dc motor speed control by pwm,newbie seek help

[iclok]

I am controling a dc motor speed with PWM.By triggering mosfet with that pulse, the output voltage is not as expected when the motor is connected. For instance,suppose duty cycle of the pulse is 10%,average voltage between the motor is much more than 10% of Vs(around 30-40%). However when the motor is taken out(no load),the voltage measure between the motor is acurrate, thats mean Vout is nearly 10% of voltage source. Why is it happen? Is that i need a snubber circuit for mosfet? How snubber circuit work? How we calculate the value of R and C? Is that this is the solution? Or the problem occur because of motor? I have add a diode across the motor already.

 

[Styx]

you say when the motor is not there the voltage is correct? since it will be pulsing how are you measuring it?

If it is a DVM then ok the average voltage would come out.

The addition of a high value capacitor (with sufficent current ripple rating) AT the motor +ve terminal will provide tank capacitance to the motor and make the voltage supply look more like a voltage source than a current source.

Also add a low ohmic 10Ohm in series with the Gate, this will stop the creation of a pearce oscillator..

Also that diode, is its characteristics matched to the FET's switching performance, if it isn't you could get turn-on problems


Since it looks like you are wanting to maintain the voltage I would suggest you put a close-loop around the system to maintain the voltage. This way you will be able to maintain the voltage. Remeber a motor is in effect just an inductor.


As to a snubber for this type of circuit with this type of voltage/power it will not be needed.

A snubber is only used when switching losses are such that it can damage the switching device or the turn-off voltage overshoot is very high that it will kill the device.

Snubbers are the very last resort. Better layout of power devices with low-inductive busbars are always better. All snubbers do is move switching losses from the device into the snubbers R component..

 

[iclok]

you say when the motor is not there the voltage is correct? since it will be pulsing how are you measuring it?

If it is a DVM then ok the average voltage would come out.

The addition of a high value capacitor (with sufficent current ripple rating) AT the motor +ve terminal will provide tank capacitance to the motor and make the voltage supply look more like a voltage source than a current source.

Also add a low ohmic 10Ohm in series with the Gate, this will stop the creation of a pearce oscillator..

Also that diode, is its characteristics matched to the FET's switching performance, if it isn't you could get turn-on problems


Since it looks like you are wanting to maintain the voltage I would suggest you put a close-loop around the system to maintain the voltage. This way you will be able to maintain the voltage. Remeber a motor is in effect just an inductor.


As to a snubber for this type of circuit with this type of voltage/power it will not be needed.

A snubber is only used when switching losses are such that it can damage the switching device or the turn-off voltage overshoot is very high that it will kill the device.

Snubbers are the very last resort. Better layout of power devices with low-inductive busbars are always better. All snubbers do is move switching losses from the device into the snubbers R component..

Ya I measure the average voltage. The main problem is when there is no load the average voltage come out accurately. When the motor is connected, the voltage measure become larger.
Actually I am not very clear to your suggestion may be can you please show me the connection of circuit? May be the "close loop"?