SN74LXC8T245: Delay from OE and DATA OUT from A to B

Where do the other pulses on the output come from? Is there anything else connected to the output? Do you still have the problem if you disconnect the B pin?

Hi Clemens

The other pulses are coming from another driver that is disable before to enable this driver.

We have tried removing all the resistors connected to the B outputs, but we still have the problem.

This is very strange. This looks as if only the first input signal edge after /OE going low enables the output.

Is there some delay between /OE going low and A going high, or do they happen at the same time?

We try different combination of delay between the OE signal and the Data Input up to 120nS of delay (see the following image), but the issue stil present.

And this waveform shows that the output gets enabled some time after the data goes low, so it is probably not related to any signal edges.

This exceeds the specified enable time. The only explanation I can think of is that the device got damaged. This would typically be caused by overvoltage on some input/output/supply pin. I do not know if there is a pattern of channels or chip where this occurs.

I don't think is an overvoltage problem because the device is powered with 3.3 Volt on both side (A and B).

This morning we discovered that the issue it is present if the OE signal remain into the disable state for more than 400/500 uSec (see first image) and it deseappers if the disable state is short (see second image)

Hi Doriano,

Schematic is not fully shown. Please help confirm if the below is applicable, thanks.

Best Regards,

Michael.

Hi Michael,

we are using the SN74LXCH8T245 device with the Bus Hold.

Sorry for the mistake

Hi Doriano,

Please help confirm below:

• Outputs are still toggling with device disabled when the device is replaced with a known good unit
• Verify power supply rails are clean and stable without any noise or fluctuations
• Ensure outputs are not being externally pulled up or down even when device is disabled
• Double check /OE is not enabled until power supplies are fully stable

Best Regards,

Michael.

• Yes, of course the outputs are still toggling when the device is disable. It is on a bidirectional bus that is shared with other devices, but note: we didn't replaced the part.
• The Power supply are clean and absolutely stable.
• The output are floting when no device drive the bus. Note: the device we are using is that having the 'Bus Hold' feature on the 'A' and 'B' side.
• The Output enable of the device is driven by a FPGA which has the same supply of the side 'A' of the device. The Output enable is driven high by the FPGA at the Power-on Reset, therefore it follows the rising of the Power supply.

Please note what I wrote last time: when the time between two consecutive activation of the device is less than about 400 microseconds, the device behavies well. When this time is greater, the device exibits the delay that I have documented.

We suspect that it has a something like a 'deep power down' which trigger when it stay disabled for a so long time.

Hi Doriano,

Thanks for your feedback. I am double checking with the design team to help clarify your suspicion.

In parallel, could you help confirm if a known good unit reveals similar observations? Thanks.

Best Regards,

Michael.

Hi Michael,

Sorry, what do you mean with "In parallel, could you help confirm if a known good unit reveals similar observations? "

To find out whether the chip got damaged (by something that happend earlier), solder a new chip in place.

Hi Doriano, 

Team confirmed device do not have special features per your observations. 

If a new device observes the same issue, concern will have to be due to bus contention. 

I.e. output will not toggle if input is not toggling regardless of OE (per the first image provided) and concern is not device related but system related, as such pulse must be present on the line with or without the LXCH device, thanks. 

Best Regards,

Michael.