afraid I have to say that I have never heard like that.

I have read other methods of voltage control too. Click here.

Kindly see fig. 8 in that link. By "chopping the AC waveform", effective RMS value of the applied voltage is controlled, as result of that, the induced EMF in the rotor is controlled and thus the torque produced is controlled. Due to difference between the load torque and produced torque, a dynamic accelerating or decelerating torque is produced depending upon the sign of the difference. Thus speed is reduced by lowering the voltage at the cost of reduced torque. I apologize and enlighten me if I am wrong.

I definitely agree with the fact that "The rotational speed of an induction motor is related to the frequency of the applied AC" but I don't think that rotational speed depends "ONLY" upon frequency. Correct me if I am wrong.

Thank you so much for your time and patience. Thanks a lot once again extremely sorry if I am wrong.

Sorry, the "thing" that determines power drawn by the motor is the mechanical loading on the fan; i.e., the air resistance. Altering the frequency, voltage, current, etc., are simply ways of varying the electrical power that the fan can supply to move a given quantity of air against a certain resistance.

The probable reason this "system" is failing is improper interfacing/buffering of the low-level Arduino signals with the higher current demand of the firing circuitry. The proper (and easier) way around this is to use the correct Arduino motor drive circuitry and code, especially to avoid the potential danger of interfacing with AC line voltages, to do otherwise is a formula for disaster.

The original request from the OP was to control the fan speed, not the power. The proper way to do that is to use a VFD. The alternative, if one is determined to go down the switch on-switch off route, is to emulate standing by the switch to flip it on, wait until it starts to revolve, and then switch off again. In other words, instead of firing millisecond pulses, the on phase should last at least a second, to allow the motor to start turning. The interval between these 1-second pulses would be the determinant of the overall speed. In that way the poor motor would at least get a chance to start up.

Oh, so some how the power to turn the fan is independent from the fan speed? That's not how motors work, especially under simple fan control, but you should check out the fan affinity laws to verify.

As I hinted, a VFD would supply full power to the fan, whatever the speed. Flipping the on-off switch could also be used to run the fan more slowly, but the power delivered would be less. Clearly you can achieve the same fan speed with different input powers, as long as the motor is only lightly loaded. This becomes relevant at the lower end of the rpm range, where the motor might stall with the on-off switch technique.

I understand your point. You are saying from industrial application point of view; correct me if i am wrong.

The “same speed at different input powers” is the reason we go for VFD.

http://1.bp.blogspot.com/-4lJhIlvX2Lk/VqxXjb4Gb-I/AAAAAAAABWE/sbX2_NOa5mM/s1600/VF%2Bcontrolled.jpg

In the image found at the above link, we can see the speed torque characteristics of VFD controlled IM. Imagine a line drawn parallel to speed axis. This line intersects the family of curves that could be obtained for various values of voltage keeping the v/f value a constant. Till rated voltage, the maximum torque that can be produced by the machine remains constant, that means till rated voltage, maximum rated load can be applied to the machine and can be made to run at various speeds without stalling.

From another perspective this can be seen as for a fixed speed below rated speed, the load can be varied (increased or decreased) “without” affecting the speed (steady state speed, of course). Whenever load is increased, motor will consume more power. Thus at various load torque values, same speed can be obtained. So speed is constant, but torque is variable and hence power is variable. This is the point you are saying, am I correct? The machine will absorb more power “only” when the load increases. But in my case, the load is air resistance and inertia of the fan blades. This is a fixed and light load, isn’t it? So there is no need to ensure that the machine is capable of producing maximum torque. It is just enough if I make sure that my motor wont stall.

This ceiling fan is one device which we use to move the air downwards to get cooling effect (wind chill effect) .I can compromise on the maximum torque producing capacity of fan to control the speed unless it stalls.

So ultimately what I am trying to say is using VFD is way too much expensive for my need. A simple voltage control method will suffice “at the cost of maximum torque” that can be produced.

I think I am correct, please forgive me if I am wrong and enlighten me because I don’t have much practical exposure, only a very little amount of mediocre theoretical knowledge I have. I am sorry if I am not clear. I am not trying to offend you; you have my apologies if you feel so.

Thank you so much for your time.

I am not quite sure, but I think maybe you do not understand how a VFD works.

A VFD, drives an AC motor with a variable frequency, but a fairly healthy voltage. This means that the motor is "synced" minus load slip, to that it revolves at almost the same frequency as the VFD i gibǘing out, as good as nothing to do with the mains frequency!

I have ignored the fact that turning "normal" AC motors too slowly, can cause extra heating, so for "fan" work, they need to be both designed for slow speed as well as having good cooling as well.....

I speculated that you have a motor that covers these points.

So you may find it difficult to control the speed without following the VFD route, also a speculation on my part!

