IEC 61000-4-6 Immunity to conducted disturbances - CDN questions

[AndersJ]

I am having problems with a digital input.
It connects with a 2 conductor cable.
One is signal ground.
The other one is the active high input.

The test is done by injecting 10 Vrms common mode RF with a CDN.
The RF is amplitude modulated, 80% at 1 KHz.
Since the signal input is common mode there should be no
differential voltage between the two conductors.
BUT there IS a differential voltage.

My circuit is unbalanced, thus requiring a CDN-AF2 for unscreened non balanced lines.
I suspect that the testhouse is using the wrong CDN,
a CDN-T2 intended for a BALANCED pair.

My question therefore is:
What output signal will a CDN for balanced lines create in the unbalanced scenario described above?

Many thanks for your comments.
Anders J

[joeqsmith]

You know, reading your post I don't think I have ever sent something to a lab to be tested without being present.   

[Helix70]

As the CDN-T2 has two anti-phase transformers in the HF path, this will cause line A to be disturbed in the opposite polarity to line B, referenced to ground. That is for a given pulse, A will be driven positive while B will be driven negative (a balanced signal) with respect to the ground.

By contrast, the CDN-AF2 omits the anti-phase transformers, so the disturbances will be driven in the same polarity with respect to ground, that is, in common mode.

If they are connecting your ground wire to ground, and the other to A (or B), then not a lot of difference will occur between the T2 and the AF2 i wouldn't think.

If they are connecting to A & B without ground, you will see differential voltages due to external capacitive coupling to ground and the series inductance of the cables. Is your circuit externally earthed?

[AndersJ]

Thanks Helix70 for commenting on my issue.
I'm not sure how to interpret your last 3 sentences.
A rephrasing would be much appreciated.

Your comments confirm my initial suspicions.
As an experiment i "built" my own CDN-T2,
and it DOES create a unbalanced ground referenced signal.

I am pretty sure this explains why the test house managed to
fail my device by creating a differential signal input where it
should have been common mode, which my input would have ignored.

[T3sl4co1l]

It sounds like your input/termination impedances are unequal, therefore a common mode current (present on both conductors simultaneously) will result in a nonzero differential voltage, as seen by your circuit.

Heavy filtering on external inputs goes a long way towards preventing these sorts of problems.

What does your input circuit look like?

Tim

[Helix70]

If the lab connected your signal wire to "A" and your ground wire to "Ground" on the CDN-T2, it will not be very different to the CDN-AF.

If the lab connected your signal wire to "A" and your ground wire to "B", then you will see differential voltage.

Do you have the test report?