TXB0104 not working at 1.8V with 12Mhz SPI Clock

Is this the reason why is works at 3.3V and not at 1.8V?

Also I am using TPS73018 LDO to generate 1.8V from 3.3V. This is suppling power to VCCA and controller2. This LDO can source only 200ma. Will it be sufficient enough to provide current?

Hi

Based on the schematic and requirememts shown, the TXB0104 should work without any issues.

Please clarify what happens " when its not working" - no output, or erroneous output etc.

Thanks

HI Jennifer,

    I took the traces from the analyser. It shows 3 lines: CS, MOSI and CLK. When I expand the waveform, shown in fig2, I see that the clock is skewed a little bit. The data on the MOSI line is different (@1.8V, 12Mhz) than the data I get when I run  (@3.3V, 12Mhz). These traces are taken at the VCCA side of my diagram.

Any clue on why the clock is behaving like this?

Regards,

Vineet.

Hi,

   The analyser above is sampling at 25Mhz, hence the clock may have appeared like the one shown above. When I changed the sampling rate to 100Mhz, the clock looks ok. I reduced the wire length on the prototyping board to less than 3 inches on either side of the shifter to make it work.

Now at 1.8V my SPI is working clocking at 12Mhz.

But I have another issue: When I try to probe the MISO line from the 1.8V side, then the system does not work. I tried with both the oscilloscope and salae logic analyser. And also, it works at 3V, 12Mhz even when the MISO is probed?

Do you know why? Because probing the signals shall be required on the PCB I make later.

Regards,

Vineet.

Hi

One possible reason is that there is leakage current on that line, that is discharging through the probe capacitance.

Is having a pull down an option?

Thanks

Hi

Since the scope probe has a ground, you  have a path to gnd. My thought process is that if SPI is infrequent, it will lead to a floating input, which might cause this anamoly. Maybe we can tackle it from the other end, ensure that the inputs to the TXB0104 are always at a defined state.

Thanks

Do you know what your probe capacitances are?  also, what does you layout look like?  Do you have any other components hanging on the signals lines?  My concern would be to much capacitive load on the signal lines and it could cause one shot retriggering or some sort of bus contention or oscillation.