Is there any way to filter out TRIAC humming

[localcrack]

As I designed TRIAC based Fan Regulator with microcontroller. I add RC snubber to filter humming generated due to phase cut.
But till some times some Fan generates noise.
Is there any other way to smooth the sine curve.

[Magnox]

Use an IGBT instead of a TRIAC?

[localcrack]

Hello Magnox,
Can you give me some schematic for IGBT to control fan speed.

[Sideshow Bob]

Here is a book (it is the GE SCR manual from 1972). It may be of some help.

[borberk]

In case motor is producing noise I would check voltage waveform of motor supply regarding triggering of triac is regular.

[Magnox]

Hello Magnox,
Can you give me some schematic for IGBT to control fan speed.

Best bet is to look up IGBT app notes on onsemi's or fairchild's (and others) websites and see what fits closest to what you are doing.

[cesare]

On STM site there are several AN about PWM motor drive and you can find also several code example

PWM drive of a motor has both advantages and drawbacks when compared to a triac scheme.

Triac scheme is normally very simple and cheap for several reasons :

1) The triac can be driven directly from the microprocessor

2) The triac is switched at mains frequency so that switching losses are negligible and often you do not need a dissipator for the triac

3) Average power needed to operate the triac can be  very low because you normally need only one trigger pulse for each mains semi-cicle

4) Usually the triac scheme use a low cost capacitive supply

The main drawback of the triac drive is the noise produced by the motor while the triac is partializing. The noise is due to the High current peaks.

The PWM drive runs more smooth than TRIAC because of high PWM frequency (> 20 Khz). The drawbacks of PWM are:

1) High switching losses due to high switching frequency (normally a dissipator is required)

2) The IGBT (or high voltage mosfet ) cannot be driven directly by the microprocessor and you need some form of driver

3) The IGBT or Mosfet normally require a 10-15V gate drive and so you normally need at least two voltage to supply the regulator

4) The PWM switch must be continously switched and so requires more power than triac and so the power supply normally is not a simple capacitive one

Normally Fan speed controller are realized with triac control.

Regards



 




 

[pickit2]

you never heard of digital logic mosfet?
if you need to control a/c then how about 2 mosfet back to back.

[localcrack]

Please keep in mind that I just want to build normal small fan regulator that will control wall fan or ceiling fan.
I do not want to build giant AC Motor controller.

If you know any application notes then kindly post it.

[cesare]

Here attached an application note from ST for a fan controller.

Regards

[solutions]

If that fan has a synchronous motor, I suggest you do not use one made in China.

You are not going to solve the motor noise problem in a controller, IMO

[localcrack]

@ cesare

I checked ST application notes it uses triac firing method same as mine so it also produce humming in fan motor due to phase angle cut.

 

[cesare]

In my opinion the triac is the correct solution for a fan speed controller. A PWM control would be too complex.

Normally fan motors are some kind of induction motor that run with AC voltages. To drive an AC motor you will need 4 BJTs and the fan speed controller would be very similar to an inverter.

The motor noise depends also on the motor mechanical structure (resonances). Take also into account that some kind of motors are not dimmable.

Regards

[max]

see the attached ckt of a fan dimmer with filter.

[Brainbox]

Is the humming comming from the choke inductor, or is it comming from the fan itself?
Humming from the inductor is inherrent to the ferrite material of a choke.
In that case it would help to chose a more stable ferrite or use an air inductor.

[pickit2]

it's like a silly pissing contest, with so many opinions on a topic, with so little information supplied by the original topic starter.

[localcrack]

@ pickit2

what information is missing ?
I think most member who replied to this topic is clearly understand what I want to say.
is anything is missing then please give your suggestion

[pickit2]

@ pickit2
I think most member who replied to this topic is clearly understand what I want to say.

That's the problem, you could have used so many different components and types of motor.
every one who has worked with a triac  can have a number of fixes for noise in a fan or motor.
Some I have used in past, include snubber , multi-pulse trigger of triac and zero crossing, that being said, Who offered these?
or did they automatically think you had taken these steps to remove motor noise.
That being so, you would not have had any problem, and no post would have been made on this topic.

As said by most who answered with a part solution was hinting, to move away from a triac, The Last time we used triac in a product was in about 1992.

I still think it is a pissing contest.

 

[Sideshow Bob]

@ pickit2

what information is missing ?
I think most member who replied to this topic is clearly understand what I want to say.
is anything is missing then please give your suggestion

What strikes me is that you have not bothered taking the time to even post a schematic. Asking for help in an electronic question and not posting a schematic is asking for a futile thread like this. You have also been given a lot of pointers but seem more insterested in some magic one page easy to read "application note", that will solve all your problem. I do not think you will find one of those. Next time you ask for help in any electronic problem (or other problem for that matter). Please take the time and provide enough details so other not sitting at your workbench can really understand your problem. Good questions draw good answers

[metal]

come on, even with a schematic, he can't understand that this hum/noise is the very nature of the thing he is trying to do,  let's not forget this very basic info about such a project...

[#tester]

I just seen why, it is humming

LOOK>> It Don't know the words<LOOK

[Sideshow Bob]

Strange this thread did not give any good vibes. Now I am just humming alone and feel entertainment, how come?

