Just curious if anyone has ever removed and soldered in a new GPS module battery? Wow they are small, I think I'll need smaller tips, any tips on fitting except being careful.
GPS battery replacement
- Thread starter Mrgs1
- Start date
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Easy to do if needed. But why do you need to replace it? It should last a way longer than the oldest Typhoons are at the moment. Does the GPS do a cold start every time you start up (although with NEO M8 it should be around 30 seconds)?
Doesn't need replacing, just asking in preparation. Some have had to replace module when they've had a failure, battery is a lot cheaper than module in UK.
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Ah, OK. ;-) It is hard to tell about the lifespan of the battery there, but usually that kind of memory back up batteries last around 5-10 years. So plenty of time to get that sharp-tipped soldering iron. On the other hand; you might need it sooner for other little soldering excercise, too. Soldering few wires to the small soldering pads on the mainboard... You know what I mean! ;-)
However, I actually do not even know, what happens if that battery drains on the NEO M8. Very likely not a module failure; I've never seen a GNSS module fail without a battery backup. In that case, every startup will be a slow cold start. The hardware integration manual says the following:
If the module supply has a power failure, the V_BCKP pin supplies the real-time clock (RTC) and battery backed RAM (BBR). Use of valid time and the GNSS orbit data at start up will improve the GNSS performance, as with hot starts, warm starts, AssistNow Autonomous and AssistNow Offline. If no backup battery is connected, the module performs a cold start at power up.
(https://www.u-blox.com/en/docs/UBX-15029985)
Older chipsets, such as Sirfstars, had very long cold start times, ~12.5 minutes, that it took to receive orbit data from the satellites before any positioning solution could be calculated at all. This is not the case with the Typhoon's GPS, since start-up times for the NEO M8, according to the datasheet (https://www.u-blox.com/en/docs/UBX-15031086) are the following:
Last edited:
Always give it about 5 minutes or more before takeoff, EU DFS slows it down anyway 90 seconds or more.
I have topshells with module in ready to go, I'm not sure the copter would be adversely affected by a non charging battery.
I have topshells with module in ready to go, I'm not sure the copter would be adversely affected by a non charging battery.
The battery voltage is easy to check if you have the top shell removed.
The compass is also located on that module. I have always assumed the battery was providing power to maintain the compass calibration and the almanac data. But that data is only good for 24 hrs, so it takes longer to get a good GPS fix if the H hasn't been flown in a couple of days. Waiting on the first flight of the day is always a good practice.
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Hi Steve,
That top shell is a pain to remove, those plastic clips are good for only few mandatory removals.. ;-)
The compass calibration data is saved on the STM32's FLASH memory, as there is no user-accessible memory, volatile or non-volatile, on the magnetometers (HMC5338 or IST8310 in later models). The magnetometers are not connected to the GPS either, they are on their own I2C bus, while the GPS is on UART, independent of each other. Waiting is required (on stock firmware ;-) only for that long that there are 10-20 satellites "acquired" (Navstar + Glonass at least, I have not yet taken a look into the NEO M8 settings if other constellations are enabled there), that gives you a very good navigation and positioning performance regardless when the drone was last flown. ;-)
Br,
TR
Last edited:
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
5 minutes is really not needed for the GPS. 10-20 satellites and GPS Ready status: you are good to go. The CGO3+ on EU FW will take it's time, but there is no reason to wait for 5 minutes only for the GPS if you have coordinates and 10-20 sats (or good HDOP value and home position set in that other firmware... ;-)
Turning on ST16 first, putting on props and checking and making sure everything looks OK, inserting battery, turning on copter, and waiting for camera to connect, 5 minutes is no big deal, especially if it's not been used for a while, everyone has their own regime as they say.
You should just pick up a spare GPS sensor. I have two. I purchased them both from eBay for 10 bucks each.
I already have spare top shells complete with modules. It was about sharing tips if anyone had any.
Last edited:
@Mrgs1,
Good point, but there IS a but here...
In your example, 5 minutes from ST16 power on is, agreed, no big deal. But... 5 minutes after powering on the aircraft itself CAN seem like a long, long time to wait.
