• State TI Thinks Resolved
  • Locked Locked
  • Replies 10 replies
  • Answers 3 answers
  • Subscribers 47 subscribers
  • Views 852 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

ADS1258 mux input leakage current

ADS1258-EP: Leakage Current in MUX

Aidan Phillips
Prodigy 140 points

Part Number: ADS1258-EP
Other Parts Discussed in Thread: ADS1258, ADS1220, ADS124S08

Hello,

I am not able to find a spec on leakage current from the analog MUX in the ADS1258-EP.  I will have high impedance input buffers on the output of the MUXOUT pins going to the ADCIN pins, however I'll be placing RTDs directly on the AinXP and AinXN input pins (with some filtering).  If ON channel leakage flows back into my RTD sensor, I'll have an offset that may have a high temperature influence.  Has this leakage ever been characterized?  I saw one similar forum post from many years ago that was inconclusive.

Aidan

  • 0
    TI__Guru 58981 points

    Hi Aidan,

    As the post you linked to implied, we do not have characterization data on the mux leakage for the ADS1258.

    However, let me look into your concerns a bit more and get back to you tomorrow with some comments about mux leakage current in general.

    -Bryan

  • 0 in reply to Aidan Phillips
    TI__Guru 58981 points

    Hi Aidan,

    I had tried to take some measurements on my EVM today, however I was having technical difficulties getting the setup to work unfortunately. Also, I did not have a buffer between the mux and ADC, so it would have been challenging to get the data you requested. It should be possible for you to take the same data however, by measuring the leakage current using a precision DMM e.g. Keysight 3458A when applying an input signal to the ADC. This should give you some idea of what this value is.

    I also spoke with our design team about this request, and while we don't have data to provide you, they estimated that the ADS1258 mux leakage current should be on the order of 10s of nA at room temp assuming the buffer provides sufficient "isolation" between the input and modulator. Since this information is not in the datasheet, it is not guaranteed and you use it at your own risk.

    If you are using a 1mA current to excite your RTD, 10nA of leakage current is approximately 0.001% error. This is likely negligible given the error of the reference resistor and the ADC itself.

    You are correct that this leakage will get worse at higher temp, though I do not have any data - anecdotal or otherwise - to tell you how much this might be.

    I hope this info helps.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 140 points

    Hi Bryan thanks for taking the time today.

    Unfortunately, our RTD excitation is in the 10s of uA, so we have to consider this as an error source.

    Would you be able to point to a datasheet with a similar MUX structure that does detail how much worse the leakage current gets worse over temperature?  If the leakage doubles every 10C then we're up to 640nA at 80C for leakage.

    Aidan

  • 0 in reply to Aidan Phillips
    Prodigy 140 points

    Hi Bryan,

    I just had a thought too - If you made the input to the ADCIN pins open circuit, but put one channel in 'Fixed channel mode' to always have it ON, and then measured the leakage from any channel to ground or through a representative 20kOhm resistor, would that be low effort?

  • 0 in reply to Aidan Phillips
    TI__Guru 58981 points

    Hi Aidan,

    For my own curiosity, what RTD are you using in this application? The largest resistance value from an RTD that I am aware of is ~ 4kohm from a Pt1000 at max temp (850C). So even 500uA of bias current could be used in this case, depending on the compliance voltage of the current source and assuming you are using a similarly-sized reference resistor (~4kohm). Are you measuring multiple RTDs in series then? Or are you measuring thermistors?

    The ADS1220 is an ADC that uses a similar process technology as the ADS1258 and also has a mux. That is about the closest I can offer you in terms of similar devices. However, as with the previous information that I provided, you make and use any assumptions at your own risk.

    I am trying to get my system up and running in the next day or so, I will try to measure the leakage current with the mux disconnected from the ADC inputs. You can try the test you mentioned, though it would make more sense to just measure the current using a precision DMM. If the current is very small, you will be limited by the resistor tolerance and the capabilities of the measurement equipment.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 140 points

    We are in an extreme temperature situation where excitation current >100uA will cause too much self-heating in our RTD element.

    The ADS1220 datasheet does not show a graph or estimated value of leakage current some I'm not sure it's of much use to know that the technology process is similar.

    The test I mentioned would still measure current with a DMM.  You would leave the ADCIN pins disconnected, and then add resistors to ground on the AIN inputs and measure the current with a precision DMM.

  • 0 in reply to Aidan Phillips
    TI__Guru 58981 points

    Hi Aidan,

    I was able to get a different ADS1258 setup working yesterday so I was able to measure the mux leakage current by turning on one of the channels (in this case AIN14), turning the MUX bypass off, letting the input float and connecting my 3458A in series with the MUXOUTP pin and ground.

    The current I measured was in the 10s to 100s of pA. This result is why I was hesitant to suggest that your experimental setup might work, because this very-low current would probably just look like noise if you were trying to measure it across a 20kohm resistor. You would need very high resolution measurement equipment to be able to detect it which, if you already had this equipment, you might as well just use the setup that I did. This is how we would test such a parameter in production.

    Some caveats to these results: this test was at room temp in a controlled environment and did not account for any bias currents from the buffer that you would use. If you want to use the ADS1258 with a buffer and are very concerned about these leakage currents, I would suggest performing the same tests I did with your whole system biased so you can measure the cumulative effect of all components. For this test you would need to measure at the ADC input however, since I assume the MUXOUTP has a hard-connection to the buffer input. I would also suggest measuring the leakage current over the full input range using a low noise, low impedance source.

    I am not sure how far along you are in your design, but a last resort might be switching to a different ADC that does characterize the ADC input currents such as the ADS124S08. This device was designed specifically to have low input currents, which are specified over temperature. This ADC also has integrated current sources for RTD biasing.

    Let me know if you have any additional questions.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 140 points

    Sweet, thanks Bryan.  I was never suggesting to use a 20kOhm resistor as a sense resistor.  I was saying that it would be a more representative resistance for the RTD resistor and reference resistor i have in series.  Having this impedance might further reduce leakage current, but connecting the input straight to ground seems like the bounding case.

    It would be very interesting if you applied a hot air gun to the chip and got a few rough data points.  You could even hold a thermocouple to the IC to correlate with temperature or watch the chip with a FLIR infrared camera.

    If the current doubles every 10C from 100pA then going from 20C to 80C would move the leakage to 6.4nA, which may be acceptable.

    Would you be able to run the quick thermal test?

  • 0 in reply to Aidan Phillips
    TI__Guru 58981 points

    Hi Aidan,

    I do not have the time or ability to take additional measurements at this time. In fact, troubleshooting the existing board and getting the new board up and running took considerably more time than I anticipated, which halted work on the other projects that I already had in progress.

    Moreover, if this error source is as detrimental to the system performance as you suggest, I strongly believe that you need to test this parameter in your system to get a valid understanding of how it impacts the actual measurements. Your additional components and the input voltage will impact the total input current you see at the pin (not to mention the temperature), so just having the mux leakage current from a floating input with no buffer will not necessarily provide the information you need.

    I hope this is understandable.

    -Bryan