What to look for in a DAC

Sep 15, 2006 15 Replies

Hi all,



Can anyone please tell me to choose a DAC, what parametersI should look at in its datasheet? What is the relation between the SNR and the resolution of the DAC(number of bits)? Is there any formula?


Thanks a lot, fEIT



The resolution is 2^N with N being the number of bits. DACs usually are monotoneous, and the manufacturers claim accuracies in the order of one bit. So with a supply and a reference that is stable and noise less, the SNR would be 1/(2^N), or the log of it.

Further parameters are the supply voltage, the internal or external reference, voltage out or current out, and whether the interface is parallel or serial. Then come speed, how many of them are in a case, price...

Rene

Ing.Buero R.Tschaggelar - http://www.ibrtses.com & commercial newsgroups - http://www.talkto.net

Are you a student by any chance ?

Graham

Thank you very much for your respond Rene.It was very helpful

fEIT

I was going to suggest that a lecturer should have explained all this but PhDs are different.

The parameters of most interest will depend on the application which you haven't yet mentioned.

The SNR will be approx (6xN)-3 dB for Nbits of resolution.

Graham

haven't

Thanks for your respond. I am trying to generate sine waves in range of 10kHz-100kHz with 80dB SNR. I generate the digital sine wave using DDS technique using FPGA.The sampling rate of my digital sine wave is 1.5MHz. The data of the sine-wave is generated comes from the FPGA in parallel.

Mandana

In message , dated Sat, 16 Sep

2006, Eeyore writes

Shouldn't they either know already or be able to use a library?

OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk There are benefits from being irrational - just ask the square root of 2. John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

In message , dated Sat, 16 Sep 2006, fEIT writes

Well 16 bits is enough, and is an industry standard. I would advise you not to try to use fewer.

Your biggest problem will be actually building the device, because 80 dB signal-to-noise ratio requires constructional techniques and skills that you almost certainly have never been exposed to.

OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk There are benefits from being irrational - just ask the square root of 2. John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

PhDs

haven't

This is 'almost' an audio application.

The resolution / SNR is trivially simple to acheive with audio DACs. Why do you use

1.5MHz sampling with Fmax of 100kHz ?

Some high-end audio gear now supports 192kHz sampling which is almost good enough for this application. If you have several converters 'in parallel' and provide them with 'staggered' samples you could do this quite readily.

Graham

Damn good point. Maybe *we* are the 'library' ?

Graham

Nah. Much too noisy in here. ;-)

Cheers! Rich

Yes, and you and the donkey are prime contributors. To the noise.

Lessee, 6 words and a smiley, vs. two sentences, comprising

12 words, and _no_ smiley - I'd say that's about 3 db.

Can you say, "Pot...Kettle...Black"?

Cheers! Rich

Shouldn't we send him to England, so they can co-habitate?

Service to my country? Been there, Done that, and I\'ve got my DD214 to prove it. Member of DAV #85. Michael A. Terrell Central Florida

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required