generating PWM

[frozenpod1]

I have run into difficulty generating PWM pulses.

The circuit I am using is a 555 square wave which is integrated to form the triangle wave. The triangle wave is compared with the sinewave to generate the pulses.

I have stability issues and I am looking for a way to introduce dead time control.

If anyone could assist me I would greatly appreciate it.

 

[Gorilla]

>The circuit I am using is a 555 square wave which is integrated to form the triangle wave.

As far as I understand there are some nicer ways to generate a triangle wave than this. In anycase when using an integrator, depending on the frequency and/or linearity of the triangle wave you require, the high order harmonics can be many times your fundamental frequency requiring a relativly fast (op)amp for your intergrator. Additonaly you may also run into limitations of the 555. Search this forum or the web.

However if using a 555 have a look at the datasheet, there is a method of directly generating PWM output directly using only a 555. I don't know how good this is but if you are using a 555 may be good enough.

>I have stability issues and I am looking for a way to introduce dead time control.
Dead time control is very dependant on the the method you are using to drive you MOSFETs. It would be nice in my opinion if someone started a thread dedicated to the topic of MOSFET gate driving techniques.

 

[classd4sure]

>The circuit I am using is a 555 square wave which is integrated to form the triangle wave.

As far as I understand there are some nicer ways to generate a triangle wave than this. In anycase when using an integrator, depending on the frequency and/or linearity of the triangle wave you require, the high order harmonics can be many times your fundamental frequency requiring a relativly fast (op)amp for your intergrator. Additonaly you may also run into limitations of the 555. Search this forum or the web.

However if using a 555 have a look at the datasheet, there is a method of directly generating PWM output directly using only a 555. I don't know how good this is but if you are using a 555 may be good enough.

>I have stability issues and I am looking for a way to introduce dead time control.
Dead time control is very dependant on the the method you are using to drive you MOSFETs. It would be nice in my opinion if someone started a thread dedicated to the topic of MOSFET gate driving techniques.

Hi Gorilla,

There's been several threads started on that already in this forum, have you seen them?

There's also a few other threads that may contain good discussions regarding gate drive that the topic may not allude to.

One of those might be the class d reference design thread, and another excellent possibility is one called something like "which chip, whatever happened to mueta".

And finally a third possibility is called "Distortions in class d".

All excellent reads anyway, but the more obvious gate drive threads will contain circuit examples and very good discussions.

So now that you've run into the problem of non ideal triangle waves (a non ideal setpoint...hmmm...not good?) and obviously the need to "deal" with inherent circuit delays such as dead time..

There's only one smart way to handle all of that and it has to do with simply making use of a self oscillating design!

It is fun to experiment with a 555 though and is also my first class d amp, didn't sound very good but I got a kick out of it. Mine was a 1Mhz crystal oscillator divided by 2 which clocked the 555 and only had a low side mosfet switch because that's all I knew how to do

Good luck to you both,
Chris

 

[sovadk]

You can build a discrete triangle generator. It works very well.
Follow this link:
Discrete triangle generator
You can also find a triangle generator using opamps or a comperator + a couple of transistors. Follow this link
Triangle generator 2

However the best performance is achived using a selfosccilating design. You can find a lot of threads about it on this forum.