GPS M10 still poor performace in BF4.5 RC3

[radiant meteor]

Poor performace of GPS in BF4.5 RC3. After 30 minutes the satellite acquired is still around 0-5, and in the only time it has a fix, the fix is maintained for only 30 seconds.

Test setup:

1. GPS: HGLRC Mini M100 GPS Module, correct wiring as it shows up in the GPS tab
2. GPS module under clear sky for 30 minutes
3. No interference from VTX (at 1mW)
4. Test repeated for 3 times

After 30 minutes, the average satellite is around 3.

Is it still a Betaflight config issue or module issue?

Thanks

[radiant meteor]

My settings below:

set gps_sbas_mode = AUTO
set gps_ublox_use_galileo = ON
set gps_set_home_point_once = ON
set gps_rescue_allow_arming_without_fix = ON

[woven plover]

Disable SBAS by setting to off.

Most likely it’s a dud GPS module.
Betaflight doesn’t have anything to do with the count of satellites. We just report what the module is able to achieve. If it only finds 5, there were only 5 it could connect to at that time.
You could hook it up to pyGpsClient or uCenter and likely it would be no different.

If it says it only got 5 sats then that’s all it could find.
Physically small gps units are not as good as ones with larger antennas

[fair pagoda]

do you use the Walksnail video system?

[radiant meteor]

I was hoping it was a config issue that Betaflight was sending non-optimal config to the GPS during initialization. Otherwise it indeed seems to be a hardware issue

Thanks for your reply

[radiant meteor]

No I dont

[woven plover]

It’s possible that the particular module you have is not being auto-configured properly. If it was blocked from reconfiguration and had a locked configuration that we could not update, then it may behave differently. There is a gps_connection debug and what should happen is that each stage of the connection should walk up as described in the wiki. If that fails we keep trying but will parse whatever data we receive.
If it won’t reconfigure then usually it will only connect at one specific baud rate. So if you try several baud rates, and it connects on each, and if the debug in sensors shows the steps progressing as they should, and if the gps update frequency is what is requested, then we are configuring it properly.
Also, connecting it to pyGpsClient or uCenter will show you what it is doing.
It’s probably easier to just get a new gps module.
Can I check that you tested it also with usb power only? Ie without powering the whole quad up?

[rare fossil]

Of those having issues with 4.5 GPS, a common factor seems to be HGLRC modules - I’m wondering if anyone actually has one working without issue.

[fair pagoda]

used dozens of foxeer and flywoo m10 modules - they work perfect. have gps fix instantly

[azure turret]

Hdzero and walksnail will effect sats. Especially hdzero. I had to mount my vtx 6 inches away

[radiant meteor]

That is quite concerning to hear, I am no fan of HGLRC modules but to learn that no such module is working with BF4.5 is still surprising (as it indicates something wrong on software side)

[radiant meteor]

Yeah I only had access to HGLRC modules, and I even got another brand new HGLRC gps with the same issue. Probably getting a flywoo soon

[radiant meteor]

I use analog but thanks for the reply

[radiant meteor]

I see your point and logic. Indeed the HGLRC module is accepting BF config with multiple baud rate.
And yes I test with VTX unplugged, under clear sky, and wait long enough.
I am so confused right now because I just switch to another brand new HGLRC gps, and still observing the same issue. I have set ground assistance to None, and GPS update rate to 5hz.

[azure turret]

The geprc M10 gets 20 sats in 2 minutes 31 sats in 5 minutes. I'm using the geprc m1025-MI

[woven plover]

@radiant meteor no, nothing is wrong on the software side. It means that the manufacturer is selling modules that do not have standard UBlox command responses. We always have required GPS modules to respond to standard UBlox configuration requests. If they don’t, then their behaviour is unpredictable. Sort of partly work and may be unreliable.

Satellite count has nothing to do with Betaflight. It depends only on what the GPS reports.

Now that does depend a bit on which Satellite systems are enabled, eg whether or not Galileo is included. And that depends on our ability to configure the module behaviour. If we can configure it, and you enable Galileo, you will get it. If it cannot be configured, and is set to not use Galileo, you’ll never see those sats.

So your best bet is to get a decent GPS. Lots around.

And keep in mind that small antenna size is far worse than large ones. Can’t be recommended for GpS rescue at all.

As I said already like 5 times get uCenter or pyGpsClient and check the befhaviour of your module there. Stop suggesting that there is something wrong with the code when you haven’t actually checked that the module itself is ok.

[woven plover]

@radiant meteor If you can make a log with debug set to GPS_connection and I can look at that. Since the connection happens at boot time, you have to temporarily go to CLI and set blackbox mode = always. I only need say 30s after arming to confirm that it goes through all the auto-config steps successfully.

[radiant meteor]

I see your point that BF config should not be the issue. I am in the process of importing a config into u center 2 and take a log there as well. I'll get back to you but thanks for the reply!