controlling fall time on PXI-6551

[ajog]

Hello,

I find that high to low transitions on one channel say A will couple into channel B and vice versa.  I have tried obtaining a better ground, using bypass capacitors etc. to no avail.  Is it possible to slow down the high to low transition so the HSDIO drivers sink less current?  Equivalently, instead of a sharp transition, can the slope be reduced so that the high to low occurs in a finite amount of time say 50 nanoseconds?

Any other thoughts on reducing channel to channel transition coupling are greatly appreciated.

In my setup, the high is obtained via tri-stating a node to which a pull-up resistor is attached.  Instead of tri-stating (or 'Z') if I drive a '1' the crostalk is understandably much less.  However, I really do need to tri-state the channel.

Please advise.

Thanks.

Anand

[ajog]

Hi All,

This has been resolved.  The oscilloscope probe ground must be kept as close to the DUT ground as possible and the perceived crosstalk diminishes.

Any tips on reducing the HSDIO current sink capability?  Right now it is preset to 50mA per channel.

Thanks.

Anand

[Keith.Shapiro]

Hi Anand,

I'm not sure I understand.  The current sink is the amount of current that the buffer can source from its output to ground and still present Vol(max).  If you do not have any current paths on the line (no other drivers, and no resistor pulls to other voltages), the amount of current being sunk by the buffer is virtually zero (just the leakage current).

Changing the current source/sink capability of the driver is a) impossible, and b) will not affect the slew rate of the driver, simply the amount of current that can pass through it before the part no longer meets the specifications for output voltage level.

Can you be more specific as to what you're trying to do?

Thanks,

Keith Shapiro

National Instruments R&D

[ajog]

Hi Keith,

Thanks for your response.  We are indeed trying to change the slew rate of the driver when it switches tri-state to low.  There is a pull-up of 500Ohms.  The HSDIO driver is strong enough to pull it low enough.  However, we are worried that such a fast transition (2.5ns for a switch from 3.3V to 0V - per HSDIO spec.)might create signal integrity issues in our system.  

Since the slew rate is not adjustable this is not possible right?

Anand

[Keith.Shapiro]

Hi Anand,

There is no slew rate control on the NI 655x product line, however, if you add additional capacitance to your line, you can slow down the effective slew rate at your receiver.

Keith Shapiro

National Instruments R&D