GTA02: NavBoard V3 causing GPS interference
Dr. H. Nikolaus Schaller
hns at goldelico.com
Fri Feb 8 20:50:14 CET 2013
Am 07.02.2013 um 14:36 schrieb Pascal Gosselin:
> Yesterday I was able to confirm that the installation of the Golden Delicious NavBoard V3 very seriously degrades the performance of GPS when using the built-in GPS antenna. Getting a 3D fix is taking over 20 minutes on average using the built-in GPS antenna. When using an external GPS antenna, the issue goes away and an unaided GPS cold start can be achieved in about 41 to 44 seconds.
> I suspect that the use of unshielded cabling between the GTA02 is a likely cause. Being in the Avionics business, I am familiar with the use of shielded twisted pair wiring, but generally in large 22AWG (will have to figure out where to source very thin STP cabling). The wires we use to perform this mod is solid core, probably 30 AWG. In avionics, when it is desired to keep a signal from radiating outside a cable (such as Headphone/Microphone wiring), then the shield of only a single side of the cable is terminated to a good grounding point. When instead it is desired that the signal inside a cable be protected from EMI/RFI from the "outside", then the shields on both sides of the cable will be terminated to ground.
> I theorize that only the SCL/SDA would need to be shielded (together in a shielded twisted pair cable ?), Power and Ground may not need to be shielded.
> This issue seems very similar to the SD Card access/GPS issue of the earlier GTA02s, so I also wonder if a capacitor fix should be looked at in this case as well.
I am not aware that others have had such a problem but different cabeling may lead to different issues.
The key problem with the SD card access is/was that the SD/MMC interface runs at exactly 25.000 MHz and the clock signal switches fast. And 63 * 25.000 MHz = 1575 MHz which is the center frequency of GPS. I.e. the 63th harmonic of the SD interface covers the very weak GPS signals in noise... Damping this SD clock by a 10pF capacitor reduces noise significantly.
For the Navigation board the communication is running through I2C. But I don't know the I2C clock frequency. It may be 400 kHz which would mean the 3937th harmonic which is quite unlikely to be the reason (because energy of harmonics goes down by the number).
So I rather would assume that the SD clock GND (or VCC) noise or something similar is brought too close to the GPS antenna.
Maybe Christoph has some additional hints.
More information about the community