Need Higher Frequency Resolution than PWM module on PIC

[savgrace]

Hey guys.... Final Post to this thread Sorry to drag this on... I just wanted to post the solution I thought up. It is relatively simple and won't require me forking over cash for a DDS. Since 1024 step digital pots were a little spendy and still don't give me the 44900 steps (100hz to 45khz) I would need for my frequency range, i thought up the idea of combining two 256 step digital pots.

With a 50k digital pot /256 = about 200ohm increments
With a 200 ohm pot / 256 = will divide up that 200 ohm increments of the 50k pot

So combining the two effectively gives me a 16-bit digital potentiometer (256 x 256).

I'll use this to control the frequency and mark/space on a simple 555 PWM circuit.

 

[MrAl]

Hi,


Hey that's an interesting idea there. That way you can control the 555 frequency at least to some extent and set it with the PIC and not use up any on board peripheral on the PIC too.

I'd like to hear more about how your double pot works for this application once you get it built.

 

[carbonzit]

So you're basically planning on having a "coarse" and "fine" adjustment.

Wonder how much programming it would take to combine them into one super-control with the resolution you want.

 

[alec_t]

A nifty pragmatic solution. That will, of course, give a theoretical 1Hz resolution. As has been pointed out above, the frequency instability of a 555 is likely to be > 1Hz.

 

[geraldfryjr]

Yes,That was an idea that I had in the back of my head and would work good on a TL494.
Although, I can't remember as to how you would implement it as a manual pot replacement in the Mark/Space of a 555 circuit.

If recall there was a circuit that did this published in EDN recently and I think it took three of them to do it.

Just remember that if you change or don't keep the overall resistance of the Mark/space adjustment the same it will change the frequency of the 555 timer.

I did alot of messing around with it and it looks easy but the timer does perform very well when you get up high in the frequency range although it will work,sort of.

I was trying to use it in a simple switching supply but my frequency of operation was to high.

And I also tried to implement it in a PWM for a Class d amp and I couldn't get a stable wide range with it without it changing frequency.

The Dutycycle worked but the frequency would change I even tried the two timer configuration but the performance didn't meet my expectations.
So, I ended up using the traditional method and it worked flawlessly.

Maybe you will have better luck !!!

I have also found that of all of the timers that I tried not all of them would work in the range that I needed although the data sheet implied that they would.
I was trying to make them work well above 100khz, though.

It was in EDN that I saw that circuit and I will look for it and post it here when I find it again.

jer