Raspberry Pi I2C Shield (PR2-2) has incorrect pull-up voltage

Hey there,

We are using one of the I2C shields for our application (https://store.ncd.io/product/i2c-shield-for-raspberry-pi-3-pi2-with-outward-facing-i2c-port/). We recieved Rev. C of this board. While debugging some I2C issues, I found that the pull-up voltage between the Raspberry Pi and the Shield was 3.6 and 3.7 volts.

Does anyone know why the pull-up voltage is above spec for the Raspberry Pi? Does anyone know a particular Shield that does not have this problem?

Thanks!

Hi,
The SDA and SCL lines are pulled hight to 3.3V using a 4.7K.
SDA and SCL lines go into a level shifter, which pulls them to 5V.

Thanks

Hi Bhaskar,

I have evidence that the lines ARE NOT being pulled up to 3.3v. Do I need to show it to you?

How many of the Shield variants have you used? Who can I call to get my order returned?

Hi,
All our I2C shield come with 4.7K pull ups enabled including the PR2-2.
I am attaching the sch.

I just ran one quick test over here and did this
Connected the shield to a 5V supply ( shield is not plugged in the pi) and measured voltage at SCL and SDA lines and the lines are pulled up to 3.3V.

we have sold thousands of these and never had any issue with pi. If you can explain more about the issue we can run some more tests.

The issue is that the pull up voltage is not correct. We have 80 of these boards. I’ve tested ~5 and all have this issue.

Hi,
Please use this fillout an RMA and it will have return instruction.
https://ncd.io/contact-us/product-returns/

Thanks

Hey Bhaskar,

OK I will do that. Can you tell me what board you were testing on that gave you the correct pull-up voltage?

We need a solution for this ASAP. Do you have any idea which of the shields currently available actually have the correct pull-up voltage?

Hi,
I did more digging into it. I am using PR2-2.
My test input power supply was around 4.4V or so and it was producing 3.3V. Once i noticed that, i changed the supply to 5V and the voltage started to go close to around 3.6V or so. The root cause was the level shifter.

We changed the Vref2 voltage ( level shifter) on the board and then the I2C voltage went back to 3.3V.

Once we get your boards we will get these fixed and will ship back next day.

Thanks