TM4C1294NCPDT: DAC Availability.

Hi,

Tiva series does not have on-chip DAC. You must externally add such device,  example DAC7513, 12 bits, SPI interface. However, you should specify/search for  appropriate device for your project.

Equivalent devices from competitors often do have one or two DACs. However, they are usually run-off-the-mill types, 10 to 12-bits with settling times of several microseconds. That is good for frequencies below 100kHz, i.e. the audio range. And, you don't gain much with integration, both in regard to bandwidth and noise immunity.

As Petrei said, it much depends on your application requirements.

Hello Gaurav,

As Petrei and Bruno both highlighted, it depends on what is the application you intend to target, that will decide what type of DAC is needed.

f. m. said:

10 to 12-bits with settling times of several microseconds

While in general that's true - we find that 100KHz rate is well exceeded when, "Less than full-scale" DAC level changes are commanded.   (admittedly - ours is a minimal sample set (maybe 3 devices w/internal DAC).

SPI & I2C DACs (especially I2C) appear to limit bandwidth due to the delay imposed by the serial interface - is not that so?    Internal DAC avoids (most) of such delay.

While the (always) "Unguided Posters" will never reveal their application - the use of even a relatively simple MCU based DAC - proves of great value in automated test/verify - especially when the DAC's output is selectively "steered" via analog multiplexer.    (and the fact that a DAC appears on a, "sub 1 (USD) ARM M0" highlights the "missing man" from vendor's brigade...)

cb1_mobile said:

While in general that's true - we find that 100KHz rate is well exceeded when, "Less than full-scale" DAC level changes are commanded.   (admittedly - ours is a minimal sample set (maybe 3 devices w/internal DAC).

I didn't try to reduce resolution even further - 12-bit is not really excessive. I used this DACs mostly for audio applications, where the settling times are appropriate.

cb1_mobile said:

SPI & I2C DACs (especially I2C) appear to limit bandwidth due to the delay imposed by the serial interface - is not that so?    Internal DAC avoids (most) of such delay.

For a 100kHz update rate, the SPI bus is not quite the limiting factor. Feeding it via DMA wouldn't even stress the core too much.

But for most "average" applications, an internal DAC would suffice - and has a distinct advantage via price and pin usage...

School exercise:
- Link 8 GPIO's from the MCU into a dividing bridge of strategically chosen resistors.
- Pick all these pins on the same port base
- Use DMA to transfer the desired value at once to all the pins (or you can also do it with the simpler GPIOPinWrite and a mask)
- Use a OpAmp on the exist of your diving bridge to have "a more powerful analog signal"
Voila, you have an 8-bit DAC that is very fast, based on a TM4C... (at the expense of several pins and the cost of "precision resistors")
Bruno

But often too expensive for commercial endeavors. You get serial DACs for less than the cost of the opamp + chicken food (SMD R+C components) - and the GPIOs used up. And to use DMA, you need to limit your GPIO selection even further.

But a good exercise, as you said.