1
\$\begingroup\$

I am currently using SN74ABT5402ADW as a buffer and level shifter for my ADC, however, the datasheet does not explicitly say that it can be used as a level shifter. I provide it with 5V power, but the input signals (out of the ADC) are 3.3V. The performance is bad: when I scope the signal on the output pins, the signal is not even close to look like binary HI/LOW. It does have rise/fall edges, but HI/LOW values seem to float, especially at xxx1111 to xx10000 transitions. I suspect that it does not convert the signal from 3.3V to 5V properly. Could anyone tell me if I am misusing it as described, and what can I do to make it work?

I use BNC2110 and CB-68LP connector blocks that connect to three PF50 connectors shown on the schematic. The bloks, in turn, are connected to NI PCI6259.

enter image description here

Measured signal at the input (yellow) and the output (blue) of '5402A when it is not connected to the NI connector blocks.

enter image description here

Measured signal at the input (yellow) and the output (blue) of '5402Awhen it is connected to the NI connector blocks via 8" unshielded wires.

enter image description here

Power at the power pin of the '5402A:

enter image description here

All ground pins on both NI connector blocks are interconnected internally.

  • Cable 1-3: Digital lines
  • Cable 4: trigger signal and GND wire
  • Cable 5: GND wire

enter image description here

\$\endgroup\$
  • 3
    \$\begingroup\$ Is the input voltage to the '5402 at least 2 V? In what way is the performance "bad"? What kind of load are you driving and have you considered the effect of the '5402's 25-ohm output resistance? \$\endgroup\$ – The Photon Jul 23 '14 at 22:49
  • \$\begingroup\$ @ThePhoton See post update, please. I can not find from the datasheet what kind of load I am driving. And I did not consider the effect of the 25ohm output resistance. What does it do? \$\endgroup\$ – Nazar Jul 23 '14 at 23:21
  • \$\begingroup\$ When you put the scope probe on the input pins of the '5402A, are the high levels above 2 V and the low levels below 0.8 V? How often are these levels changing? \$\endgroup\$ – The Photon Jul 23 '14 at 23:51
  • \$\begingroup\$ Also, do you have a common ground between this circuit and the NI-DAQ board? Is there a connection between the ground of this board and one of the ground pins of the NI-DAQ inputs? \$\endgroup\$ – The Photon Jul 23 '14 at 23:53
  • \$\begingroup\$ @ThePhoton Yes, the ground is common. The voltages on the input pins are above 2V and below 0.8V for high and low respectively. However, the outputs measure a mess. \$\endgroup\$ – Nazar Jul 24 '14 at 0:01
0
\$\begingroup\$

Even with the information from the scope traces, there's still a few things this could be:

  • Ground disconnect: Make sure the ground connection between the ADC board and the DAQ board is well-connected.

  • Cross-talk: When one line switches it could induce a signal on anther line. This is particularly likely since it looks like you have many more signal lines than ground/return lines. Two ways cross-talk could be created:

    • Mutual inductance: If you move the connection wires around relative to each other, does that reduce the problem? If that's it, try to keep all the signal wires as close to the ground wire as possible. The ~50 ohm series resistance you have in each output should mitigate this, but you never know.

      To reduce this problem, keep the victim signal line as close as possible to the ground wire.

    • Return line impedance: If there's substantial resistance in the return wire (due to a dodgy connector, for example, or if the return wire is much longer than the signal wires creating excess inductance) that could cause a ground bounce that allows a switching current in one wire to affect all the other wires.

  • Ground loop: This is less likely, but I'll mention it for completeness. If the ADC board has two different ground connections back to the DAQ system (like one direct connection and one through the power supply to the wall to the other power supply) then this loop could pick up interference. Since the "interference" here seems to be synchronous with the data signal, this is pretty unlikely, but maybe your data is synchronous with some external event?

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ I shielded all the cables and connected shields to the ground at the board and at the connector boxes, but the problem remains. Look at the third image from the top (blue signal). This is what I read on one of the pis of '5402 when the connector box is connected. Right after the rising/falling edge, the signal drops/rises respectively. The drop is fine, since it is still higher than the minimum "high" voltage. However, the sudden rise after the falling edge looks like is over 1.5V, whereas should be less than 0.8V to be considered zero. Any other suggestions? \$\endgroup\$ – Nazar Jul 28 '14 at 19:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.