AMC1301: Sigma Delta converters on TIDA-1606: Data Problems from the AMC1301 for grid current measurement.

Hi Team,

A gentle follow-up on this please? Thank you!



Kind Regards,

Jejomar

Hi Jejomar,

have you already carried out scope measurements to find the cause?

Kai

Hello Kai,

I observed the real current and voltage via oscilloscope as well as the clock input and data output signals of the SD converter (at the position of the SD converter) and found no problems.
But if I try to observe the clock output and data input signals at the F28379d controller card, I can see disturbances. However I can see disturbances even if I hold the probe slightly above the associated pins.
That means there is a lot of electromagnetic field coupling into my voltage probe. Note, that the controller card is located nearly to the gate driver cards.

My conclusion regarding to this measurement is that I just measure EMC disturbances in my probe not on the data or clock line of the SD converter. Do you agree with this?


Kind Regards,

Jejomar

Hi Jejomar,

yes, this sounds reasonable.

A good test is to use such a ground spring and to probe with both pins of this modified probe signal ground on the printed circuit board. This can be a ground via in the ground plane or the ground terminal of a decoupling cap. With both pins of the probe being connected to signal ground, no signal should be seen in the scope. Anything else is injected EMI and is an artefact.

Eventually, you might want to add some shielding between the controller card and the driver card? A sheet of metal or FR4 with a copper plane connected to a proper signal ground could do the trick.

Kai

Hi Kai,

Apologies for the delay. I tested some shielding on different places on the board:
1. I surrounded all three gate drivers with an isolated, grounded copper plane
2. I surrounded the control card with the grounded copper plane

In both cases I could not see any effects to the controller cards internal measurement. To verify if the copper planes has a generally effect, I measured the radiated field via unconnected voltage probe and I have seen, that the shielding was working in a notable manner.

My conclusion regarding to this measurement is that the radiated field is not a problem for this measurement.

But with further measurements via oscilloscope I recognized a remarkable effect:

I was trying to measure the voltage drop across the shunt resistor of the AMC sigma-delta converter and was a bit careless because I connected the ground of my scope directly to the phase output of the device. But fortunately there was no short circuit (I think the DC sources are isolated to the lab mains).

But with that I saw that the problem was solved. I could no longer see disturbances in the controller board data when I connected the scope's ground to the unit's output phase. Now, I'm a bit confused because I'm not sure how to interpret this measurement. Do you have any idea?

For a better understanding, please see the attached overview picture.




Kind Regards,

Jejomar

Hi Jejomar,

do you have a connection between "GND_AMC1" and "Vinv_A", as recommended in the datasheet of AMC1306?

Is 5V_1 stable and noise-free?

Kai