Constant average output voltage with PWM - Electronic Circuits Projects Diagrams Free

Futterama · 13th March 2006, 12:52 PM

Hello forum,

I'm looking for a circuit that can give a constant average voltage with PWM regardless of the input voltage.

Example: I need 1.5V from a random voltage source. The 1.5V should be the average of a PWM pulse. Eg. the input voltage is 3V, the PWM duty cycle should be 50% to give an average output voltage of 1.5V.

The load for the 1.5V is resistive.

Any suggestions?

Regards,
Futterama

**Nigel Goodwin** · 13th March 2006, 01:04 PM

It sounds like a normal switchmode regulator?, the filtered output voltage is fed back to control the mark/space ratio. However, your supply voltage is rather low and will make a design more difficult.

What exactly are you trying to do?.

Futterama · 13th March 2006, 01:26 PM

I'm trying to drive a glowplug for a RC engine, from a 2-18V supply using PWM.

chemelec · 13th March 2006, 02:53 PM

You would be better off using a Switching power supply. "Down Converer"

Not PWM.

hantto · 13th March 2006, 03:45 PM

Quote:

Originally Posted by chemelec

You would be better off using a Switching power supply. "Down Converer"

Not PWM.

I disagree. Use a PIC with PWM and a A2D then code a program that adjust to PWM so that the A2D remains at 1.5V. You'll need a low pass filter to feed the A2D. My 0.02€ (but maybe that's a switching power supply anyway :lol

**Nigel Goodwin** · 13th March 2006, 03:56 PM

Quote:

Originally Posted by hantto

Quote:

Originally Posted by chemelec

You would be better off using a Switching power supply. "Down Converer"

Not PWM.

I disagree. Use a PIC with PWM and a A2D then code a program that adjust to PWM so that the A2D remains at 1.5V. You'll need a low pass filter to feed the A2D. My 0.02€ (but maybe that's a switching power supply anyway :lol

Just as I said above, it's a regulated switchmode PSU.

Futterama · 13th March 2006, 07:04 PM

Quote:

Originally Posted by hantto

Use a PIC with PWM and a A2D then code a program that adjust to PWM so that the A2D remains at 1.5V. You'll need a low pass filter to feed the A2D.

A low pass filter like this:

Right?

**Nigel Goodwin** · 13th March 2006, 10:14 PM

Quote:

Originally Posted by Futterama

Quote:

Originally Posted by hantto

Use a PIC with PWM and a A2D then code a program that adjust to PWM so that the A2D remains at 1.5V. You'll need a low pass filter to feed the A2D.

A low pass filter like this:

Right?

Yes, with suitable component values.

Futterama · 13th March 2006, 10:28 PM

Quote:

Originally Posted by Nigel Goodwin

Yes, with suitable component values.

Which is? 100n and 10k? I'm gust guessing...

**Nigel Goodwin** · 13th March 2006, 10:33 PM

Quote:

Originally Posted by Futterama

Quote:

Originally Posted by Nigel Goodwin

Yes, with suitable component values.

Which is? 100n and 10k? I'm gust guessing...

You need to calculate them, if you check the MicroChip website they have an application note about just that!.

philba · 14th March 2006, 06:36 AM

you need to use a fairly high PWM frequency and pick a cut off frequency well below the PWM frequency. the 3db frequency is defined as f = 1/(2*PI*R*C) You want to attenuate the PWM component pretty heavily. Though ripple may not be an issue here.

to get a feeling for this, try SPICEing it.

what you don't say is how much current which has a fair bearing on the matter...

Futterama · 14th March 2006, 06:47 AM

Quote:

Originally Posted by philba

what you don't say is how much current which has a fair bearing on the matter...

I don't know the exact current, but I think it is a few ampere, like 3 or 4A.[/quote]

philba · 14th March 2006, 08:05 AM

well, that's going to take a bit more than just a PIC and a little filtering...

I think I'd look at a buck converter like the lm2852 if its under 2A. Minimal components: 4 caps, 1 coil and the chip.

