Apalis TK1 EMC (Electro Magnetic Compliance)

Hi

Our systems Uses Apalis TK1 On IXORA Carrier as platform.
We now comply to EMC standard EN55014-1 with the whole system .
During test we have a “Fail” ( ~6DB above threshold) frequency of 200MHz
radiated from Apalis + IXORA.

We added ferrites to cables hooked to board and masked the boards with conductive foils.
It improved the emission, still not pass !

Any idea of the exact source of the 200MHz ?
Any suggestions how to suppress it locally ?
What are the boards clock oscillators ? ( 8 MHz, 12MHz, 25MHz ?) any others ?

Please advise
Thanks
Oded

Any idea of the exact source of the 200MHz?

Probably SD card.

Any suggestions how to suppress it locally?

If you do indeed have an SD card then I would start removing that one or one could also limit its frequency/speed in software e.g. here:

http://git.toradex.com/cgit/linux-toradex.git/tree/arch/arm/mach-tegra/board-apalis-tk1-sdhci.c?h=toradex_tk1_l4t_r21.6#n105

Limit SD card frequency/speed. Drive strength may also be tuned:

http://git.toradex.com/cgit/linux-toradex.git/tree/arch/arm/boot/dts/tegra124-platforms/tegra124-apalis-pinmux.dtsi?h=toradex_tk1_l4t_r21.6#n1483

Details may be found in the TK1 datasheet or TRM resp.

What are the boards clock oscillators ? ( 8 MHz, 12MHz, 25MHz ?) any others ?

12 MHz OSC1 for NVIDIA TK1 SoC

25 MHz OSC2 for Intel i210 gigabit MACPHY

32.768 kHz OSC3 for ASM AS3722 PMIC

8 MHz OSC4 for NXP K20 companion MCU

Thanks Peter,

I will try testing without the SD card tomorrow .
Hope it will help.
Thanks Again

Oded

Another question ,
If the SD card is installed but I do not read/write to it
will it still be quite ( with regards to the 200MHz)
or must I physically take the SD out of the IXORA ?
Thanks
Oded

If the SD card is only plugged in, there is still an idle clock as far as I know. I do not know whether this would also create an interference, since we have tested it only under a load. We were running our tests while doing a continuous read and writes to the card.

Hi Oded,

We have tested the TK1 in combination with the Ixora a couple of month ago in the EMC lab. We have also seen a peak at 200MHz. We have been able to reduce that peak by adding cable ferrites and were passing the EMC limits. In our case, the frequency was coming from the SD card interface. In order to confirm that, we swapped the card with a lower speed grade card (a card without UHS-I). I recommend you to do the same (or do a test without an SD card). In our case, the 200MHz peak disappeared.

Since we identified that the SD card interface could cause issues, we are going to do some improvements in the next revision of the Ixora carrier board. We will improve the decoupling of the card power supply. We will also change the cable shield circuit in order to reduce the coupling into the cables. However, we do not know whether these counter measurements will improve situation until we have tested the new boards. Unfortunately, I do not know when we will start the PCN and the new version of the Ixora is available.