• State TI Thinks Resolved
  • Locked Locked
  • Replies 18 replies
  • Answers 5 answers
  • Subscribers 47 subscribers
  • Views 472 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

ADS1262: ADS1262 negative power short to ground and couldn't automatically recovered

Irene Qin
Intellectual 2000 points
Part Number: ADS1262

Hi team

This is FAE supporting MT, Could you please help them about below question?

They use the positive and negative power supply scheme (as shown in the figure below, AVDD connected to +2.5V, AVSS to -2.5V), and then when the system is instantly connected to a large load, it will cause a short drop of +2.5V and -2.5V, and then the ADS1262 chip will not work properly.

Further verification, confirm that the problem is -2.5V, that is to say, artificial short circuit +2.5V power supply, the ADS1262 can resume work immediately after the short circuit is lifted, but when the -2.5V short circuit, the chip can not work normally. At this time, we can artificially reset the ADS1262 (20 feet down) and cannot return to normal.
In fact, in the case of normal +2.5V and -2.5V power supply, the ADS1262 is artificially reset (20 feet pulled down), and then the chip can not work normally, the phenomenon is exactly the same as the -2.5V short circuit (the DOUT pin output period will become 20Hz, as shown in the figure below).

Some other AD chips have been verified, but they are all powered by a single voltage (AVSS is connected to GND), and no matter how the AVDD is short-circuited, it can automatically resume normal operation immediately after the short circuit is lifted.

The requirement for the project was that even if the AD power supply was briefly dropped or shorted, the AD would work properly without the need to reboot the system after release (previous products have been able to meet this).

The question is: why does the AD operation abnormality occur once the negative supply drops or short circuits and cannot be automatically recovered, while the positive supply drops or short circuits do not have such a problem? Why does artificial reset not only not solve the problem, but also cause the same abnormality? What are the solutions or circumventions to this problem?

I've already loop MT's engineer Chen in this discussion, due to I will be in annual leave in the next week and will limit to check emails.

Thanks for your support in advance.

Irene

  • 0
    TI__Mastermind 42871 points

    Hi Irene Qin,

    Some questions so we can help diagnose this issue:

    1. Can you provide the power supply schematic so we can see how the +/-2.5V supply is generated?
    2. What is the "large load" (signal magnitude, etc.)
    3. How is the load connected instantaneously to the ADC inputs? Is there a switch between the load and the ADC?
    4. What is a "short drop of +2.5V and -2.5V"? Can you explain how much the supplies droop and for how long? A scope shot would be helpful

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Dear Bryan,

    1. For the power supply schematic, Irene will forward our AD option board SCH to you when she gets back to office. It is not allowed to paste it here.

    2. The "large load" means heavy load.  +2.5V and -2.5V are generated by +5V and -5V through two LDOs, +/-5V are the excitation voltage of the load cells. The number of the load cell can be 1 to 16, so the heavy load means many load cells are connected to the +/-5V，such as 16 units. It should be noted that 16 load cells are connected at the same time, so a large transient current（larger than 10A， see the below picture) will be generated at the moment of connection.

      

    3. there is no a switch between the load cells and the AD option board,  the load cell cable connected to the AD option board by a plug（J5 on page 4 of the SCH）.  Plug in will lead to the abnormal work of ADS1262.

    4.  for the "short drop of +2.5V and -2.5V"， please refer to the below picture when the load cell cable was plugged in.

    The key point of this issue is: Why is when +2.5V short circuited to ground,  ADS1262 can back to normal operation immediately after the short circuit is removed, but ADS1262 cannot work normally after the short circuit is removed, when - 2.5V short circuited to ground ?  This fault will occur 100%. Maybe you can verify it on your Demo board.

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Thanks for providing this additional information, it helps us narrow down the possible issue that could cause this behavior. I will see if I can recreate this behavior on my EVM.

