AD2 Scope crosstalk

[attila]

I've got the following question:

Something popped up yesterday that I would like your input on.  I had signals from two antennas going into the two analog channels of the AD2.  One signal was 10 mV p-p at 1.4 MHz and the other was 500 mV p-p at 1.4 MHz.  Quite a difference in signal strength between the two channels.  I noticed that when I dropped the strength of the 500 mV channel the other one dropped as well, not as much but certainly noticeable.  It appeared that the stronger channel was feeding the other channel until I dropped is strength.  I swapped channels and cables and moved everything around physically but got the same results.  When I fed the two signals into the two channels of my scope it did not happen.  No cross-talk.

Is this something that you have seen before?  Is my AD2 unit defective?

I would appreciate your input.  Thanks.

[attila]

The crosstalk is normal effect, since wires act as antennas.
Shielded and/or differential twisted wires are used to reduce this effect.
Try to post some screenshots and photos from your setup.

Here Scope channel 1 receives 4V pk2pk 1MHz signal from Wavegen and channel 2 wires are floating:
Similar effect could be with long wires too.

Here Scope channel 2 input is shorted with a jumper, no noticeable cross-talk in time view:
The FFT indicates about 13dBV peak on the noise floor, which is on device cross-talk.

Here Scope channel 2 is connected to Wavegen and channel 1 input is shorted:
The FFT shows about 5dB crosstalk.
The 5 and 14 dbV difference is due to the Scope channel 2 being closer to Wavegen than channel 1.

Feeding signal from external device to the AD2 shows no crosstalk between the device Scope channels: 

Crosstalak with BNC adapter board is about 4dbV, channel 2 is shorted in the BNC plug:

Crosstalk with BNC probe is about 32dbV, channel 2 is shorted on the probe cable tip:

[duncan crane]

Thanks for the detailed response Atilla.  I duplicated your tests to make sure I understood what you were saying and to verify my AD2 is working properly.  Screen shots attached.

I used an external signal generator for all my tests rather than the built-in signal generator in Wavegen.  You did not get significant crosstalk using an external signal generator but I did.  Shorting out the other channel helped a little when I was NOT using the BNC adaptor and it really helped a lot when I WAS using the adaptor.  So our test results were similar.

It is interesting to note that when feeding these test signals or other signals coming in off my two antennas directly into my old analog oscilloscope there is never any hint of crosstalk.

So bottom line is yes, my device is working as advertised; and yes there will be crosstalk occurring within the AD2 device.  I do not believe any of the crosstalk I am seeing is coming from external wires or cables.

My particular application is all about phase difference between the two signals I am bringing in on the 2 analog channels of the AD2.  Both signals will be the same frequency and at times one or both of them may be as strong as a volt or more.  Any crosstalk will throw off my measurements of phase difference in a big way.

As I think about it, the only way that I can go forward using the AD2 would be to purchase a second one and run my signals through separate AD2 devices.  Some questions:

1. Will I be able to display all four channels in a single Scope window?

2. Can I click a Single sampling and get readings on all four channels, synchronized readings?

3. Is it reasonable to assume that this approach will eliminate any AD2 device originated crosstalk?

Thanks for your excellent support Atilla!

Duncan

Crosstalk.docx

[attila]

Hi @duncan crane

In step 3 I suppose you only have shorted the 2+ to GND and the 2- pin behaves like an antenna.

The scope inputs are differential, so the +/- should be shored.
The inputs are differential but not floating, so a ground connection between the devices/circuit is also needed.

Could you post a photo from your setup?

 

 

1,2,3. One WaveForms application can control one device at a time. Separate applications are needed for each device.
The devices have independent oscillators and especially at high frequency analysis the sampling would show jitter.
 

[duncan crane]

1. Here is my setup using the BNC adaptor and the external signal generator feeding channel 1 with nothing on channel 2.

2. Looking at WaveForms Scope I see 4 V p2p, at ~1 MHz on channel 1...

3. ... and due to crosstalk I see 50 mV p2p on channel 2.

4. Now I attach the short on channel 2 ...

5. And the crosstalk goes way down.

6. The testing with and without the short on channel 2 is informative and demonstrates well the existence of the crosstalk.  But when I use the AD2 in my application I have signals coming in on channel 1 and 2 so of course I cannot short one out.  You have answered my questions regarding the existence of crosstalk, as well as the possible option of using a second AD2 to eliminate crosstalk, which apparently is not an option for me because I would no longer see both signals in one  WaveForm Scope window and therefore wouldn't have a synchronized view of both signals. 

So, I am going to see if there is some practical maximum level I can attenuate my signals down to that will reduce crosstalk to a level where it is no longer an issue.  That might be 100 mV, maybe 50 mV, I shall see what I come up with.  If I can do that on both channels without loosing the signals true phase shift I am in business.  Stay tuned!