Futterama · 14th March 2006, 12:10 PM

Well, I got something that works now. I just measure the input voltage with the PIC A2D, and calculate the PWM duty cycle from that, so it will output the average 1.5V square wave to a MOSFET that drives the glowplug.

Thanks for all your replies.

Regards,
Futterama

Oznog · 14th March 2006, 05:55 PM

Your might want to revisit your requirements.
A resistive glow plug should take a PWM signal with no problem whatsoever. You can use a voltage divider to read the source voltage and adjust PWM period based on that.

Code:

Source voltage  PMW period 
1.5v                 100%
  3v                   25%
15v                     1%

The RC filter is very lossy since it burns up all the extra voltage in the resistor. I don't know what a glow plug takes but I'm guessing in the amp range. This may require a beefy resistor and the cap would need a fairly sizable capacitance and high current capacity to filter it into DC effectively.

If you want DC, usually you use an inductor to make a step-down "buck" converter to do this sort of thing. Those change the voltage without generating a lot of heat.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

13th March 2006, 12:52 PM	(permalink)
Futterama	Constant average output voltage with PWM Hello forum, I'm looking for a circuit that can give a constant average voltage with PWM regardless of the input voltage. Example: I need 1.5V from a random voltage source. The 1.5V should be the average of a PWM pulse. Eg. the input voltage is 3V, the PWM duty cycle should be 50% to give an average output voltage of 1.5V. The load for the 1.5V is resistive. Any suggestions? Regards, Futterama

13th March 2006, 01:04 PM	(permalink)
Nigel Goodwin	It sounds like a normal switchmode regulator?, the filtered output voltage is fed back to control the mark/space ratio. However, your supply voltage is rather low and will make a design more difficult. What exactly are you trying to do?. __________________ PIC programmer software, and PIC Tutorials at: http://www.winpicprog.co.uk

13th March 2006, 01:26 PM	(permalink)
Futterama	I'm trying to drive a glowplug for a RC engine, from a 2-18V supply using PWM.

13th March 2006, 02:53 PM	(permalink)
chemelec	You would be better off using a Switching power supply. "Down Converer" Not PWM. __________________ I No Longer accept Private Messages on here. All Emails to me Must Contain the Word \"Electronic\" in the \"Subject Line\" or they go Directly to my Junk Mail Folder. Email me at: chemelec@hotmail.com Website: http://www3.telus.net/chemelec

14th March 2006, 06:36 AM	(permalink)
philba	you need to use a fairly high PWM frequency and pick a cut off frequency well below the PWM frequency. the 3db frequency is defined as f = 1/(2PIR*C) You want to attenuate the PWM component pretty heavily. Though ripple may not be an issue here. to get a feeling for this, try SPICEing it. what you don't say is how much current which has a fair bearing on the matter...

14th March 2006, 08:05 AM	(permalink)
philba	well, that's going to take a bit more than just a PIC and a little filtering... I think I'd look at a buck converter like the lm2852 if its under 2A. Minimal components: 4 caps, 1 coil and the chip.

14th March 2006, 12:10 PM	(permalink)
Futterama	Well, I got something that works now. I just measure the input voltage with the PIC A2D, and calculate the PWM duty cycle from that, so it will output the average 1.5V square wave to a MOSFET that drives the glowplug. Thanks for all your replies. Regards, Futterama

14th March 2006, 05:55 PM	(permalink)
Oznog	Your might want to revisit your requirements. A resistive glow plug should take a PWM signal with no problem whatsoever. You can use a voltage divider to read the source voltage and adjust PWM period based on that. Code: Source voltage PMW period 1.5v 100% 3v 25% 15v 1% The RC filter is very lossy since it burns up all the extra voltage in the resistor. I don't know what a glow plug takes but I'm guessing in the amp range. This may require a beefy resistor and the cap would need a fairly sizable capacitance and high current capacity to filter it into DC effectively. If you want DC, usually you use an inductor to make a step-down "buck" converter to do this sort of thing. Those change the voltage without generating a lot of heat. __________________ I thought what I'd do was I'd pretend I was one of those deaf-mutes.