NI5761 cross talk

[abikutn]

Hi

We have a NI5761 (DC coupled version) with NI 7951. It is configured to run with an external clock rate at 80MHz. And an external trigger is applied to the trigger in port also.

When we are trying to test the cross talk of difference channels, there some problems with result. When we just give a sqare pulse signal to CH1, channel 2 3 4 are giving 0.615V dc signal (as showing in picture 1). However, When I give the signal to Ch2 (or Ch3 or Ch4), Ch1 always showing some noise like result, as showing in the attached picture 2.

I am so confused by this result. Can some one help to explain?

[A_Donfigulation_Error_Has]

Hi abikutn,

Can you go into more detail with what kind of cables you are attaching to your 5761? Are there any cables connected to Ch1 in screenshot "No signal to all of the channels.png" and "only give signal to channel 2.png", or is this channel completely disconnected? Even though you are not providing a signal to the channel, I am wondering if you have anything at all connected to the channel? We can further troubleshoot by swapping the cables of your system around to see if this noise appears on any of the other channels. I also recommend you run the NI 5761 - Low Speed Clock Select example in the NI Example Finder. By running this example, we can determine if this noise is a result of a faulty hardware connection or of a software bug.

Also please note that there is no Ch4 for this device. The channels are Ch0, Ch1, Ch2, and Ch3, however I believe throughout this whole post, you are referring to Ch0 as Ch1.

Regards,

Jason L. 

[abikutn]

we connect a sma to BNC cable to the input of the adc. 

With "no signal", I mean there is not cable connected to certain channel.

Yes,  I have made the Ch0 as Ch1.

I will try the example to see where the problem is coming from.

Thank you!

[abikutn]

Hi 

I have just tested your example program "Low Clock Speed sample". It shows clearly that Ch0 behaves quite different from other 3 channels: When I only connect Ch1 with our square pulse signal, and leave Ch0, Ch2 and Ch3 connecting to nothing.

I have attached the tested results. I think there are some hardware issue with this board.

[A_Donfigulation_Error_Has]

Hi Abhi,

I have taken a look at the schematics on Fig. 2 of this document: http://www.ni.com/pdf/manuals/375509a.pdf , and I see where your concern is, as it looks like there are only 2 ADC's, so at first it seems strange that the values shown on Ch0 and Ch1 are not similar. If you look at the document here: http://www.ti.com/lit/ds/symlink/ads62p49.pdf regarding the actual component, figure 2 indicates that there are actually 2 phsyical ADC's within the chip. 

Measureing a signal that has nothing connected to it is suspectable to electromagnetic fields and other souces of noise to your system, so I would try to physically ground your signal. Even though Ch0 appears to be noisy, it seems to be within the microvolt range, which seems really low. I would try to set the scales of the graphs so that they are not auto-scaled, so you can see the full range of the clipping in channels 1-3. 

Regards,

Jason L. 

[abikutn]

Thank you for your reply!

I have tried to give signal to Ch2 which should be on the second ADC chip, and leave ch0,1,3 unconnected. But Ch0 still showing this kind of noisy result, meanwhile Ch1 and 3 are showing the same as 0.615. Which means the variation or the noise of Ch0 is much larger than Ch1 and Ch3, at least more than 10 times. I do not understand why it showing quite different behavior while they are the same type of ADC chips.

[A_Donfigulation_Error_Has]

HI Abikutn,

I don't think you can make any assumptions on what the values on the channels can be if you just have them disconnected. This is because the channels will become floating voltage sources that are suspectable to different kinds of noise. On top of this, there are also offset errors and gain errors associated with this device. I would begin by providing a ground to these channels to determine if the offest error is within specification and then by providing a known calibrated voltage sources to the channels to assess the gain error. Seeing variation between channels even though they are on the same ADC chip is not unexpected.

Regards,

Jason L.