I found this:-

DIY VFD

Which might help you better than the previous links I posted and I expect there to be special chips around.....plus, there must be a circuit somewhere to use an Arduino to provide the intelligence:-

http://www.nxp.com/files/microcontrollers/doc/white_paper/MC3PHACWP.pdf

You might like these videos too:-

http://www.youtube.com/watch?v=XE_vslqL6x0

http://www.youtube.com/watch?v=5l5xAarbbvc

http://docs.google.com/document/d/1kf-5Ncz-gQ3qPFr1N8VTj8d0WK6dz8UdwPOJLtexr2M/edit

http://www.instructables.com/answers/How-can-I-make-a-Ac-to-Ac-converter-with-variable-/http://www.instructables.com/answers/How-can-I-make-a-Ac-to-Ac-converter-with-variable-/

http://www.google.de/search?q=using+arduino+AC+motor+Variable+frequency+drive??&client=ubuntu&espv=2&biw=1041&bih=505&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwjdiuGv0KfNAhVDKcAKHd_6C2EQsAQIXw&dpr=1.25

Will look into it. Thank you once again. :)

You're welcome. In the short term I suggest you use the Arduino to switch on and off at quite a low frequency, 0.1-0.5Hz. Perhaps later I can encourage you to program the Arduino to act as a VFD. The aim would be to start with a 240V DC source, obtained from the mains via a bridge rectifier and smoothed with a capacitor. You then use a switched output via an opto-isolator to develop a stepped waveform resembling a sine wave, to drive your fan at full power at any speed you like.

Thank you. Will do it this way :)

http://www.futurlec.com/Diodes/BT136.shtml

Check I/O specs if its compatible(both Adruino & motor).

Read or research more about Analog, Digital communication and interfacing. Google can help.

If what you say is true, I can only think of a few possibilities (there could be more!), but as you did not mention the exact motor type, that might have helped us all here better to help you further:-

1) Motor is defective - check it on the mains directly

2) Wiring to motor is open circuit or there is a high resistance connection, or even insulation is "trapped" and isolating the motor....

3) If a capacitor start motor (single phase induction), one field is open. Though this usually makes noise implying voltage is getting to the motor....so far less likely.

Guessing only that no current is flowing through the motor!! Which might demonstrate a possible direction to search in, only voltage is being measured, that can be very misleading. Adding a current measuring meter or shunt in circuit can help in such situations as here....

Thank you. The motor is a single phase induction motor but don't know if it's a capacitor start or capacitor run. When it is connected to the mains it works well and when its speed is regulated through a "commercial fan regulator-bought from electrical shop" (which uses RC of variable R to alter the firing angle and triac), it works well too.

I also think that sufficient current is not flowing.

Is there any possibility that with reduced gate current the triac current would be reduced? I know that concepts of holding and latching current. But still the most basic structure of TRIAC is BJT right? So is it possible that gate current could affect current through it?

It looks like your fan can be controlled by a fan controller, so your problem is not related to many of the other possible problems. If your control system can control a light (assuming an incandescent light of similar power rating to the motor), then the problem is related to your Triac configuration and the inductance of the motor. An oscilloscope may be needed to troubleshoot the problem. Be very careful, since troubleshooting an AC line powered system with an oscilloscope may let the smoke out of the ground wire of your probes (along with other potentially lethal results). If you can get an isolation power transformer, you can add a little safety to your efforts.

Do you know that your circuit design was intended to cope with controlling inductive loads? Many Triac control circuits are designed for lamps, and will not work with motors, which is why there is a need for a "commercial fan regulator-bought from electrical shop."

Thank you so much. I will try to monitor it using oscilloscope once I could get my hands on it and will keep the safety measures that you said in my mind.

I tried to reverse engineer a commercial fan regulator bought from shop, once i figure out the circuit and identify components and their ratings I will add them in my circuit and share the results.

I did not think or expect fan's inductance could cause this much learning experience.
Thank you once again.

Educational story: While in college, I arrived at a friends house as he was ready to power up is "invention" that would "adjust the speed" of his Breeze Box fan. He had wired a radio volume control potentiometer in series with the power to his fan. As he plugged his invention in to the wall outlet, a shower of sparks was ejected from the potentiometer. He learned that his invention didn't work . . .

I made some similar "discoveries" as a kid.

Anyone here who denies ever making such mistakes is lying to himself......

The trick is to not kill yourself......

I had a similar experience with triacs and transformers. The inductive phase shift was enough to prevent the triac switching to fail full on or full off. I had to insert a 10 ohm, 100 watt resistor in series to correct the phase angle to something the triac could work with, but the resistor was a major heat source. Effectively what you are doing here is underdriving the motor so that speed is limited by the power supplied from the triac. The motor will likely run hot due to the underdriving of the motor, as will the triac due to heat generated during switching on of the triac. You probably would be better off switching to a DC motor and full wave rectify out of the triac, depending on how much power the motor draws. The result is that the motor will likely have a shorter life and you will occasionally let the smoke out of the triac.

