SN65LVDS100: AC-Coupling and Unpowered Receiver

Hello Josh,

Thank you for your suggestion. Unfortunately, the board with the receiver will be off so no voltage is available for the biasing circuit. Can you recommend another solution?

Regards,

Francesco

Hello Francesco,

Allow me to confirm a solution with another peer. Due to the holiday break, please expect some delays.

Regards,

Josh

Hello Josh,

The comment "any voltage applied to a receiver's input could damage it" is very surprising. We selected this buffer because based on its datasheet it can accept VI between 0V and 4V (not related to VCC) (https://www.ti.com/lit/ds/symlink/sn65lvds100.pdf?ts=1673197571627&ref_url=https%253A%252F%252Fwww.google.com%252F table 8.3)

We use the buffer for board-to-board communication therefore we cannot control overall power sequencing.

Can you confirm that there is an error in the datasheet? If yes, what would be the maximum clamp diodes current acceptable when the part is unpowered?

Regards,

Francesco

Hello Francesco,

My comments above come from another E2E post (link) in which another customer experienced a damaged LVDS receiver when the receiver was turned off and the driver was kept on. I haven't seen this kind of issue before but with the other E2E post that I linked, I just wanted to express caution.

Since this device has an EVM, I can run a quick lab test on it and see if providing a signal to a turn-off SN65LVDS100 will damage the device, if that's okay with you. Given the US holiday next Monday and having to order the EVM through our site, some delays will be expected.

Regards,

Josh

Hello John,

Thank you for your help. A test is useful, however is it possible to get a confirmation by the buffer designers that it can supports driven inputs when unpowered? 

I can wait until next week for an answer.

For the provided link:

1- Was it confirmed that the damage was due to unpowered receivers?

2- What was the protection solution used?

Regards,

Francesco

Hello John,

We found two forums tickets that seems to indicate that it is ok do drive the inputs even if the SN65LVDS100 is off.

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/180560/sn65lvds100d-power-question?tisearch=e2e-sitesearch&keymatch=sn65lvds100#

https://e2e.ti.com/support/interface-group/interface/f/interface-forum/648389/sn65lvds100-receiver-equivalent-schematic?tisearch=e2e-sitesearch&keymatch=sn65lvds100%2525252525252520input#

Is this info still valid?

Regards,

Francesco

Hello Francesco,

Thank you for sharing both links. Yes, given Max's response from the first link, there should be no issue with providing a voltage to an unpowered SN65LVDS100.since the device will be in a high impedance state.

However, care should still be taken with inputting a negative voltage to the device's input pins.

Regards,

Josh

Hello Josh,

Thank you for the confirmation. Coming back to the original topic:

1) Do you have a recommended protection circuit to support negative voltages?

2) Is there an acceptable leakage current for negative voltages?

Regards,

Francesco

Hello Francesco,

We don't have a protection circuit that supports negative voltages. Our test conditions are performed for positive voltages so I cannot say when what would be the acceptable leakage current for negative voltages.

Regards,

Josh

Hello Josh,

Unfortunately we cannot add the biasing circuit on the transmitter side. 

Is there an acceptable maximum current with negative voltages? We could also clamp the negative voltage to keep it very small (-0.3V)

Regards,

Francesco

Hello Francesco,

Based on the datasheet, we know there is a power off protection circuit when the Vin has a greater voltage than Vcc. For those test conditions, the allowable current is 33 uA.

However, these test conditions are for positive voltages only. I cannot say what the maximum allowable current for negative voltages is since this goes against the devices recommended specs.

Regards,

Josh