LPC824 ADC datasheet parameters

[tyassin]

Hi

Looking at the ADC section in the datasheet for LPC824, there are these parameters listed.

I have tried to search on the net and read several application notes. I sought of understand what the different terms mean. But what I lack is to understand what to expect of precision on the output of of ADC.

Some microcontrollers like the LPC1769 have an additional parameter

Et (absolute error) which I understand is the maximum non-calibrated error. This is can be several of the other errors giving this error. Also that the absolute error is not just the sum of the other errors.

Does it mean that for a 12-bit ADC with 3.3V reference, the maximum output error is 4LSB which is 3.3/4096*4=3.22mV?

My question is what these two tables give of data to predict the actual output precision of the ADC?

Kind regards

Thomas 

[tyassin]

Hi Diego,

Thank you very much for the reply.

I have gone through your links and they are very helpfull.

Reading the notes there seem to be an error given by Et and then some sort of degradation in resolution given by the INL, if for example you have 2LSB of INL that will degrade a 12bit ADC to only 11bit of resolution.

I understand the Et parameter for LPC1769, but how do you get something similar for LPC824? Do we have to calculate the Et by yourself with-root-mean-square?

Also I found this link: LPC ADC gain error vs absolute error spec where something similar to my question. Here is also refered to Et, but also adding the gain error Eg? 

Thank you for the help and clarification.

Regards

Thomas

[diego_charles]

Hi Thomas, 

I'm glad to help. 

Reading the notes there seem to be an error given by Et and then some sort of degradation in resolution given by the INL, if for example you have 2LSB of INL that will degrade a 12bit ADC to only 11bit of resolution.

I´ll  need additional time to corroborate this. By the moment, could you let me know in which specific note you obtain this?

I understand the Et parameter for LPC1769, but how do you get something similar for LPC824? Do we have to calculate the Et by yourself with-root-mean-square?

Also, I will check if is possible to provide this information. Or if there is a valid method to obtain this data.  

Also I found this link: LPC ADC gain error vs absolute error spec where something similar to my question. Here is also referred to Et, but also adding the gain error Eg? 

In the thread, the user is trying to find the relationship among the Tue(total unadjusted error) parameter and Et (absolute error) . Both make reference to errors related to unadjusted ADCs. It is mentioned that you can add up Et and Eg( percentage of difference of ideal conversion  and actual  without offset ) to obtain Tue, unfortunately I don't have a reference document to prove this.  

[tyassin]

Hi Diego,

I´ll  need additional time to corroborate this. By the moment, could you let me know in which specific note you obtain this?

I found this app note where they discuss this. Look under "Systems Requirements": https://www.maximintegrated.com/en/design/technical-documents/tutorials/7/748.html 

Best regards

Thomas

[diego_charles]

Hi @Thomas Sorensen,

I hope you are doing well, sorry for the delay.

I can point out the following general definition of INL (integral non linearity error )

INL is a measure of how closely the ADC output matches its ideal response. 
INL can be defined as the deviation in LSB of the actual transfer function of the ADC
 from the ideal transfer curve.

In the LPC824 the INL error   is present even after correcting the gain and offset errors , in  other words after performing a calibration. Therefore I can say that in a  calibrated 12 bit ADC, the INL parameter specifies that the ADC actual output can be +-3 LSB deviated from the ideal output.

The absolute error  Et corresponds to the worst case scenario for the ADC error, with no calibration .

Theoretically,  you can obtain the absolute error by doing the  root  of the  squared errors of the ADC. Unfortunately I'm not able to provide you that parameter for the LPC824.

Doing research I´ve found another interesting application note that discusses this errors and  ADC calibration techniques

https://www.nxp.com/docs/en/application-note/AN2989.pdf 

Although is for the MPC5500 family, some ADC concepts are still applicable.

I hope this can help you further.

Best regards, Diego.

[tyassin]

Hi Diego,

Yes I think I sort get the point now with these parameters.

Thnk you for the support.

regards Thomas

[diego_charles]

Hi tyassin@live.com‌

Thank you for your reply. 

I hope being helpful.

Best regards, Diego

[diego_charles]

Hi, tyassin@live.com‌

The tables can help you to predict the maximum error  on the ADC output data from the ideal conversion, in other words the errors from the ADC itself.

In the  applications we will need to take into account other factors that can generate external errors, for example the PCB, external noise, quantification noise, input signal noise. 

I assume that you may check this documentation before, but I will attach it just for reference. 

 Understanding ADC specifications  

Common ADC Terminology

How to Increase the Analog-to-Digital

 For the LPC1769, The Et parameter with 4 means that the conversion can at a maximum of 4 bits deviated from ideal conversion.

 Please, let me know if this is helpful for you!

Have a great day,
Diego

-------------------------------------------------------------------------------
Note:
- If this post answers your question, please click the "Mark Correct" button. Thank you!

- We are following threads for 7 weeks after the last post, later replies are ignored
Please open a new thread and refer to the closed one, if you have a related question at a later point in time.
-------------------------------------------------------------------------------