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.

converting a PWM output to steady voltage

Status
Not open for further replies.

Thunderchild

New Member
I have a PWM output 0-100 % of 12-15 volts @ 2 A, i would like to pass this through an LC filter to obtain a roughly steady voltage (some ripple is fine), what values should I choose for L and C ? I'm runing 10 - 25 KHz trying to compromise on space and switching frequency.
 
Why not start out with steady voltage?

Saves unnecessary back conversion from PWM to RMS. :D
 
this is the output of a power controller I'm building and it needs to be efficient ! basically is a crude smps output idea but this is a car dynamo regulator, the PWM % is pic generated and varies to control the output. The direct PWM output works but a smother voltage would help
 
If annoying sound of the motor is the problem you might be able to solve it by connecting an electrolytic cap across the motor terminals.

I tried it on a PWM circuit generating 300Hz PWM with loud noise at low rpm of the connected fan.

Adding an electrolytic of 47µF across the fan motor had a double effect:

- reduced noise
- better torque at startup

Boncuk
 
this is powering the field coil of a dynamo, its a possibility but I don't want undesirable effects. the noise is not a problem (the car engine is a tad noisier!). the idea is to provide a smoother input to the field coil of the dynamo to obtain a smoother output.
 
Put a diode in parallel (reversed!) with the field winding, then since you are using XX kHz the field winding will maintain a steady current even though you are feeding it with PWM DC.
 
you think the feild coil will act as a "condenser" ? that was my plan too, I was just worried that the pwm was interfering with the steadiness of the output and damaging components. what if I put a small condenser in paralel with the field coil ?

there is also some noticeable ripple to the output as it is but suspect this is due to a time constant produced by the dynamo, battery and controller reaction times
 
I still suggest to use an electroylic cap.

Start out at a low value of 10µF and increase until it suffices.

Boncuk
 
No it's not a condensor, it's a freewheeling diode. When you PWM a power inductor you put a diode in parallel with it (but reversed), so the inductor current remains flowing in the PWM off period. This maintains the magnetic field in the inductor and makes it very smooth, since your PWM is 20kHz and the inductance of that DC generator field winding is huge the magnetic field in the field winding will be almost free from ripple.
 
I still suggest to use an electroylic cap.

Start out at a low value of 10µF and increase until it suffices.

Boncuk
A capacitor will generate high peak output currents in the PWM output stage which will causing heating, and is not needed. As Mr RB noted, a diode is all you need for a smooth output. At 10-25kHz, the generator field inductance will provide all the filtering you need.

Any ripple you note may very well be due to instablility in your feedback loop. What is its frequency? You may have to adjust the loop response time to eliminate this (are you using a PID loop?). Simply slowing down the loop response time may be sufficent.
 
A capacitor will generate high peak output currents in the PWM output stage which will causing heating, and is not needed. As Mr RB noted, a diode is all you need for a smooth output. At 10-25kHz, the generator field inductance will provide all the filtering you need.

Any ripple you note may very well be due to instablility in your feedback loop. What is its frequency? You may have to adjust the loop response time to eliminate this (are you using a PID loop?). Simply slowing down the loop response time may be sufficent.

if I'm feeding the feild coil directly with PWM then I'll be running at 500 Hz as this is closest to the original 75-150 Hz that comes from the original electromechanical reg.

I'm using a pic running at 8 MHz with a small program as I also have current control and the wish to have control over current and voltage regulation without resorting to trimmers which may move with engine vibration, I guess my "ripple" is due partly to the speed of the program and partly to the time constant of the battery and dynamo.

I'm uncertain about pulsing the feild coil at over 500 Hz as I presume it was originally made for 100 ish Hz and I'm not sure of the effects of a higher frequency (apart from running the control mosfet a bit hotter but its quite cool at the moment anyhow)
 
Running at 500 Hz should be fine.

As I noted, if the ripple is a problem then you may need to add some lag to the loop to slow it down ( I assume you have no compensation in the control loop). You can do that by only allowing it to change the control current a small increment at a time and adding some wait states, if necessary, between increments. You can experimentally determine the right value for this.
 
oh 500 Hz is fine I am uncertain about increasing it.

Basically the main of the program is a looped "if, then, else" statement, a bit of ripple is not a problem being in a car, I am just trying to avoid adding any spikes. the increment goes in a variation of the output by 0.39 % (1 step in a 256 scale - 8 bit)

it could be that the step it increases/decreases the PWM signal by needs increasing to say 0.78 %
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top