The alternative is to pick a triac that is way over rated for the load and accept the extra heat generation.

Thank you so much. Your words help me visualize the problem. Although Andygermany and Carl van wormer said it before, I did not understand. Sorry guys.

So the problem is fan's inductance. I sense voltage and sync with it. But definitely current will be lagging and so this conduction issue arises.

I didnt visualize this current wave with the phase shift in the graph. So stupid of me. I will see wht the commercial regulators do for phase shift and share it.

Thanks a lot everybody for your time and patience.

Back to basics:

What is the purpose and power rating (and operating voltage) of the fan?

What is the problem that we are trying to solve? (why do you want to control the speed of the fan?)

The answer to these questions may help to focus the discussions.

I am trying to control ceiling fan speed. I am from India, fan is induction motor and runs on 220v 50hz supply. We have a regulator- to control its speed- attached to the socket board which is similar to the light dimmer.

I did a project to turn on/off fans and lights in a room using a remote using arduino. Now i want to improvise it by controlling the speed of fan too so that we dont have to go and adjust speed, we can simply control it using remote just like how we turn on/off the devices.

Using arduino i alter the delay after zero crossing so that speed can be controlled.

What is the power rating (or current rating)? Is this a small (200 watt) room ceiling fan, or an industrial building high power (1/2 hp or more) fan?

In using the arduino to control delay, you need to turn on the arduino output and hold it on, ideally until the AC to the fan crosses zero volts and amps. If you are taking the output high for a few thousandths of a second, you may be missing the points in the AC waveform where you can turn on the triac and have it stay on. The phase shift between current and amps complicates this. You may also have timing synchronization issues between the arduino on/off signal and the AC waveform, unless your software senses AC zero crossing and starts your time cycle from the zero crossing.

You could use an SSR10VA type solid state relay and use an analog output from the arduino to a transistor to control the SSR. The SSR uses a resistor across it's control terminals and a suitable transistor could be used in place of the resistor.

But then, if the fan is currently controlled by a variable resistor, you could substitute a power transistor for the resistor and use the arduino analog output to control the transistor and leave the rest of the circuit intact.

Its small ceiling fan. Not a big one.

You said you connected a resistor to reduce phase angle. In a commercial regulator which i reverse engineered there is an inductor in series with gate. Will this bring phase delay in gate current so that i matches mains current? Is this possible??

The resistor I used in series was on the load. In practice usually a capacitor would be added in series with the load to correct the phase angle. Using a resistor in series with the load reduces the inductive lag a bit on the load. In my case it was enough of a reduction to get the triac to function properly. I used the resistor because big capacitors tend to be spendy.

Inductance causes the current to lag behind the voltage. My guess would be the inductor on the gate is intended to cause the current lag at the gate to match the lag in load generated by the fan so that switching more closely matches with load conditions. You might try cannibalizing an inductor from a controller for the gate on your triac and see if it works.

I have done this but with a commercially available remote control. I used a motorised pot in a triac fan speed regulator, and wired the motorised pot to it. However, the motorised pot is big and needs enough space. I fitted it in the electrical panel which was luckily just above the switches.

Gajanan Phadte

Application_Notes.pdf

This document may extend your knowledge a little further.

Also, there are specialised chips around to do the job better I believe.....you will have to search on google for data sheets. Like this one:-

AC Motor Speed Microcontrollers.pdf

One other point, covered in this PDF a bit, is to find out exactly what type of single phase AC motor a ceiling fan uses!!! They are not "normal" motors, so to say!!! But normal for ceilings!!

Also, looking here may help:-

Fan schematics

The one I personally liked most was this one, which could be converted to Arduino once its fully understood and working I feel:-

Cheap AC Fan Speed Control

Thanks a lot for the links and your time. Thank you all for spending time and helping me out :)

I would be remiss if I did not thank everyone for this interesting blog, which has demonstrated to me clearly that my personal knowledge in this area is very poor.

So thanks to all for making such a problem so interesting!!

You'all have a great day!!

The fan motor is a squirrel cage motor with a single capacitor, permanently wired (in series) to one of the two windings.

Gajanan Phadte

Then I do believe that a relatively simple VFD will control its speed fully correctly!

If you are not up to making one, then something like this is relatively cheap:-

$100 VFD on Ebay

Best of luck!

How about this for under $20.00

This is unlikely to work correctly as drills/routers are either battery DC or mains universal motors......both of which are speed controlled by simple voltage levels changes.

The AC mains universal motor has a strong resemblance to DC motors.....it has brushes!!

What he needs, is a VFD for a single phase induction motor, this is not one!!

Sorry.