Cheap 12 bit DAC and ADC's

For the ADC, maybe an ADS7866 and an HC4051?

John

Define "cheap". Also, none of them truly goes to the rail if that's what you want. Many are 0V to 4.095V because that gives you a nice 1mV granularity. Others go by whatever external ref you provide.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Depends on your requirements. There are huge price differences depending on speed. And there are different interfaces, like SPI or parallel and you might not like BGA style parts.

A good source for selecting the right part is the parametric search at Digikey, even if you don't buy it there:

ADC:

DAC:

Not cheap, but nice and fast is the LTC2624, which is on the Spartan3E evaluation kit, with which I've implemented my signal generator:

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de

I like the AD5328 and the AD7928 but they aren't exactly cheap.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

I should have mentioned SPI, But I think the AD7927 is the one, half the cost of the LT part we are using now. The are some Dacs that are priced right from analog too.

If I had extra port pins then the Mux I would play around with, but thats not an option.

Thanks

That falls under "cheap"? Hmm, you guys leave too much money on the table :-)

At that cost I'd consider muxing the SPI stream so you can hang a register onto the SPI that rotates through the ADC channels. Not an option for the DACs though.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

There are some cheap high quality audio DACs, but I don't know if they can go down to DC.

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de

They used to, but at least the ADC part of audio codecs often has inherent highpass filtering in there which can't be turned off. Learned it the hard way when I wanted to hack a laptop into going down to below

1Hz. But it did go down to about 5Hz.

Also you usually get only two channels. Example of a really cheap audio DAC, for 3.3V supply though:

They don't even spec the offset of the opamp that does the current to voltage conversion. It could very well be above the 12bit granularity because it doesn't matter to the audio guys.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Sometimes it matters. If you have a sample rate change and you have to disable the output before changing the clock for the DAC to avoid any clicking noise, it could be a problem. I know a schematic with lots of transistors, op-amps and a relay, for which I've designed and implemented a software concept (including FPGA functions), which is really necessary to avoid any clicking noise for professional audio monitoring systems, where you can get deaf at signal levels of 0 dBu. I don't care, but looks like radio broadcasting customers do :-)

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de

True, don't ever touch the clock rate or data stream, then it is indeed a problem. Curing that at the output is like a squelch and it seems that even manufacturers of professional communications gear no longer fully understand how to handle this. As evidenced by the popping squelch and the schematic a guy at the German NG posted a while ago (Michael, the one with the large model railroad). That was a truly horrid circuitry.

BTW, after we had our "glorious" switch to DTV (don't even get me started on this #@^&!! ...) we now "enjoy" the occasional tsk, pop, pock and phsst sound when video content such as news reporter pieces changes. Oh, and the movie we wanted to watch yesterday pixelated out into a blue screen, as usual.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Well I could add like a 74HC595 and feed the mux with that. I think I have 1 or 2 pare port pins.

Also For the DAC I found a microchip part MCP4922, $3.

Cheers

Microchip? hmm.. For the ADC maybe a dsPic with simple program to output data via SPI, $3 Some devices have inbuild 16-bit DAC also (like dsPIC33FJ64GP802)

Well, there's the solution. All you need is to disassert the device select for the DACs/ADCs and run another CS line to the 595, then use the regular (same) SPI lines for clock and data.

But that's only a dual DAC. Plenty of those.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Do you have a few favorites?

Not really, each circuit situation is just too different and sometimes tricks need to be played with the DACs. I like the AD5328 and the AD7928 for ADC when I have to have lots of channels (if I need more than one channel it's usually lots of them). For a 2ch DAC the DAC7612 is kind of ok at $3, although a long hard search should turn up less expensive ones. Whenever I have a project I go by price and most of all widespread availability. Certain parts are on the black list, but we all know who's those are.

Most of the time my clients already have DACs or ADCs in their designs and prefer that I use the same.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Most of the audio DACs go down to DC, and can be underclocked, but:

1) DC offsets and drifts are poor indeed and not specified in the datasheets. 2) Strong 1/F noise at subaudible frequencies, and strong noise shaping residue in the ultrasound. 3) The audio DACs have the synchronous interface (I2S or TDM) which has to be updated continuously. I.e. they can't hold the value. 4) High current drain.

5) The internal voltage reference is unstable and inaccurate.

6) The output is not R-R and centered around the reference voltage.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant