Hey guys!

I'm doing a hypothetical design project of a MEMs vibrometer to go on a jet engine.

As the jet engine would obviously be very hot, I am designing a cooling system with some mechanical engineers. They've decided a fan should be run which looks something like this:

https://www.digikey.com/product-detail/en/delta-electronics/AFB0405LA-A/603-2014-ND/5799904

I'm currently designing the circuit to run it, and was wondering, how should I control it? I know for a real motor, PWM would be used, but is that necessary here? Could I just use an analogue signal?

Cheers

 

You can vary the run speed of a dc motor either by varying the DC voltage or by varying the apparent DC voltage - -IE. PWM. The later gives better torque and higher efficiency

 

Most of these small fans have brushless motors with the electronics inside. They don't really like the supply voltage being changed. There are some types that have extra connections for PWM speed control. PWM is used because it is easier for a micro-controller to generate a PWM output than an analog output.

See this for an example of fans with PWM input.

 

Most of these small fans have brushless motors with the electronics inside. They don't really like the supply voltage being changed. There are some types that have extra connections for PWM speed control. PWM is used because it is easier for a micro-controller to generate a PWM output than an analog output.

See this for an example of fans with PWM input.

so if theres only two connections for power, and i can't change the supply, presumably i have to use PWM?

 

Or have the fan going full speed all the time - does it matter?

 

Or have the fan going full speed all the time - does it matter?

aye, i reckon so - its got to be a regulated temperature?

 

Unless you really have a need for fan speed control as in increasing fan speed based on temperature rise I wouldn't worry about it. Realizing this is a hypothetical project in reality you would first look at the temperature ratings of your sensors, determine how much heat you have and how much heat you need to remove. Then decide on a fan based on airflow... I am sure you get the picture. Looking at what you have I would likely just run a fan at full speed rated voltage.

Years ago I did do some experimenting controlling fan speed by just changing the applied analog voltage. Using a fan similar to what you linked to I applied 12.078 VDC. The current draw was about 157.4 mA and rotational speed was 2,739 RPM. I found that the fan speed was proportional at 9 VDC and 6 VDC. I did not measure the airflow.

Ron

 

Hi,

Just to note, if you really do need speed control there are the two types, linear and PWM.
The PWM type breaks down into two or more types too though, the ones made for PWM and the ones not made for PWM.
The ones not made for PWM (two wire and many three wire units) probably should not be run on PWM because that means constant banging with the full voltage over and over again many times per second. This could have several undesirable effects on the life of the motor.

I notice this particular fan is very low current. It looks like must 30ma? That would lend itself to linear control without consuming too much power. At 10v a 30ma load only consumes 300mw. That's less than 1/2 watt. That is at max power, and anything less than that will consume less power. For example, at 15ma (roughly 1/2 the voltage too now) the total power dissipated by both load and control circuit is around 150mw. I would be surprised if that was a problem even thought the efficiency isnt that great.

But as others have also mentioned, the fan may not need control. Running full speed may not bother anything except for the life of the fan which will be reduced over that if it only had to dun a partial full speed for many hours.

I would also make sure that fan is actually enough to actually cool the required components. It's a very small fan.

 

Unless you really have a need for fan speed control as in increasing fan speed based on temperature rise I wouldn't worry about it. Realizing this is a hypothetical project in reality you would first look at the temperature ratings of your sensors, determine how much heat you have and how much heat you need to remove. Then decide on a fan based on airflow... I am sure you get the picture. Looking at what you have I would likely just run a fan at full speed rated voltage.

Years ago I did do some experimenting controlling fan speed by just changing the applied analog voltage. Using a fan similar to what you linked to I applied 12.078 VDC. The current draw was about 157.4 mA and rotational speed was 2,739 RPM. I found that the fan speed was proportional at 9 VDC and 6 VDC. I did not measure the airflow.

Ron

cool thanks man!

Would you additionally know how I would check whether these fans are series or shunt connected DC motors? (I guess I'm just assuming they are DC motors)

 

Says on it, DC brushless, (BLDC). P.M. rotor.
Max.

 

Says on it, DC brushless, (BLDC). P.M. rotor.
Max.

does that mean shunt or series? I dropped power engineering pretty early in my university career haha

 

does that mean shunt or series?

Neither. The rotor is a permanent magnet one.

 

This is the data sheet for your fan link. When you see fans like this the motor is a PM (Permanent Magnet) type motor. Keep in mind permanent magnet means the motor does not have a field winding, there is no field needed since the permanent magnets create the field. I see Alec already addressed this as I was typing. You may want to give this a look.

Ron

 

does that mean shunt or series? I dropped power engineering pretty early in my university career haha

It is effectively a shunt field, the field and rotor are in parallel.
It would be hard for a magnetic series field!
BLDC is a DC commutated motor turned inside out.
Max.

 

BLDC is a DC commutated motor turned inside out.

I like that and it actually is.

Ron

 

thanks to everyone!

 

thanks to everyone!

Hi,

I think i speak for everyone here when i say, "You're welcome"

 

(thought i was out of questions but here we are again)

Can i assume linear change of air flow with either PWM or regular voltage control?

For example at either 0.5 duty cycle or 2.5V linear control, could i assume half air flow?

bare in mind i won't be able to test this fan in reality

 

If directly related to fan rpm, generally the output flow curve is non-linear, with a low gradually increasing curve up to a certain rpm (~50%) where it climbs steeply up to max rpm.
IOW without able to do empirical tests I would tend to control from 50% rpm to 100% rpm for an anticipated result, assuming with a BLDC motor that 2.5v will be .5 x max rpm.
Max.