TCA9801: Unexpected low pulse on I2C bus SDA line when using TCA9801 bus repeater

Karl Kracht

Part Number: TCA9801
Other Parts Discussed in Thread: ISO1540,

Hello,

I have an I2C circuit with a TCA9801 buffer and an ISO1540 buffer connected to each other. During a high-to-low transition, from the ISO1540 buffer, to the TCA9801, I'm seeing an unexpected low pulse on the SDA line. The ISO1540 is specifically designed to produce an output low level of about 570mV, in order to avoid triggering its own input circuitry. So, I'm wondering if this unexpected low pulse, which drives the SDA line down near zero volts, could cause the I2C bus to lock up. See attached pdf for more detailed information.I2C_buffer_ISO1540_to_TCA9801_issue.pdf

10 months ago

0 Tyler Townsend 10 months ago

TI__Guru 64738 points

Hi Karl,

I will move this thread over to the isolation team to see if this is expected behavior from the ISO1540 as I do not support isolated buffers.

For TCA9801, I don't think this buffer is re-driving the LOW back the other direction for that short duration shown. If data is coming back to the microprocessor, then maybe the ISO device is creating an inverted pedestal?

If the iso team finds no issue with ISO1540, I will debug further.

Regards,

Tyler

0 Varun Kumar 10 months ago in reply to Tyler Townsend

TI__Genius 17080 points

Hi Karl and Tyler

The VOL1 of ISO1540 is not zero but rather has an offset which go up to 0.8V.

This is an expected behavior of the ISO1540 and will not hinder any communication.

FAQ on the VOL1 being 0.8V: (+) [FAQ] Why is the logic LOW level output voltage, VOL1, up to 0.8V on Side1 of the ISO1540/ISO1541 and ISO1640/ISO1641 bidirectional I2C isolators? - Isolation forum - Isolation - TI E2E support forums

I request you to read section 7.4 in the datasheet for detailed explanation on this VOL offset.

Regards
Varun

0 Karl Kracht 10 months ago in reply to Varun Kumar

Prodigy 10 points

That's not the question I'm asking. Please read my original post, and please also look at the supporting information I provided in the pdf file attached to the original post.

Karl

0 Varun Kumar 10 months ago in reply to Karl Kracht

TI__Genius 17080 points

Karl,

That's the exact reason why I asked to read through section 7.4 in datasheet.

See Figure 7.4 and read the followed explanation for transmit mode - And you'll understand that this is an expected device behavior.

Regards
Varun

0 Karl Kracht 10 months ago in reply to Varun Kumar

Prodigy 10 points

I'm confused now. My question is about receive mode, not transmit mode.

Karl

0 Varun Kumar 10 months ago in reply to Karl Kracht

TI__Genius 17080 points

Karl,

Apologies for the confusion as I misread the waveforms initially due to similarity with transmit behavior of ISO1540. Allow me to restate below:

According to the PDF and the zoomed waveforms, ISO1540 is transmitting data to TCA9801.which means SDA1 is in receive mode. And this 400ns LOW pulse cannot be generated by ISO1540.

Now, addressing your concern regarding that 400ns pulse - " My concern, is that the rising edge of this low pulse, will trigger the input buffer on the ISO1540 chip, and make the chip think that the bus wants to go to a logic high, while it's driving a logic low. This could cause the ISO1540 to latch and cause the I2C bus to latch up. Are my concerns valid? Can this low pulse cause the I2C bus to lock up? The bus is running at 100KHz"

The rising edge will not change any state on Side1 of ISO1540 but only on Side2.
1. Output D on Side2 will be turned on during the LOW state of the 400ns, but this will affect any state since SDA2 is already pulled low.
2. When signal rises back to VOL1 level and crosses VIHT1 levels - it will only switch off D which will again not affect any state of bus on either side.

Regards
Varun

0 Karl Kracht 10 months ago in reply to Varun Kumar

Prodigy 10 points

OK, thanks for your help.

Karl

Interface

Interface forum

TCA9801: Unexpected low pulse on I2C bus SDA line when using TCA9801 bus repeater