overshoot/undershoot analog output square wave

[rager]

Hi,

I am generating a voltage signal with a PCI 6115 DAQ card, connected to a BNC 2110 connector block.  I am outputting a simple square wave produced by the Simulate Signal Express VI.  The wave is 1 kHz, 0 offset, 1V amplitude.  When looking at the voltage output on an oscilloscope, I notice overshoot on the leading edge and undershoot on the trailing edge.  Similarly with a sawtooth, there is undershoot on the falling edge.  I've plotted the simulated waveform on a chart in Labview and it looks fine, but there seems to be some hardware issue here causing the over/undershoot.  I've tried different amplitudes, offsets, frequencies, and switched output channels with no luck.  Any ideas?

Leyen

[Seth_B.]

Hello Leyen,

What cable are you using with this board?  What you are seeing is most likely an artifact of additional capacitance on the output lines.  Using short cabling and connections would probably help.  Can you post a picture of the output?

[rager]

Hi Seth,

I'm just using BNC cables to look at the output.  From the computer to the BNC connector block is a 184749c cable.  Attached are screenshots of a 1kHz, 1 V square wave.  As you can see, the magnitude of the spike oscillates.  Left axis is V, bottom axis is seconds.  Thanks for your help with this.

Leyen

[Seth_B.]

What length (1m, 5m, etc) of the cable are you using?  What devices are you using to measure the signal?  And what is it's input impedance?

[rager]

I'm using the shortest possible length of cable - 50 cm.  I have measured the signal 2 ways - by looking at the output on an oscilloscope and by acquiring the signal with another BNC 2110/ PCI 6115 and looking at the output in Labview.  The latter is where the screenshots are from.

With the oscilloscope impedance at 1 MOhm I see the spiking.  However when I change it to 50 Ohm it goes away. So it looks like this is caused by impedance mismatch, as the DAQ boards output at 50 Ohm and inputs are at 1 MOhm.  THanks!