It is the latter that is most important! Getting the aircraft ready, including GPS data, is more important than the ST16. Especially if the aircraft has been idle for several days, as has been discussed often, it is highly recommended to allow the aircraft to sit, powered on, but motors off, for the 12-15 minutes GPS almanac refresh.
(Repeating here in anticipation of any readers who may not have previously read about the almanac refresh recommendations.)
That make sense?
Respectfully,
Jeff
Last edited:
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Then there is something very wrong with the GPS unit of the device if Time To First Fix (TTFF) takes 45 minutes on a modern uBlox GNSS chipset. Replacement is required here. Please refer to the datasheets above regarding the uBlox' chipset performance. ;-)
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Umm.. what on the Earth is this "GPS almanac refresh"? I've seen this before in this forum and I would like to know, where this thing comes from. ;-) I have never heard of this term or the need to wait for something if there is a fix and a good positioning solution, outside this forum. Please refer to uBlox datasheets and HW integration manuals I posted above, there is absolutely no reference of anything like that. I've actually never seen that in any GNSS receiver chipset or system manual or datasheet, real aircraft systems included, or taken this kind of behavior into account in any GNSS-enabled systems I've been using, flying, designing or involved with. Not described in Yuneec's manuals, either which I'd take as an official statement, how to operate devices manufactured by them. ;-)
What are those almanacs actually for? Please take a look at these links:
Here you can find some info what the "GPS almanac" is:
GNSS Ephemerides and Almanacs
NovAtel’s OEM GPS receivers, antennas, GPS+Inertial and software products lead the GNSS industry in reliability, innovation and support.
www.novatel.com
Here are parameters for the Galileo almanac (each GNSS constellation has this kind of data, the Navstar's data is similar):
...and here is the actual data for the Galileo if you are interested in what is actually there (I chose the Galileo as an example, as this data was easily available ;-):
The almanac data essentially is used for the first fix. You have absolutely no means of controlling what happens regarding this data inside the GNSS chipset, when it is refreshed and how, only indication of the validity of this data, is that you have a GNSS position available on your device. By waiting 12-15 minutes for "refresh", you are basically waiting 12-15 minutes for nothing.
Don't worry about things like this, the GNSS chipset takes care of all the low-level operation of the GNSS and if everything is OK, it gives you a position in return. Don't wait, go flying and have fun! ;-)
Br,
TR
Last edited:
rdonson
Premium Pilot
How To: Usage Compilation — Calibrations - Advanced Functions
Here are some good videos on performing the calibrations. Feel free to add your favorite videos on this topic. Notes: - conventional wisdom suggests removing the camera before compass calibration. It saves the gimbal a lot of stress. - Although not a calibration, conventional wisdom is to...
yuneecpilots.com
PatR
Premium Pilot
Pollo,
I’ll suggest to take a short break from toy grade drone forums to visit the world of full scale aviation and avionics. While there, research how much time will pass GPS manufactures make reference to new installation boot up time and positional accuracy. They also explain why.
Understand all their systems are FAA approved and certificated, and must be installed and tested by licensed avionics technicians and avionics service facilities. Recognize they are describing the best equipment under the best conditions, a couple things our toy GPS units are bot afforded.
Hint: you will read that some units will require up to an hour of power on time before they become usable. I can assure you that what is used in multirotors is not nearly well made, sophisticated, or accurate, and if allowed to sit powered down for extended periods of time forget where they are.
- Joined
- Jan 20, 2019
- Messages
- 183
- Reaction score
- 143
- Age
- 41
Been there. CPL(A) with instrument rating and multi-engine rating, fATPL, MCC, type rating for a high-performance twin-engine jet, having flown hang gliders, powered ones, sailplanes, weight-shift microlights, SEP's, ultralights and an Embraer jet since 2005. And pretty exotic drones, too.
Units that require up to an hour to power up are vintage mechanical INS devices in particular conditions. I think I've seen one in museum. I would not call them that sophisticated in 2020.
Last edited:
Similar threads
