This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS1113: Continuous Mode

Part Number: ADS1113
Other Parts Discussed in Thread: ADS1114, ADS7142, ADS1191

Hello,

I am considering the ADS1113 ADC for an ECG application. "Continuous mode", I believe, is the best fit. However, how is it insured that the Host uP reads a 'fresh' conversion each time it does an I2C read? I suppose the software driver could do something such as anytime the data byte read is the same as the previous, you toss it out. But that's sort of a "hack" to me. 

In general, is it expected that when using an ADC such as the ADS1113, it is the only slave on the I2C bus? I see that the ADS1114 DOES have an ALERT output to tell the Host when the conversion is done, but that won't work in my particular application, because I don't have a dedicated IRQ/ALERT line back to the Host uP.

Thanks for any feedback!

Regards,

Mark

  • Hi Mark, 

    The caveat when using continuous conversion mode is that you can read a previous conversion and may not always be 'fresh'. However, using single-shot mode, you can. In single shot mode, you tell the device when you want to perform a conversion and you can poll bit-15 of the configuration register to identify when a conversion is done. The ADS1113 will always act as a target device. 

    Is using single-shot mode something you would be capable of doing?

    Regards,

    Aaron Estrada

  • Hi Aaron,

    Thanks for your reply. When doing an application like ECG, single-shot mode isn't really an option. We want periodic samples because then we can remove spectral content via filtering. If the samples are done via single-shot, then the time delta between them isn't known, so you can't discern the spectral content of the signal. Imagine the difference between the two cases: 1) sample 100Hz sine wave at 200Hz rate and do FFT on the results [you'd get 100Hz peak and then all the alias frequencies; and 2) sample 100Hz sine wave randomly and do FFT (it would be a mess).

    Regards,

    Mark

  • Hi Mark,

    Understood! With continuous conversion mode, using the Alert pin would be the best way to know when a conversion is finished. I suppose you can read the conversion data at a slower rate than the data rate to avoid reading the same data twice. However, there is a +/- 10% variation on the data rate so that would need to be taken into account. 

    I know we do offer some other devices that would be more suitable for an ECG application so if you are interested, I can give you some part numbers. Also, if you don't mind, what drew you to the ADS1113 for this type of application?

    Regards,
    Aaron Estrada

  • Hi Aaron,

    Thanks for the feedback. The ADS111x was attractive because of the following: 1) in stock!!;)   2) small size; 3) I can try three parts ('13, '14, '15) all in the same footprint; 4) has continuous mode w/o needing to have SCL continuous (ADS7142 uses this method). But if you think there is a more applicable PN for ECG, please pass on those part numbers. Thanks!

    Regards,

    Mark

  • Hi Mark, 

    Understood! Well some of the devices I found may be a bit overkill if you are looking at the ADS111x family but I would still like to give you options Slight smile We have the ADS119x and ADS129x families of ADC's for biopotential measurements. If you only need 1 channel, then I would look into the ADS1191 which is a 16-bit, 1CH device. 

    If these are overkill then I think your best bet would probably be going with the ADS1114 due to the integrated PGA that allows for a wider full scale range. Regarding the continuous conversion mode, it will ultimately be best to utilize the Alert pin. Are you sure that your uP cannot spare a GPIO pin for the Alert signal? You can also read the conversion results at a slower data rate but after some thinking, that may cause you to read repeated data. 

    Regards,
    Aaron

  • Hi Aaron,

    Whoops, I should have also added "I2C interface" to my list-of-needs. We're looking to cable this a fair distance and we're pin-limited, so we chose I2C. Are all the "biopotential" targeted parts SPI interface?

    I did look at the I2C driver, and I guess basically the driver inserts delays to gate the reads for the samples (if I understand it correctly -- and maybe I don't). So that would make the samples "periodic", presumably with some +/-% in the period (we're not talkin' realtime SW here). 

    Regarding using the Alert, I may be able to cobble it into my existing hardware with a rework. It's just that my baseboard is already designed so, preferably, I'd like to live without needing it. But it's possible I could rework it.

    Regards,
    Mark

  • Hi Mark, 

    To my knowledge, all of the biopotential parts are SPI. I can still look around but ultimately, I think a workaround on the ADS1113 or ADS1114 will need to be done. Would be much easier with the Alert pin but I understand your constraints. 

    Regards,
    Aaron Estrada

  • Thanks again, Aaron. I'm going to go with the ADS1114 so I can hopefully get the ALERT working. I think for our low sampling rate, we'll be OK even without it. 

    Regards,

    Mark