    If you want to share your schematic with me offline, you can hover over my e2e user name ("Bryan Lizon86") and select "Request friendship". This will allow us to communicate via private message, where you will be able to share your schematic. Or, if you want to wait until Irene returns, that is fine too

    One additional question: after shorting the -2.5V supply to ground, and then removing the short, the -2.5V supply returns to normal. However, the ADS1262 does not work properly, even if you toggle the RESET pin. Is there anything that can be done to get the ADC to work properly under these conditions? Can you power cycle the system? In other words, how do you get the ADC out of this "stuck" mode?

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    I have shared the schematic of our AD Option Board to you via private message.

    So far, I have found two method to get the ADC out of this "stuck" mode:

       (1) Power cycle the system (cool reboot).

       (2) Reset the MCU on option board ( I shorted the R108 on page 1 of the SCH to ground for about one second, and then ADS1262 could work properly ). 

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Thanks again for sending the schematic.

    I missed something in the original post from Irene:

    Irene Qin said:
    In fact, in the case of normal +2.5V and -2.5V power supply, the ADS1262 is artificially reset (20 feet pulled down), and then the chip can not work normally, the phenomenon is exactly the same as the -2.5V short circuit (the DOUT pin output period will become 20Hz, as shown in the figure below).

    It is interesting that this same behavior can be seen when you toggle the RESET pin on the ADC. Let's explore this a bit more:

    I did not see a pull-up resistor on your net "AD_RESET" - do you have one and I missed it? Or did you not have a pull-up resistor on RESET?

    If you do not have a pull-up on RESET, I believe that the device is not being fully pulled out of RESET once your toggle the pin. According to the scope shot you sent, the ADC data rate changes to 20 Hz (the default data rate). Assuming you are using a data rate that is not 20 Hz, this means that the device successfully reset. However, it seems like there is a long distance (20 feet?) between the ADC and the controller. Perhaps there is not enough drive strength to return RESET high after toggling this pin on the MCU. This would also make sense why resetting the MCU allows the ADC to resume normal operation, which would not have worked if this was an ADC power supply issue.

    I would suggest looking at the ADC RESET pin on your scope to see what happens when you try to toggle this pin. If the device is not resetting properly i.e. toggling low to reset and then toggling high to release reset, I would add a pull-up resistor on the RESET pin at the ADC to +3.3V and see if this improves the issue. You might also consider adding an IO buffer to improve drive strength

    As a sanity check, I tried shorting out the power supply on the ADS1262EVM, see the scope shot below. I was unable to reproduce the behavior you were seeing. Instead, the ADS1262 worked just as well before the AVSS short as it did after. I checked multiple input voltages, gain settings, and data rates, and they all worked the same (before vs after).

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    The reply is late due to the forum maintenance and the Chinese National Day holiday.

    First, "20 feet" means Pin 20, it is a mistake and has nothing to do with distance here.  "Pin" and "foot" correspond to the same Chinese character, it is

    the reason.

    Then, I tried to add 20K pull-up resistance on RESET (pin 20) and CS (pin 10) signals separately, it could not improve this issue. The scope waveform is as follows:

    (C2/blue --- -2.5V,  C3/red ---- RESET,   C4/green ---- DOUT)

    Since the behavior could not be reproduced on the ADS1262EVM, the issue must be on our design. Is it possible that the issue is caused by the setting of any register?

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Thanks for clarifying your previous comments

    The scope shot you sent shows that the RESET pin is always high. What happens when you toggle RESET? Can you include a waveform that shows this behavior? From your previous post it seemed that when you toggled RESET the device stopped working.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    The scope shot I sent you on Oct. 6 was "toggle -2.5V" .  Please see the below “toggle RESET” pictures . We can see from these scope shot that all the power (AVDD -- +2.5V, AVSS -- -2.5V, DVDD -- +3.3V)  work well when the RESET was toggled to GND, but the DOUT showed that the device had stopped working.

    +3.3V, -2.5V, RESET and DOUT

    +2.5V, -2.5V, RESET and DOUT

    -2.5V, RESET, CS and DOUT

    -2.5V, RESET, CS and DOUT    zoom in

    Another "toggle -2.5V" scope shot, which shows the same issue.

