LPC5411x Vref Sourcing Current !

[MuhammadTharwat]

Hello all, 

We 're working on an ultra low power application which uses the ADC and VREFP pin (VREFN is always grounded). A buck-boost controlled power supply is used to provide the VREF while using the ADC. 
This Buckboost is then disabled (producing theoretically 0 voltage) when the ADC is not used and the device is put in deep sleep mode. 
We notice some different behavior using the "same board" and different LPC chips. In 50% of the cases, the current profile was correct and in the other 50% the current went higher by an amount of 35 - 45 uA. 
We doubt that this is cause by VREFP sourcing some current when it is not used. 
Is there a chance that VREFP can source a current (maybe through pull up resistor) ? 
This is very concerning and we have tried everything we could till we came to the firm conclusion that it is the LPC chips that are not acting consistently under the same board and Firmware. 
Can you advice please ???

[Pavel_Hernandez]

Hello, 

I reviewed the datasheet, Did you take this measurement in a laboratory to have the same environment? this is mentioned in the datasheet. 

• Typical ratings are not guaranteed. Typical values listed are at room temperature (25 C).

You mentioned "the current profile was correct", Could you tell me what the power consumption was correct?

Best regards,
Pavel

[BW2]

the board is quite simple .. here is what it has

1. a thermistor connected to ADC

2. a bust boost whihc is controlled by an IO

3. this buckboost outputs 3.3V

4. this 3.3V is connected to VREFP and Thermistor

5. we turn on buck boost when we read Thermistor and turn it off after reading thermistor 

6. 50% boards read around 12-16uA in sleep and other 50% read between 35-60uA

7. we swapped the MCU between good and bad board and issues followed MCU

[MuhammadTharwat]

Hello Paval, 

Thanks for reaching out. 
Our correct current is 12 to 16 uA. 
We have used few MCUs on the same board to narrow down the number of variables. 
Some MCUs consumed 35 - 60 uA while others consumed the expected 12-16 uA range. 
VREFP is connected to a voltage divider (two series resistors and ADC pin is reading from mid point) and a Buckboost converter output. 
We tried the following : 
1- If we remove the Buckboost converter , nothing changes and current is still high on 50% of the MCUs we tried. 
2- If we remove the voltage divider (and leave VREFP floating and disconnected from anything) the current goes down to our expected range (12 - 16 uA). 
3- We also tried disconnecting the ADC input pin from the voltage divider mid-point so that it becomes VREFP is just pulled down using these resistors , the current went high again ! 

We believe there is something wrong as VREFP seems to be sourcing current !
Can you confirm if that's correct ? 

[Pavel_Hernandez]

Hello,

I appreciate the information, we follow this case in the other. This case has another replay.

Best regards,
Pavel

[BW2]

what do you mean ?

[Pavel_Hernandez]

Hello,

This case has another ticket 477019 I will follow your case in that.

Best regards,
Pavel