[localcrack]

I use same method described on this page. On this page it uses Ardunio (Atmel AVR chips) however I am using Microchip PIC 16F877A

http://playground.arduino.cc/Main/ACPhaseControl



after zero crossing triac will be off till t1 time
Then we will give triac firing pulse of t2 time
Triac will remain on till next zero cross
again same steps repeated for the next half cycle of sine wave.

I am using following components that shown in attached schematics.

[Sideshow Bob]

Did you open the book I posted. It has a sections about noise reduction(16 and 17). Also please inform the rest of us how of the humming appear. A picture of your full setup would also be nice

[pickit2]

You missed the snubber for the opto...
did you google the data sheets?

sorry it still don't even come under the term project.

damn I just lifted this of a data sheet...

[localcrack]

You missed the snubber for the opto...
did you google the data sheets?

sorry it still don't even come under the term project.

damn I just lifted this of a data sheet...

it is a snubber for firing a triac, it's not necessary because I am using BTA12-600BW snubberless triac.

Posted on: September 11, 2014, 07:17:48 19:17 - Automerged

Did you open the book I posted. It has a sections about noise reduction(16 and 17). Also please inform the rest of us how of the humming appear. A picture of your full setup would also be nice

I read that section and I added RC snubber in my design

[pickit2]

it is a snubber for firing a triac, it's not necessary because I am using BTA12-600BW snubberless triac.

Posted on: September 11, 2014, 07:17:48 19:17 - Automerged

I read that section and I added RC snubber in my design

So what function is R1, C3 doing in your design? is it Belt&braces?
so the problem I can see is you can't follow data sheets, or take in what people tell you. see Attachment marked in red
I see you are using a MOC3021 and I have shown you part of a data sheet that has snubber components for that device.

you then add (again not revealing your BOM) your using a snubberless triac, then you respond to Sideshow Bob with "I added snubber in my design".

My thoughts on this topic have not changed, your saying its your design, yet you post a concept diagram that is in all data sheets for triac firing.
you have not added to design, taken from a data sheet, but you taken away components from those shown on the data sheets, I have read.

 

[Sideshow Bob]

Not meant to be nippy. But have you done a serch like this https://www.google.no/search?q=triac+motor+control+noise+reduction
Quite often noise reduction is not doing just a single approach. But the sum of several approaches or attack angles

[Ichan]

But till some times some Fan generates noise.

Would you share what fan generate noise and what fan is not?

-ichan

[PeterMcMonty]

About triac driving and snubber networks, for several years I used triacs to fire asyncronous motors (about 1/2 to 1 HP) with phase shift capacitor for gate opener controllers. I always used BTA12-600 triacs, that are very robust, and MOC3061 (or MOC3062) zero crossing optotriacs to drive triacs.
When I needed torque control I substituted zero crossing optotriacs with non-zero crossing devices such as MOC3052.
Then I have some hints about triac driving:

1) for 230 Vac I suggest to use MOC3051, MOC3052 or MOC3053 instead of MOC3021, because MOC305x are rated to 600 V while MOC302x are rated to 400 V only, and when I work on AC mains I prefer something that gives me more confidence, expecially when you consider the voltage across motor phasing capacitor.

2) I found a snubber network that is derived from that published by pickit2, but shares the same capacitor to snub both triac and optotriac thus saving one HV capacitor. I include a schematic (in this case for very light loads, but values were tailored for 1/2 to 1 HP async single phase motors and I lazily duplicated them in this design).
BTW: even "snubberless" triacs perform better with a snubber network.

3) It's not this issue, but when driving an async motor with two of such circuits may be that accidentally they fire toghether, shorting the motor phasing capacitor (that could range from 4 uF up to 25 uF and more, depending on motor characteristics). Triacs are very robust devices and they can whithstand up to ten times the rated current for half an AC mains cycle (i.e.: up to 120 A for BTA12-600), but they are destroyed by the huge dI/dt that occurs initially when the capacitor is short circuited. To protect them from the initial current rising too fast, I found that a small 33 uH / 2 A coil in series between each triac and the load is sufficent to smooth the very first rise of the current before saturating, thus saving the triac. These small coils are not shown in the included schematics.

[PeterMcMonty]

Another suggestion: if the noise of the fan motor, complained by localcrack, is due to the triac switching, rather than cycle partialization you can do cycle skipping: use MOC306x optotriac with zero crossing (as in my schematics) and a zero crossing detector (D1, D2, R25, R26, R27 and Q5 in my schematics) to drive the optotriac N over M cycles (N < or = to M).
Transistor Q5 is on (and its collector is slightly above GND) when AC mains approaches zero volts until it rises agani by some few volts of opposite polarity.
By adjusting R26 and R27 values you can adjust the pulse width at Q5 collector: when the microcontroller detects this signal going down it can drive or not the optotriac, depending by the duty cycle selected; the drive pulse can be removed when the signal returns high: the more precise zero crossing circuit, internal ti the optotriac does the rest of the job.
This signal may also be used as a rather accurate real time clock for the microcontroller.
Of course a pull-up resistor is needed at Q5 collector: this may be the internal pull-up of the microcontroller input or an external resistor.