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Now I am confused as to what the actual issue is.

    In your scope shots you toggle RESET and the ADC appears to be working okay: the DRDY pin is toggling at approximately the default data rate (20 SPS), which means the ADC interpreted the activity on the RESET pin correctly and is indicating that data is ready as intended. And in the last image you sent, this same behavior occurs even when you do *not* toggle the RESET pin.

    In all cases I see the ADC working as I would expect if you turned off one of the supplies (which is basically what is happening here). What behavior are you expecting to see?

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    My question is: What is reason when +2.5V is short circuited to ground, ADS1262 can keep work normally (DOUT continue output data at a rate of 1200Hz）, but when - 2.5V is short circuited to ground,  ADS1262 will work abnormally ((DOUT first has no data output, then output at a rate of 20Hz ). What the difference between +2.5V short circuited to ground and - 2.5V short circuited to ground？

    +2.5V short circuited to ground（C2/blue --- +2.5V,  C3/red ---- DOUT）:

    - 2.5V short circuited to ground(C2/blue --- -2.5V,  C4/green ---- DOUT):

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Thanks for clarifying what the issue is. From the previous posts I was under the impression that the ADC did not work at all or was non-responsive after shorting out AVSS. I understand now that the issue is simply that the ADC resets when AVSS is shorted out, and this is not the behavior you want.

    However, I must ask: if your power supplies collapse to ground, why do you want the ADC to keep running? It cannot be guaranteed that the ADS1262 will be still be functioning properly or that the registers will hold their values when the supplies are shorted out. I would think you would want to reset the ADC when this condition occurs since you cannot trust that the register settings or output data are valid. Why do you want to operate the system this way?

    I will keep looking into this issue, though it may take some time to try to replicate the issue and then, if successful, get feedback from the design team regarding what is going on here. Thank you for your patience as we try to resolve this issue for you. 

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    please see my reply below.

    Bryan Lizon86 said:
    I understand now that the issue is simply that the ADC resets when AVSS is shorted out,

    I am not sure whether the issue can be defined as “ADC resets when AVSS is shorted out” ,  because from the scope shorts, the RESET signal works well （keeping the high level) when the -2.5V is shorted. But from the DOUT output, indeed, the behavior is the same as ADC be reset.

    Bryan Lizon86 said:
    if your power supplies collapse to ground, why do you want the ADC to keep running?

    Some products using other AD chips have been verified, no matter how the AVDD is short-circuited, it can automatically resume normal operation immediately after the short circuit is removed. The only difference is these AD chips are all powered by unipolar (AVSS is connected to GND)，but the current project is powered by bipolar (AVSS is connected to -2.5V).

    The requirement of our current project is that even if the AD power supply is dropped transiently due to heavy load added suddenly or short circuit occurs, the AD chip should work properly without needing to reboot the system. This is an excellent gene of our product family, and we can't give up on this generation. All previous products can meet the requirement. 

    Bryan Lizon86 said:
    It cannot be guaranteed that the ADS1262 will be still be functioning properly or that the registers will hold their values when the supplies are shorted out.

    when AVDD is short circuited to ground, ADS1262 can keep work normally , but when - 2.5V is short circuited to ground,  ADS1262 will work abnormally. Why?  I think all the registers should powered by DVDD (+3.3V), no relevant to AVDD and AVSS?

    Bryan Lizon86 said:
    I would think you would want to reset the ADC when this condition occurs since you cannot trust that the register settings or output data are valid. Why do you want to operate the system this way?

    It is just a try, do not care it. The real need is when the AVSS short is removed, ADC and MCU can resume normal data transmission, just like the behavior of AVDD.

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Thanks for your feedback, let me clarify: the ADS1262 power-on reset (POR) is triggering when the supplies collapse (see section 9.4.10.1 in the ADS1262 datasheet). So this has nothing to do with the RESET pin, it has to do with the internal POR circuit activating once the supplies drop below the POR threshold (see Table 9-30).

