I concur and I have done a slightly different GPS mod to my Solo. (I actually did this last November)
The datasheets for the u-blox 7N module refers to the backup battery that maintains the Almanac, Ephemeris
and time during power off, discharge current is typically 15uA. Settings can be saved in internal Flash EEPROM (or so all the product web
sites say), but not Almanac & Ephemeris.
https://www.u-blox.com/sites/default/files/products/documents/MAX7-NEO7_HIM_(UBX-13003704).pdf
https://www.u-blox.com/sites/default/files/products/documents/NEO-7_DataSheet_(UBX-13003830).pdf
That battery on the Solo is a MS621FE, which is the cell recommended by u-blox.
http://www.sii.co.jp/en/me/files/2014/11/BAC3011EJ-01C1409_web__.pdf
On page 6 of the battery datasheet we see:
voltage 3
capacity 5.5mAh
charge/discharge current 15uA (although later on the graphs indicate you can charge it at 100uA)
cycle life at 100% DOD 100 (Depth Of Discharge)
cycle life at 20% DOD 1000
When I disassembled my Solo after a week of intermittent use and measured the MS621FE voltage, it was 1.4volts,
indicating that it was very close to fully discharged, see page 7, 8.
Page 7 indicates a charge current of .1ma to 3.3v for 96 hours will result in the capacity being maximized, and results
in about 1% charge gain per hour. The u-blox datasheets indicate 1.6V minium for data retention.
But, to charge for 96 hours one would have to have the Solo powered on (flying or not) for that whole time.
If you have only one battery and at 2 hours per charge and don't fly at night, you MIGHT get 1-2 hours per day of charging time.
With 4 batteries, you might be able to get 12 hours of charging time, and at about 1% charge gain per hour (and .3% per hour charge
loss) you might actually gain some charge over the course of a day! That's if you fly (or power on) continuously for 12 hours straight.
Of course you don't need really need a fully charged backup cell.
At 15ua discharge current a fully charged battery MIGHT last 366 hours, or about 2 weeks, that's about .3% charge loss per hour.
(The Almanac and Ephemermis decay in validity as well during this time so the GPS would still need to
re-download some of each after extended power off time). Time, once off by > 20 seconds, makes sat acquisition difficult.
Consider this, if the Solo spends 2 weeks in the supply chain after manufacture, it's backup battery will be pretty dead by the time
it gets to its buyer, and won't be able to reach full charge unless extraordinary measures are taken. Partial charge will
effect cycle life and be terrible if you can ever get it recharged enough to matter.
And that assumes the the battery is fully charged when it is installed in the factory.
This MAY be why new owners see 3-4-5 minutes to get enough sats to fly.
The result of this is that the GPS module will loose its time/Almanac/Epheremis and do a cold start almost every time
the Solo starts up. (1.6 volts is at the minimum necessary to maintain SRAM integrity)
TTFF for the u-blox 7 is specified as 28 and 29 seconds for warm and cold starts with 1 second for hot start. This is
for the first 3D fix of 4 sats. Solo wants >=7 and a HDOP of <2 so TTFF should not be relied on to indicate how long
it will take for the Solo to be ready to fly. In addition, these factory tests were performed with the GPS signal provided by a simulator,
so no interference, and signal levels all at -130dbm which is certainly not reality. I've had 9 sats sitting for 10 minutes and not be
ready to fly. Why? the HDOP was 4 due to.... whatever... bad geometry, high Kp, phase of the moon...
I have modded my Solo and replaced the MS621FE with a non-rechargable CR2032, also removing the diode from the charge path so there is no possibility of charging that cell. Which might result in "Bad Things(tm)" happening.
http://data.energizer.com/PDFs/cr2032.pdf
With a capacity of 240mAh, at a 15uA discharge it should see a lifetime of 1.5-2 years.
This is "engineering accuracy" and only pretends to be at best +/- 20%, but is certainly good enough for this seat of the pants analysis.
At the same time I did this mod I noticed that the copper foil shield was very close to/touching the antenna/lna components. I took
this as an opportunity to add a small plastic box around those components as shown in the photos "just in case". Whether or not this
is a cause of problems remains to be seen.
My Solo came from Amazon with 2 day shipping, but was the last one they had for almost a week, so who knows how long it was in
transit from the factory in China.
u-blox specifies 28/29 seconds for a warm/cold start, having the Almanac/
Ephermis/time has made a difference to my modded Solo.
I believe that 3dr made a mistake in designing the GPS in line with the u-blox design guide, which was intended for "always on" applications where the small battery will not be discharged too much and will have ample time to recharge during normal operation.
>>>The Solo is NOT the same use case. Especially considering what initial shipping times from China might be.<<<
My u-blox 6 & 7 modules all have the same battery as the Solo and they are all always dead. The u-blox 6 can store
the data in an eeprom, unfortunately the u-blox 7M can't and has to rely on the backup battery. The u-blox 7N used in the
Solo can store some data in eeprom, but it can't keep time when the battery is dead, or do a warm start.
IMHO it would be better to have a 1-2year non-rechargable battery that guarantees Time/Almanac/Ephermis are retained during long power down times (like shipping from China), (the Almanac may still be ok, but the Epheremeris won't) rather than one that will discharge during shipping (or idle time) and never be able to be fully charged again. My tests seem to prove this out.
The ephemeris and clocks parameters are usually updated every two hours, while the almanac is updated at least every six days.