- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

What can we do to test linearity (INL) of 12bit SAR ADC with sinusoidal signal source which has only 10bit DAC ?

Simply put how can i increase linearity performance of 10bit DAC source?

Appriciate your comments

Thanks

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Is there a reason you are restricted to a 10 bit DAC? Why not using one with a higher resolution (even using 12-bit one will give you the problem that the DAC itself might not be linear).

And what is the reason for you to test the linearity of the ADC at all?

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

You might want to read Jim Williams on such a matter: cds.linear.com/docs/en/design-note/dsol11.pdf

Also, you can use two IDACs, one with low current and the other one with high current, and combine their outputs. There is a white paper available: www.cypress.com/

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Use dithered dac component as one possibility -

http://www.cypress.com/?rID=86073

http://www.cypress.com/?rID=47478

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Thank you @hli

Its more of query whether it is possible to improve linearity of DAC by any simple means.

If we wanted to use 10bit source and improve its SNR then we could just use a bandpass filter followed by buffer ryt ?

In a similar way is there anything we could do to improve linearity of source, if yes please explain briefly.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

The following might assist -

Improve non linearity http://www.cypress.com/?docID=32344

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.48.7296&rep=rep1&type=pdf

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Yopur original question was about resolution, not about linearity of the DAC. So now I'm confused...

If you want to improve the resolution, look at the documents Dana I gave you. They show multiple ways to increase the resolution of the PSoC IDACs up to 12 bit. If you need better resolution, you can a PWM as DAC as PSoC73 suggested.

If you want to improve the DACs linearity, ask Dana, she knows her stuff 🙂 (Much better than I do, I should add...)

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

It might be useful if you could shed some light on -

1) Target specs needed, DC, AC, temp, noise......

2) Application, what is the DAC output being used for

A filter and buffer can improve noise performance, or destroy it, it all

depends on specs you are trying to achieve and how you implement

the additional signal path components. As you approach higher perfformance

very subtle effects cause increasing problems, like the crossover distortion

in the input of R-R OpAmps the industry does not like to talk about.....

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

thank you.

@hli my question was how can i improve the linearity of 10 bit DAC such that it is equivalent to a 12bit DAC source.

@dana application my source is limited and i would like add some **external circuitry **to improve its linearity so that i can test ADC with that. So the question is what external circuitary should i add to serve this purpose ?

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Your first question, dithering, per link posted in prior part of thread.

One approach to SAR testing, use 16 bit PWM with integrator ( R-C network ), DelSig to read PWM

output of R-C, and feed R-C to SAR. Since DelSig good to 20 bits, that in turn insures you know what

the V source (PWM) is doing, and measure against SAR output. I would use averaging on readings to eliminate

un-correlated noise.

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Scratch the SAR testing comment in prior post, forgot there is no DelSig

in PSOC 4.

If you used 16 bit PWM approach, and supplied a precision reference for Vdd on the

part, then PWM would in effect become a precision DAC. Note this is only if you need

absolute accuracy. If all you need is relative no reference is necessary. The PWM is

inherently montonic, only its output integrator, if you used an OpAmp integrator vs the

simple R-C network, would affect montonicity.

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Thank you dana

can you provide any sample project file in which PWM is used a DAC.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Did you look at the first two links Dana gave you? The contain an AppNote about increasing the resolution of the PSoC DACs, and even a complete dithered VDAC (which I didn't knew existed as finished component - thanks for pointing me to it, Dana!)

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

"yields a DC outoput directly proportional to DC."

That might be confusing in prior post, should be

"yields a DC output directly proportional to Duty Cycle."

Regards, Dana.

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Thank you dana.