Octal eight-bit DAC for 80 cents

The settling time and and integral nonlinearity specs are very Meh;

I think the TI quad DAC08 variants with SPI like D084S085 are pretty nice for a buck

Well, it's to replace a bunch of PWMs. ;)

Those are nicer, for sure, but twice as much money.

Cheers

Phil Hobbs

BH2221 12-channels.

Have you used it for anything?

Thanks

Phil

-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510

I use the ATTiny85 as a "general purpose low-frequency thing building block" that's under a dollar, I got this software suite that lets you drag-and-drop functional blocks like clocks, counters, PWM outputs, etc. "wire" them up into a system, assign port numbers, push a button and it generates C++, compiles it and uploads it to the uP for testing in under

The Rohm chip might make a good team with the ATTiny, the main flaw in a lot of applications is it only has three hardware PWM outputs on fixed pins, one of which ties up an ADC channel if you use it.

Stepping up to a core with any more than three hardware PWM channels is a significant increase in footprint and price

Not that one, but other Rohm stuff, I'm happy to believe their specs.

I'm getting down on using PWM outputs, DACs are so cheap.

There is a crapload of microes that has fast 12bit DACs

PWM has slow response, if you need 12 bit accuracy. And injects noise from the supply

Cheers

Klaus

The ROHM devices are R/2R ladder based, so they too will pass noise from the supply (no internal reference, or reference pin).

... if you get the twelve-output one, and only need nine, could it help if one were to put filter capacitor(s) on the other outputs, and set 'em all high? That would put a decoupling capacitor on the chip side of any Vcc wiring drop.

Heck, you could use an AC-coupled amp to do an active filter...

Impressive! Why the 2 packages? The big one is rather wasteful of space..