Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Variable speed fan drive circuit is OK?

Status
Not open for further replies.

Flyback

Well-Known Member
Hello,

Please could you check the attached fan drive circuit for driving the "ebm papst 8218 JH3" fan? Please advise if this circuit is OK?

Basically, as you may see, it is driving the fan from a Buck converter.
Do you believe that the 10uF capacitor across the fan is enough capacitance?
The PWM is at a frequency of 100KHz, and its duty is variable by a microcontroller.

Please could you also supply the coil inductance of the fan?

Fan datasheet:
**broken link removed**

Fan series datasheet:
**broken link removed**
 

Attachments

  • Fan drive circuit.pdf
    13.7 KB · Views: 284
Why such a high PWM frequency? The 18K to 20K is normally for motor drive circuits. To keep it in the high range of human audio hearing, so it is not heard when operating.
 
Why such a high PWM frequency? The 18K to 20K is normally for motor drive circuits
If you are driving the motor with AC. BUT here the PWM is filtered to DC so the switching frequency is not important. I like 100khz or higher because the inductors and capacitors can be small.
What is the point of the current monitor?
I agree. Why?

This fan was not designed for variable speed. You will need to experiment. I have found that when the fan gets old it will start harder than when it is new. I start the fan with 100% for 1/2 second or until the fan is up to full speed. Then drop back to slow speed. If you start out with 50% the fan might start now but will not when it is old.

Do you use LT Spice? I would build the MOSFET and fan part of the circuit with out the current sensor. In place of the fan use a 96 ohm resistor. I think the inductor should be 470uH 1A. Using spice you can see what the ripple voltage is on the capacitor.
 
Surely this fan is designed for variable speed, otherwise whywoudl it say its supply voltage can be 20 to 58V?
The current monitor is needed because certain currents correspond to certain speeds, and we must monitor the current, the micro reads the current and varies the duty cycle which speeds up/slows down the fan to the required speed.
Do you think its a single phase BLDC fan, or a 2 or 3 phase BLDC fan?
Also, we have a 100u electrolytic across the fan right now but would like to make it non electrolytic and just 10uF, do you think that would be OK?

Regarding start up, surely this is a potential problem due to the BLDC not locking up to the drive signals?....whether they be from an encoder or resolver, or from back-EMF sensing....back emf sensing is more likely for space/cost reasons.

Where on the datasheet does it say its a fixed speed fan?
 
Last edited:
Surely this fan is designed for variable speed, otherwise whywoudl it say its supply voltage can be 20 to 58V?
I looked all over the data sheet for voltage range and did not see it yesterday.
Today I had the computer look for "58" and there are no "58"s. It also could not find "20" having to do with voltage. None
My PDF reader has 1/2 the pages blank. Maybe it is there.

Flyback, You know PWMs. You know how to find the peak current in a inductor. You know how to find the ripple current in a capacitor. (ripple voltage)
The fan data sheet does not give a hint as to capacitance on the supply. Or at least I can't see it.

1) drive the fan from a bench supply. Separate the supply and fan with a 10mH inductor (and a 100 ohm resistor ) and use the 10uF cap. Measure the capacitor voltage ripple produced by the fan? Can you live with that ripple?

2) Use math to see ripple voltage caused by the PWM. Can you live with that.

Recently I had a load with a high frequency switching frequency. I used a capacitor with a low resonant frequency so the cap was inductive at that frequency. The whole thing oscillated violently. The voltage would swing so low that it reset the internal oscillator. I have no idea what is inside the fan and how it will react to 10 volts of ripple. You have to determine that be experiments. Also you capacitor ripple current could be >0.5A.
 
The 20V to 58 V is in the "fan series datasheet" in the top post,
The fan series datasheet says that this fan has a "PWM control input" but gives no explanation of it. Does anyone know what it is?
 
Check out the 4-wire fan description here
 
The 20V to 58 V is in the "fan series datasheet" in the top post,
The fan series datasheet says that this fan has a "PWM control input" but gives no explanation of it. Does anyone know what it is?
Is this the data sheet you are talking about?
upload_2014-7-28_21-15-30.png
 
Strange. That fan has only two wires visible, so no 'PWM input'. The spec also says "Protected against reverse polarity and locking." IMO that implies some internal circuitry, which may or may not be happy if the supply is PWM'ed.
 
OK now I see. It is not in the fan data sheet. It is in the fan series data sheet. Sorry for my confusion.
 
My two yen on the subject;
Years ago, we used a good quality fan on a high volume product.

Similar to your proposal, we also used lightly filtered PWM to slow down the fan (to lower the noise).

Within a year, we started receiving failed fans, which were very early in their service life. Remember, this was an expensive and high quality fan.

To make a long story short, the explanation was that the PWM would somehow cause additional vibration. This would cause premature bearing failure.

I personally found this very hard to believe. However, we implemented the recommended corrective action, and the failures disappeared.

My recommendation: contact the fan manufacturer's technical support department and ask for their opinion.
 
However, we implemented the recommended corrective action, and the failures disappeared.
What action?

I had the same problem. I blamed it on old bearings. (1 year) My action is to start the fan at full voltage and then drop back to slow. The fan failure rate is very good.
 
Lower the ripple applied to the fan.
It helped us in our situation. But..it may or may not apply to the particular brand/model fan you are using.

However, my recommendation universally applies: When in doubt, contact the fan manufacturer's technical support department and ask for their opinion.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top