Maker Pro
Maker Pro

PWM frequency ?

C

cb

Jan 1, 1970
0
Greetings,
Can anyone tell me if there is an ideal frequency, in terms of efficiency, for PWM speed
control on a low voltage DC motor ?
If this is correct can the frequency be calculated or is it a matter of trial and error ?
Thanks
CB
 
P

Paul Burridge

Jan 1, 1970
0
50khz to 200khz is the mid range of whats used mostly in pwm controlers i
think, 50khz is usualy considered a minimum to avoid it being audible and to
use reasonably small inductors, (some setups can miss every other pulse)
200khz is quite fast, so you need smaller inductors etc , 500khz is
acheivable but you start to need to use special ferrites and special
switching considerations .. for motor drive however you can just use the
inductance of the motor so particularly hi speed isnt an advantage.

You're only making life difficult for yourself with these higher
frequencies. 400hz is my preference. The motors generate an eerie
howling noise as the speed changes, which is quite cool - at least it
was in the fighting robots I was building at the time.
 
C

colin

Jan 1, 1970
0
cb said:
Greetings,
Can anyone tell me if there is an ideal frequency, in terms of efficiency, for PWM speed
control on a low voltage DC motor ?
If this is correct can the frequency be calculated or is it a matter of trial and error ?
Thanks
CB

50khz to 200khz is the mid range of whats used mostly in pwm controlers i
think, 50khz is usualy considered a minimum to avoid it being audible and to
use reasonably small inductors, (some setups can miss every other pulse)
200khz is quite fast, so you need smaller inductors etc , 500khz is
acheivable but you start to need to use special ferrites and special
switching considerations .. for motor drive however you can just use the
inductance of the motor so particularly hi speed isnt an advantage.

Colin =^.^=
 
G

Gareth

Jan 1, 1970
0
cb said:
Greetings,
Can anyone tell me if there is an ideal frequency, in terms of efficiency, for PWM speed
control on a low voltage DC motor ?
If this is correct can the frequency be calculated or is it a matter of trial and error ?
Thanks
CB

Since the most power will be wasted as the FETs are in the process of
switching, the less switching you do the better. Therefore I think a
lower frequency would be more efficient.

As Colin said, you may want to switch well above audiable frequencies to
avoid the motor whining or buzzing, but if you don't mind that then a
frequency of a few hundred Hz should be fine, and as Paul said, in some
applications the noise could be considered an advantage.

Gareth.


--
 
P

Paul Burridge

Jan 1, 1970
0
Since the most power will be wasted as the FETs are in the process of
switching, the less switching you do the better. Therefore I think a
lower frequency would be more efficient.

Depends how much time overall your FETs spend in the ohmic region, I'd
have said. That surely is the overall measure of efficiency: whichever
gives rise to the least amount of heat produced.
 
M

MG

Jan 1, 1970
0
colin said:
efficiency,
for PWM speed

50khz to 200khz is the mid range of whats used mostly in pwm controlers i
think, 50khz is usualy considered a minimum to avoid it being audible and to
use reasonably small inductors, (some setups can miss every other pulse)
200khz is quite fast, so you need smaller inductors etc , 500khz is
acheivable but you start to need to use special ferrites and special
switching considerations .. for motor drive however you can just use the
inductance of the motor so particularly hi speed isnt an advantage.

Colin =^.^=

Why inductors? I thought that in this kind of motor controller the motor
provides its own inductance, there is no need to *filter* the square wave.

If this is the case then the winding current will have a triangular ripple
superimposed, as long as the current ripple is low and the torque smooth the
lower the freq the more efficient. Yet once the switching losses are much
smaller than the conduction losses there is no need to lower the freq below
that.

MG
 
K

Ken Smith

Jan 1, 1970
0
MG said:
Why inductors? I thought that in this kind of motor controller the motor
provides its own inductance, there is no need to *filter* the square wave.


Motors aren't very good inductors. The high frequency components in the
switching waveform tend to cause losses. As a result, you can improve
matters by putting a smallish inductor in series with and perhaps a small
capacitor across the motor. This will reduce the losses by keeping the
high frequencies out of the motor.
 
C

cb

Jan 1, 1970
0
cb said:
Greetings,
Can anyone tell me if there is an ideal frequency, in terms of efficiency, for PWM speed
control on a low voltage DC motor ?
If this is correct can the frequency be calculated or is it a matter of trial and error ?
Thanks
CB

Thanks to all for your replies. Will do some experiments and report back if there is anything
interesting.
CB
 
Y

Yzordderex

Jan 1, 1970
0
One other note. Or two.

Is application one where you will need a lot of dynamic control? If
so the minimum pwm rate may be a function of the bandwidth required by
controller.

Also (as peviously mentioned) you may want to consider audible noise.
Certain pwm rates can be very nasty to live with.

You may also want to consider resonant action in the switching in
order to minimize switching losses. Certain topologies are prone to
these advantages. If you can minimize losses by resonant action or
perhaps with some combination of turn-on and turn-off snubbers you may
like the effect.

Best of luck,

Bob
 
Top