ISO1640: There is the spike on the waveform

Hello Moon, 

I believe this post is a duplicate of ISO1540: ISO1540 related issues. As stated in the other post, I do not expect to see the spiking behavior from either ISO1640 or ISO1540 based on the schematic assuming both sides have pull-up resistors.

In the schematic shown in this thread it looks like the application is using the ADI. Can you confirm that the spikes are from a TI device? 

Best,
Andrew

hi Andrew，

Thanks for the reply.

Yes they are the same issue. I confirm it use our ISO1640 to replace the device of ADI in the schematic.

They added the pull-up resistor on both of sides, but the spike issue was not solved.

Do you have other recommend ways to solve this problem?

Another question, can ISO1640 exchange 1 and 2 side? for example, we use 1 side to connect slave and use 2 side to MCU?

BR,

Moon

Hello Moon, 

If both sides of the isolator have pull-up resistors, then there is nothing on the schematic side that would cause this issue.

The official recommendation is that Side 1 is meant to be connected to one master controller or MCU. Side 2 has a higher capacitive load and is designed to be connected to the I2C bus. 

After reviewing the information again, can the customer please explain how the waveform is produced? these rise to an intermittent level. I do not expect the ISO1540 or ISO1640 to produce an intermittent high like this unless the VCC has been changed to 1V. Is it possible that the customer's VCC is shifting?

Best,
Andrew

Hi Andrew,

Thanks for your reply! After checking, I found the waveform from 2 side have some spikes and the correspond waveform on 1side appears. 

I also found the pic9-7 in datasheet as below:

Is that the reason why the low voltage level on 1side can't reach 0V?

BR,

Moon