Post Go back to editing

Recommandations for fixing EMC problems caused by digital isolators ("isoPower" devices family)

Good afternoon,

I am an electronics engineer working for the CEA in Grenoble (France) and I am actually working on the electronics of a product made for medical applications that was designed few years ago.

We need to pass a special standard for Medical Devices called IEC 60601, that's why our product needs to have good EMC performances.

After testing our product in a EMC laboratory, we observed a heavy peak around 350 MHz that makes us out of the EMC template we need to respect for the standard.

After some investigations, we found out that this high frequency peak was possibly due to digital isolators implemented in the main PCB of our product.

Those isolators are Analog Devices products parts of the "isoPower" devices family.

Here are some of the isolators that were implemented in the PCB of our product: ADuM4190; ADuM4160; ADuM2250; ADuM6404

We recently discovered the AN-0971 Application Note which gives some recommendations regarding radiated emissions when using those products.

This note mentions a 360 MHz peak that is appearing in some worst-case scenario. This peak seems very similar to the peak we observed during the EMC test.

 

I would like to know if some users have already had trouble with this family of products regarding EMC.

Do you have any recommandations for reducing this high peak without changing completely our design or totally rerouting our board (for example, adding capacitors or inductances at a certain location or adding a kind of metallic shielding over the components) ?

Feel free to ask me if you have any questions of if you need more information and details about our product.

Thank you for your help.

Sincerely,

Antoine.

Parents
  • Hi Antoine:

    Judge from the noise frequency point, ADuM6404 should be the root for noise.  some application notes have been released to address the radiated emission issue with isopower technology built in, such as adding stitching capacitor or power beads...... But unfortunately i am afraid most of them need PCB layout modification as well. 

    if the design and layout has to be changed, we would rather recommend our new generation isopower products. frequency dithering and special package design is adopted to isolate high frequency noise from isopower circuit to others. the most close part to ADuM6404 is ADuM6424A. The sample schedule is October this year. 

    if you can not wait, we would recommend ADuM6020/28 + ADuM240 solution. ADuM6020 is a stand alone isopower which adopts the same technology with ADuM6424A. 

    Regards

    Chris

Reply
  • Hi Antoine:

    Judge from the noise frequency point, ADuM6404 should be the root for noise.  some application notes have been released to address the radiated emission issue with isopower technology built in, such as adding stitching capacitor or power beads...... But unfortunately i am afraid most of them need PCB layout modification as well. 

    if the design and layout has to be changed, we would rather recommend our new generation isopower products. frequency dithering and special package design is adopted to isolate high frequency noise from isopower circuit to others. the most close part to ADuM6404 is ADuM6424A. The sample schedule is October this year. 

    if you can not wait, we would recommend ADuM6020/28 + ADuM240 solution. ADuM6020 is a stand alone isopower which adopts the same technology with ADuM6424A. 

    Regards

    Chris

Children
No Data