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.

ads1241 - energy getting into measurement circuit when MUX channel selected

Other Parts Discussed in Thread: ADS1241

Was just at my client's site tracking down some of these ADC problems and found another very strange one that I need help on. It looks like a component problem but I need to get an opinion. Again, I am using the part in strictly register read/write mode as with my other posts.

Symptoms:

On certain channels (channel 1 in particular), the ADS1241 seems to be injecting energy or outputting a voltage when the MUX channel is selected. This is causing the 4.1V signal to change to 4.6V on channel when the MUX is selected to channel 1. The input stays at 4.6V until a different MUX channel is selected and then returns to the 4.1V level.

I have at least 2 parts that show this problem and probably a bunch more.

Replacing the ADS1241 with a known working part resolves the problem.

This product has been in production for 10 years without any problems on this ADC until about 1 year ago.

Circuit:

The driving circuit is a buffered sensor output that is driving the signal through a 2 pole RC low pass filter. The first pole is 1k/1.0uF the second pole is 10k/0.1uF. This puts a total series resistance of 11K between the sensor and the ADC. The sensor output is capable of driving 100uA of output current. The sensor output is generally static as it measured atmospheric pressure and so you can really treat this as a fixed DC source.

ADC Register read/write loop:

This uses the older version of the read/write loop.

0. set channel to 0
1. poll the ACR register for DRDY 0
2. read DOR (cmd 1)
3. set the MUX channel (register 1) 0xN8 where N is the channel
4. increment channel, if 8 then set to 0
5. wait about 88 ms
6. go to step #1

ADC Configuration:

0x50, // Writer register command starting at register 0
0x0C, // number of registers (minus 1) to write

0x00, // Setup Register
0x01, // Multiplexer Control Register
0x04, // Analog Control Register
0x00, // Offset DAC
0x00, // Data I/O
0xFF, // Direction Control for Data I/O
0x00, // I/O Configuration Register
0x00, // Offset Calibration Coefficient (Least Significant Byte)
0x00, // Offset Calibration Coefficient (Middle Byte)
0x00, // Offset Calibration Coefficient (Most Significant Byte)
0x59, // Full-Scale Register (Least Significant Byte)
0x55, // Full-Scale Register (Middle Byte)
0x55 // Full-Scale Register (Most Significant Byte)

Test Data:

I can provide schematic of this application and scope traces of the problem. I would rather not post those on the forum for reasons of IP protection for my client. Probably not a huge IP issue but I would need to get permission from my client first.

Failed Parts:

I have two failed parts that were removed and I can send to TI for failure analysis and testing.

  • Forgot to add a section on what has been tried so far.

    1. I tried to set the MUX register to 0x88 prior to step 3 in the register read/write loop. I thought that maybe the previous channel was injecting the signal into the circuit when the MUX switched. Setting the MUX channel to 0x88 provides common to both inputs which should result in 0V on the sample and hold cap. This did not change the symptoms.

    2. Since the analog inputs can also be digital outputs then I put a register read of the ICON register in the register read/write loop and set a break point if it ever came back non-zero. The break point was never hit so this tells me that all the pins are configured as analog inputs. This ruled out a spurious digital output configuration problem.

    Another Symptom:

    1. Other channels have this same problem but to a lesser degree. This may just be a coincidence, but if the input voltage is lower then the problem is not as severe. Channels with input voltages in the 1.6V range did not have a measurable shift when the MUX is selected for that channel.

  • Hello John,

    I will be happy to assist you on this issue.   I will be contacting you directly via email.

    Thank you and Best Regards,

    Luis

  • Problem was caused by a blown AVDD 5V LDO. The ADS1241 can source current from the AVDD pin when the LDO opens. This was supplying 5V to the AVDD power rail and powering the other analog circuits thus it appeared the AVDD power was working correctly. Removed the ADC and measured AVDD at the pad and it was only about 200mV.

    Be careful when troubleshooting power problems on the ADS1241 because the part can source current from the AVDD pin and make it appear the power rail is OK even when it has failed.

  • John,

    Thanks for the update! Yes, it is possible to back drive the analog supply if a voltage higher than AVDD is applied at the analog inputs or VREF inputs.  This happens through the ESD protection diodes.  This can be very confusing when troubleshooting your system.

    Best regards,

    Bob B