Re: GTA02: NavBoard V3 causing GPS interference
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. Or can it be that the U.FL cable for the GPS antenna module is broken? I know that it is a weak point of the Openmoko antenna module design, but it can be repaired by cutting off the broken end and resoldering. If the cable becomes too short, get a precrimped 100mm U.FL - U.FL jumper cable and cut in the middle. In that case it could be that after mounting everyting, the LNA of the GPS Module is no longer working and you loose 20dB of signal amplification. You can easily check by measuring the DC-voltage at the GPS Module. Or the resistance at the U-FL end of the cable (should read ~50 Ohm). Nikolaus ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
I have found a good reference on how to best handle I2C cabling. It's a PowerPoint Presentation from Texas Instruments, I made a screenshot of the relevant image: http://s7.postimage.org/t9bfk47hn/Screen_Shot_2013_02_08_at_11_00_03_AM.png As I suspected, only one end of the shield gets grounded, individual coaxes for SCL and SDA. Google I2C cable issues.ppt to find the original document as the first hit. Now I have to source some extremely tiny minature coax, ideally with a solid core (for easier soldering to the pads on the NavBoard and the GTA02). Looking for recommendations, ideally something I can order from Digi-Key, Newark or Mouser in smaller quantities (it's not like a need a 1,000 foot roll!). It's not clear IF the power supply wires should be shielded. I think I will try to see if it makes a difference for the GPS. -Pascal ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
Am Friday 08 February 2013 17:10:00 schrieb Pascal Gosselin: I have found a good reference on how to best handle I2C cabling. It's a PowerPoint Presentation from Texas Instruments, I made a screenshot of the relevant image: http://s7.postimage.org/t9bfk47hn/Screen_Shot_2013_02_08_at_11_00_03_AM.png As I suspected, only one end of the shield gets grounded, individual coaxes for SCL and SDA. Google I2C cable issues.ppt to find the original document as the first hit. Now I have to source some extremely tiny minature coax, ideally with a solid core (for easier soldering to the pads on the NavBoard and the GTA02). Looking for recommendations, ideally something I can order from Digi-Key, Newark or Mouser in smaller quantities (it's not like a need a 1,000 foot roll!). It's not clear IF the power supply wires should be shielded. I think I will try to see if it makes a difference for the GPS. -Pascal You should use fine pitch flat ribbon connector or flat plastic connector like on LCM, and shield it with a metal braid sticky tape, again like originally done in GTA02 LCM connection. For further crosstalk and EMI shielding run a GND wire as every 1., 3. aso wire in your ribbon cable. Ulrathin coax is *extremely* expensive and hard to source /j ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
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. BR, Nikolaus ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
GTA02: NavBoard V3 causing GPS interference
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. -Pascal http://www.wi-flight.net/ ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
A few comments inside: Pascal Gosselin píše v Čt 07. 02. 2013 v 08:36 -0500: 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 We are usually trying to refrain from connecting shileding to ground on both sides as a prevention of grounding loop. . 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. We have never twisted two differnet signals together. What about cross-talks? The only type of signal we are twisting together are for example positive and negative leg of RS422 bus. Or A429 (you may be familliar with). 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. -Pascal http://www.wi-flight.net/ Zbynek ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
Dne 7.2.2013 18:00, Hrabosh napsal(a): A few comments inside: Pascal Gosselin píše v Čt 07. 02. 2013 v 08:36 -0500: 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 We are usually trying to refrain from connecting shileding to ground on both sides as a prevention of grounding loop. . 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. We have never twisted two differnet signals together. What about cross-talks? The only type of signal we are twisting together are for example positive and negative leg of RS422 bus. Or A429 (you may be familliar with). Just ground betwen all data lines is common solution to this. Like GND|DATA|GND|CLK||GND Emited radiation is equal of surface of closed loop betwen GND and signal line. 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. -Pascal http://www.wi-flight.net/ Zbynek ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: GTA02: NavBoard V3 causing GPS interference
Jiří Pinkava píše v Čt 07. 02. 2013 v 18:18 +0100: Dne 7.2.2013 18:00, Hrabosh napsal(a): A few comments inside: Pascal Gosselin píše v Čt 07. 02. 2013 v 08:36 -0500: 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 We are usually trying to refrain from connecting shileding to ground on both sides as a prevention of grounding loop. . 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. We have never twisted two differnet signals together. What about cross-talks? The only type of signal we are twisting together are for example positive and negative leg of RS422 bus. Or A429 (you may be familliar with). Just ground betwen all data lines is common solution to this. Like GND|DATA|GND|CLK||GND Emited radiation is equal of surface of closed loop betwen GND and signal line. Yes, this is quite common for ribbon cables (like the one for connecting ATA100 hard drives for example). But shielded TWISTED PAIR was mentioned originally... 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. -Pascal http://www.wi-flight.net/ Zbynek ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community ___ Openmoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
