I read a lot of issues about losing the gps signal due to interference of some kind. Happened to me once. It never happens with the gps signal in my Iphone 6s.
Is the quality of the gps module in the spark not good enough?

The main reason is this gps in your drone needs to be in direct line of sight to the sky to receive enough signals to fly.
A cell phone is basically a sophisticated two-way radio. Towers and base stations, arranged into a network of cells, send and receive radio signals. Cell phones contain low-power transmitters that let them communicate with the nearest tower thus signal from many different sources.

hallmark007 Posted at 2017-12-25 16:12
The main reason is this gps in your drone needs to be in direct line of sight to the sky to receive enough signals to fly.
A cell phone is basically a sophisticated two-way radio. Towers and base stations, arranged into a network of cells, send and receive radio signals. Cell phones contain low-power transmitters that let them communicate with the nearest tower thus signal from many different sources.

Thanks for the explanation!

hallmark007 Posted at 2017-12-25 16:12
The main reason is this gps in your drone needs to be in direct line of sight to the sky to receive enough signals to fly.
A cell phone is basically a sophisticated two-way radio. Towers and base stations, arranged into a network of cells, send and receive radio signals. Cell phones contain low-power transmitters that let them communicate with the nearest tower thus signal from many different sources.

I already experienced some GPS signal loss in middle of nowhere, no obstacles of anykind...
Hopefully they where either short, or I could bring the Spark in ATTI mode.

My "best guess" is that there is some INTERNAL source of interferences that could sometime trigger and mess up with GPS .
GPS L1 freq in in the 1.575 GHz range, so messing up signal should probably be "bad harmonics" from another microwave source.

In the Spark, it could be : the 2.4 GHz Wifi, or the internal Movidius vision chip that also features a USB 3 PHY which operates in the 2.4GHz band.

Last thing would be bad GNSS antenna that "self jam". A GNSS antenna is not something "passive" : there is an amplifier inside (call LNA : low noise amplifier). In some cases (bad design or bad manufactuing), a part of the amplified signal could be re-emitted though the antenna and become a jamming source right in the GNSS L1 band.

I already faced both first cases (Wifi and USB 3 short range GNSS jamming) in my professionnal environment. the jammed receivers where high end geodetic receivers.
I never met the "self jamming" issue, but I met the issue where a low end GNSS device was short range jamming a high end device.

It would be interesting to take note when those issue occurs if the RC to AC link was in 2.4GHz or 5.8 GHz mode (which should be less likely to have "strong" leak signal in the 1.575 GHz band)

For flight safety a higher level of GPS accuracy is required than is needed for a phone. Therefore the Spark needs to see satellites spread across the sky, the more spread the better as the circle of error gets smaller the more spread the view is.

A phone is also taking a reference from the GSM signal from the mast so it can compare the two sources, and it does not have to be that accurate for most phone purposes anyway. Some phone map apps show the circle of error as a circle around the estimated location. You could actually be anyway in that circle, being "anywhere" is not good enough for flight.

Matioupi Posted at 2017-12-26 02:52
I already experienced some GPS signal loss in middle of nowhere, no obstacles of anykind...
Hopefully they where either short, or I could bring the Spark in ATTI mode.

Gosh, this sounds very complicated. Thank you for trying to explain however.

Thank you all for the replies!

As others have mentioned, your Spark needs more satellites “locked” in order to accurately pinpoint it’s position. It uses a concept called HDOP (Horizontal Dilution Of Precision) to determine how accurate the location is. If the HDOP number is too high the quad will “complain” about it whereas the phone, not needing the extreme precision, will not. If your phone shows the location 50ft off most people would never even know it as that would be more than good enough to find the road you need to turn on. One other item that affects the GPS reception of the Spark compared to a phone is the Spark, due to having large motors and speed controllers etc, creates much more radio/magnetic interference making the relatively weak GPS signals harder to receive when compared to a phone.

In my experience of building mini quads with gps. the biggest problem is electro magnetic interference impacting on the magnetometer. We always mount the GPS antenna and magnetometer as far offboard as possible (normally on a mast) I have tried all sorts of shielding, but non ever worked as well as a simple mast. My  gut would say that everything in the spark is to close together in its little body. wouldn't be hard for DJI to put it on a retractable mast for V2. Blade went that way with the qx350 v3 and the Chroma, they even increased the size of the GPS antenna on the later Chroma.
I think form is more important to DJI than function !
I haven't had any problems with my spark so far, but I'm very careful to head any on screen warnings. I don't feel as confidant with the spark as I did with my Chroma.

Matioupi Posted at 2017-12-26 02:52
I already experienced some GPS signal loss in middle of nowhere, no obstacles of anykind...
Hopefully they where either short, or I could bring the Spark in ATTI mode.

That's some good to know technical details, thank you for laying that out.

Hello,
Last Saturday I've managed to crash my Spark and now, it doesn't fly anymore. I get "check redundancy state" error everytime I power it on. I've red some topics that said it might be the GPS module broken(i have a hit right in that area) and I have opened the drone and saw that the module is intact but 2 of the capacitors of the board, are broken. Anyone knows how can i get a datasheet or a schematic of the GPS module? Sorry for my english and thank you in advance.

sunete Posted at 3-11 02:47
Hello,
Last Saturday I've managed to crash my Spark and now, it doesn't fly anymore. I get "check redundancy state" error everytime I power it on. I've red some topics that said it might be the GPS module broken(i have a hit right in that area) and I have opened the drone and saw that the module is intact but 2 of the capacitors of the board, are broken. Anyone knows how can i get a datasheet or a schematic of the GPS module? Sorry for my english and thank you in advance.
[view_image]

Hi sunete,

How do you know these 2 caps are bad? These are multi layer film capacitors, which do not leak since they have no fluid electrolyte in them. The yellow colour can be anything, coating or such.
However, these ML-caps are prone to shorting when microcracks form inside the cap because of mechanical stress.
No way to get datasheets. You can desolder them from a known good board and measure the capacity, plus the voltage they see (when unit is workin), plus some generous reserve.

HTH,
Wolferl

Hi Wolferl,

Thank you very much for your answer. The photo attached is not made by me, I only used it for refference compared to my board, where those 2 caps are missing(one is broken and one is not on the spot. I thought I might get the third one off the board and measure it, because it looks the same as the 2 with problems, but first i said that i might try my luck for the datasheet online :p.
OK, thanks for your answer.
BR,
sunete

Hello everyone,

The 2 caps turns out to be coils. I managed to put others instead and guess what.. it works like a charm My spark it's alive!

sunete Posted at 2019-3-12 11:55
Hello everyone,

The 2 caps turns out to be coils. I managed to put others instead and guess what.. it works like a charm  My spark it's alive!

Hello, I have a spark with the same problem.
I tried another gps module and it works so I'm sure it's a problem with that board.
I already changed the ceramic antenna but it still has the problem of redundancy.
What value are the inductors that you changed? I measured them with a multimeter and they are not open, I would have to take them out.
Could you tell me a little more about your repair?

Thank you very much.