    Therefore, it makes sense why the ADC has reset under the circumstances you described. Moreover, this should be the expected behavior - it is not clear to me why the ADS1262 did not reset when AVDD was shorted to ground. I tried replicating your system on the EVM and I saw the following behavior:

    • AVDD = +2.5V and AVSS = -2.5V
      • When AVDD (+2.5V) was shorted to ground, the ADS1262 POR circuit did not trigger
      • When AVSS (-2.5V) was shorted to ground, the ADS1262 POR circuit did trigger, and the ADC reset
    • AVDD = +5V and AVSS = ground
      • When AVDD (+5V) was shorted to ground, the ADS1262 POR circuit did trigger, and the ADC reset

    I know you mentioned in your previous post that the ADC did not reset when you used a unipolar supply (+5V). However, this was not the behavior I saw on the ADS1262EVM.

    I will look into this issue further, though again it may take a few days. However, I will reiterate that you should expect the ADS1262 to reset when the supplies are shorted.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    Today I spent some time doing some further research. there are two pieces of information to share with you.

    (1) I changed the AVDD of our system from +2.5V/-2.5V to +5V/GND, When AVDD (+5V) was shorted to ground, the ADS1262 POR circuit triggered, and the ADC reset. The behavior is the same as what you see on the ADS1262EVM.

    (2) I measured another product which used AD7195, AVDD was powed by +5V and GND, when AVDD (+EXC) was shorted to ground, the behavior of DOUT (MISO) as follows, we can see that the POR circuit did not trigger.

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    One other thing to consider is the amount of capacitance you have on your supplies. I noticed in my tests that no matter how fast I tried to ramp down the supply voltages, there was always some delay due to the decoupling capacitors on the ADC. For larger capacitance e.g. 10uF, it took several milliseconds for the voltage to completely discharge from the capacitors and the POR to trip. So if the transient on the supplies is short enough, the ADS1262 POR will not trip. If the transient lasts for 10s of milliseconds, the POR will almost certainly trip. The point of the POR circuit is to ensure that the ADC is a known state should the power supplies collapse.

    Another concern with the supplies collapsing is how this affects the analog input voltage versus the supplies. If the ADC analog input voltage is somehow larger than the supply voltage, you could potentially be outside the absolute maximum ratings (where AIN < AVDD +0.3V). Since you have multiple power ICs in your system, I presume it is possible that this set of conditions could occur as the transient propagates through the system. Are you able to generate an independent supply to power the sensors? This way if the supplies collapse it will not affect the operation of the sensitive ADC circuitry. I would strongly encourage you to consider this as an option.

    I did not find any information on the POR circuit in the AD7195 datasheet, so it is not clear to me when / how it triggers (other than at power up). As stated previously, the point of the POR circuit is to ensure that the ADC is a known state should the power supplies collapse. It is up to you if you want to take this risk.

    -Bryan

  • 0 in reply to Bryan Lizon86
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    One other thought: how do you account for the data that is taken when the supplies collapse? If the ADC continues to sample during this time, you would have to throw away this data if the supplies, input, reference, etc., are all collapsing. However, you could never really know for sure when the ADC was providing good data again since you cannot assume what you're reading is due to changes at the input or due to power supply issues.

    I would think the POR would be very helpful in this case, as the ADC is held in RESET until the power supplies stabilize. Then you know the ADC is ready to provide good data again, and you can re-initialize the ADC to your desired configuration settings.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 85 points

    Hi Bryan,

    Thank you very much for your detailed clarification and suggestions, which are very valuable to us. We will make some improvements based on your advice and in combination with the actual situation of our project. If there is any good news, I will tell you.

  • 0 in reply to YongLi Chen
    TI__Mastermind 42871 points

    Hi YongLi Chen,

    Great, I am glad we could help. Let us know if anything else is needed

    -Bryan