I want to controle a small dc motor, blower fan, using a pwm module. The actual setup is a 12v DC transfo plugged to a pwm module. The motor is connect to the output of the module.

At low speed, the motor won't start. I do understand now that these motor need minimum voltage or current to get in movement.

How can I supply an higher voltage for a brief moment to kick start the motor then let the pwm module to control it.?

 

For a start, does the motor have a reverse diode across it?
And, as you have PWM, just try a higher PWM % for a short burst to get it moving. But, if you want consistent low speed ( or any speed ) control, some feedback indicating the motor speed will be needed, and then a control loop to adjust the PWM accordingly.
If your fan has a third yellow wire, a speed signal can be found on that. I'm not sure if it is a pulse or a voltage.

EDIT: It looks like it may be pulses...
http://pcbheaven.com/wikipages/How_PC_Fans_Work/

 

This might do the trick: http://www.romanblack.com/onesec/DCmotor_xtal.htm

 

Roman's solution is for a DC brushed motor, it depends on whether the OP's motor is a 2 wire PWM BLDC fan with internal electronics or a simple DC brushed motor.
Max.

 

To start a DC motor at an extremely slow speed and control it smoothly to the desired full speed was investigated in depth back in 1969 by Linn Westcott

It was a discrete component circuit built for model railroading. The problem was the same. Modelers wanted their model locomotives to start smoothly. Prior to development of this circuit, the locomotives would start like a jackrabbit, going from 0 to 60 in a flash.

So you all aren’t model railroaders. Why did I bring this up? The published article contained detailed graphs of the output waveforms. As well as an incredible amount of technical detail.

This may be of great value in developing a modern solution. The output waveform included varying pulse widths (from narrow to wide) which decreased in amplitude (from maximum to zero) on top of a slowly increasing (from 0 to maximum) pure DC component.

Search for “TAT IV Throttle”, possibly adding “Linn Westcott” to your criteria.

 

To start a DC motor at an extremely slow speed and control it smoothly to the desired full speed was investigated in depth back in 1969 by Linn Westcott

It was a discrete component circuit built for model railroading. The problem was the same. Modelers wanted their model locomotives to start smoothly. Prior to development of this circuit, the locomotives would start like a jackrabbit, going from 0 to 60 in a flash.

So you all aren’t model railroaders. Why did I bring this up? The published article contained detailed graphs of the output waveforms. As well as an incredible amount of technical detail.

This may be of great value in developing a modern solution. The output waveform included varying pulse widths (from narrow to wide) which decreased in amplitude (from maximum to zero) on top of a slowly increasing (from 0 to maximum) pure DC component.

Search for “TAT IV Throttle”, possibly adding “Linn Westcott” to your criteria.

Signs of recovery ? Thanks for contributing