AVR Controlled LM317

[Richard Head]

I have recently used a 10-bit DAC instead of a PWM. It works like a dream with no calibration required. Unfortunately the DAC is damn expensive. Luckily this application can afford it.
I wonder if they"ll ever have onboard DAC's in microcontrollers? I love using them, so precise.

[dannyf]

I wonder if they"ll ever have onboard DAC's in microcontrollers?

Wonder no more - tons of them out there.

[Richard Head]

dannyf
Now you've got me looking at the TI micros with onboard DAC's. I never knew!

[dannyf]

Not just TI, other folks too: Freescale is top notch when it comes mixed signal mcus (Silabs great on ADC). STM32F3 is another option. Even some of the lowly PICs now have DAC (or Vref generator with an output pin).

[miguelvp]

And Cypress, don't forget about them, even with programmable analog blocks.

[Lizerd]

Almost all mid-high end STM32 has 12bit DAC.
Even quite a lot of the low-ends have 12bit DAC

update on my lab psu, read more at www.lizerd.se/?p=893

[themacman33]

After a brief break, I have decided to just use a lm2576-adj with i2c from an AVR to control a SMPS. I find that it will be easier to implement, lower noise, and more accurate.
Lizerd, that looks like a really neat project, I'll take a look at it!

[dannyf]

update on my lab psu

Nicely done.

I have decided to just use a lm2576-adj with i2c from an AVR to control a SMPS.

The same principle applies to a linear regulator-based power supply. Switchers do have the advantage of being more efficient, making them more compact.