ADS1225: DRDY pin remains high for 442 ms in case of temperature 0°C / in case of temperature 20°C it takes only 17ms

Hi Richard Schreck,

I am not aware of any temperature dependency on the ADS1225 startup time.

On how many boards do you see this issue? You said "most do not have this problem", so how many do, and out of how many total?

If you continue to decrease the temperature (or increase I guess), do you see any additional dependencies?

Have you looked at the output signals on a scope instead of a logic analyzer to see if the analog behavior matches the digital representation i.e. no large over/undershooting, ringing, slow ramping, etc.?

-Bryan

Hi Bryan,

Thank you for your quick reply and sorry for my late answer. In the meantime we did some additional tests in relation to your questions. See it described below:

There are currently about 500 boards in use at the customer, at the moment 25 boards show this issue. Further 1500pcs are already assembled on the board and ready for use in the product for the customers.

After the startup time, the ADS1225 works without any problems. The conversion time and the AD values are OK.

The issue only exists when the supply voltage is applied and the AVDD/DVDD pins are at 0 volts before. If the AVDD/DVDD pins have a voltage of approx. 400 mV before the supply voltage is applied, the issue does not exist.

I think that in this case the IC does not make a POR. The issue only exists when the ADS1225 makes a POR.

If the IC does not make a POR, after applying the supply voltage, the DRDY pin remains low and it works without issues.

If we change the temperature, we do not see any additional dependencies.

Below some images measured with a scope.

Image 1 (10 ms/div Signal AVDD/DVDD and DRDY/DOUT)

Image 2 ((1 ms/div Signal AVDD/DVDD and DRDY/DOUT))

Image 3 (50 µs/div Signal DRDY/DOUT and SCLK)

Hi Richard Schreck,

Richard Schreck said:

The issue only exists when the supply voltage is applied and the AVDD/DVDD pins are at 0 volts before. If the AVDD/DVDD pins have a voltage of approx. 400 mV before the supply voltage is applied, the issue does not exist.

To be clear, this means that the issue happens every time you power on the system, because the supply voltage is always 0V at this point? And then you have to droop the power to ~400mV so the POR triggers, and then everything works fine?

-Bryan

Hi Bryan,

On bad boards the DRDY pin stays high at 20 degrees for 15 ms after we apply the supply voltage. See image 1.

If we cool down the temperature of the ADS1225 to 10 degrees, the DRDY pin will stay high for 600 ms after we apply the supply voltage. See image 2.

If we cool down the temperature of the ADS1225 to 0 degrees, then the DRDY pin remains high for 974 ms after we have applied the supply voltage. See picture 3.

The problem is that it depends on the temperature how long the DRDY pin stays high.

Do you have any idea why some ADS1225 show this behavior? This problem does not occur with good boards.

With these good boards, the DRDY pin remains high for approx. 17 ms regardless of the temperature.

Pin assignment for the measurement

AVDD/DVDD = 3.3 VDC

BUFEN/TEMPEN = GND

MODE = 3.3 VDC

SCLK = GND

START = GND

Image 1 (50 ms/div temperature 20 degrees)

Image 2 (200 ms/div temperature 10 degrees)

Image 3 (500 ms/div temperature 0 degrees)

Hope you can give us an idea about the problem.

Best regards,
Richard Schreck

Hi Richard Shreck,

In your second to last post, you made it seem like as long as there was some voltage on AVDD then there was no temperature dependency - is that correct? I am just confused because this last post seems like a repost of your first post, even though you found some (seemingly) different behavior in the interim

How are you cooling down the ADC? With some sort of freeze spray? Or using a thermal chamber?

FYI I am on business travel this week so my responses might be delayed a bit until next week

-Bryan

Hi Bryan,

yes, the temperature dependence is the main problem. We now see quite a few ADCs with this very unusual behavior depending on the temperature.

We have different cooling methods available. No matter whether we slowly lower the temperature in the climate cabinet or with cold spray, or whether we put the circuit board outside at night.
We always see that the PIN stays high for longer depending on the temperature on some boards, the colder the more extreme this becomes.

And yes, this always occurs when the voltage is completely switched off.

We need a solution for this, I don't think this behavior is standard. In general, we have thousands of the ADS1225 in stock that should be used now, but our customers are having significant problems with those that are already in use.

We therefore primarily ask ourselves the following questions in the hope of an answer from you: 
-Do you have any idea what we might be doing wrong in this regard that could cause such an error? (We have over 30 years of experience with A/D converters, our expert with the most experience in this area has now also worked on it, but no one has seen such effects before)
-Could it be a production defect in the ADS? This is what we currently believe....

Looking forward to an idea from you.
Best regards,

Hi Richard Shreck,

I have not seen this issue before with these devices, and we have shipped many of them over the years, so it is difficult to provide feedback as to how to proceed. Let me ask some additional questions:

• According to the info you have provided, it seems like there are some devices that show this behavior, and some that don't - is that correct? Have you tried swapping a "good" device with a "bad" device to see if the issue follows the device or the board?
• You also said "we now see quite a few ADCs with this very unusual behavior...". Is that to imply that you are seeing more than initially thought (you originally said 25 out of 500, or 5%)?
• It seems like a customer identified this issue - is that correct? Did you perform any over-temp testing before putting these boards into production? I am just curious if the boards were "good" when they shipped and then something went wrong, or if this behavior was just not at all characterized before being sent out to customers?
• Does the DRDY high time get shorter if you increase the temperature to >20 degrees?
• Is there any consistency from device to device with respect to the DRDY high time across temperature? For example, could you provide a table for the DRDY high time of all 25 "bad" boards at the three different temperatures you are testing (0, 10, 20)? Or is there a lot of variation from device to device?
  • What about consistency for a specific device? For example, if you run one board / ADC from 20 to 10 to 0, then back to 20, then 10 then 0, and so on for maybe 10 cycles, are the DRDY high time readings consistent? Or again, a lot of variation?
• Can you send a schematic of your system?

-Bryan