PilotAware
British Forum => Technical Support => Topic started by: PhilPassmore on April 23, 2019, 03:59:53 pm
-
Hi, just recently bought PAW to go in my SSDR and therefore not fitted it in the aircraft yet. I am bench running the setup, and on the whole am well satisfied with the results I am seeing. There does seem to be one small problem though. Despite running on a very clean, well regulated and multi-amp power supply, I am getting times when SkyDemon seems to lose connectivity with the PAW. At least initially, the PAW wifi network is still present, as it will sometimes re-connect. However, after a number of drop out/ reconnect cycles, it seems to get to the point where I am losing connectivity completely. I have tried upping the wifi power in configuration, and my tablet running skydemon is only a matter of a meter or so away from PAW. After powering off PAW for a while, all seems happy again. Is this likely to be a cooling issue? I have PAW inside a window at home, but the most recent failure occured with no direct sun, and not particullarly warm. These failures have happened after anything from a couple of hours of running, to 24 hours continuous.
-
Phil
My first suspicion would be 5V power to PAW even with your regulated supply.
Can you give some details on that supply?
Regards
Gerry
-
It's a home brew power supply (I am pretty electrically/electronically savvy). Should quite happily deliver in excess of 5 amps.
-
Hi Phil
I would suspect the power supply
First thing to check is the home page after running for a while, there is a warning generated for potential power issues
Could you post your home page please ?
Thx
Lee
-
I will try to replicate the problem and reproduce the home page here. Thanks for the help.
-
Phil
It isn't a Lenovo tablet by any chance?
-
Phil,
With some devices, too high WiFi power can be worse than too low. Try decreasing the WiFi power level to say around 20mW. Remember you need to reboot for WiFi changes to take effect.
Let us know how you get on.
Regards
Peter
-
Phil
It isn't a Lenovo tablet by any chance?
No, it's Galaxy Tab A6
-
Now running on a commercial 5A USB power supply module, with only one output in use. WiFi power turned down as suggested above. Will update if difficulty replicates.
-
Please update regardless of outcome Phil. It confirms the fix in case anyone else has the same issue.
-
Hi Vic
I'm intrigued by your question about the possibility of it being a Lenovo tablet.
What do you know about Lenovo tablets that i don't I've recently purchased one to run the new EasyVFR 4 that requires a 64 bit processor & min 2 gig memory to run it.
Alan
-
Morning Alan,
Predominantly Historic now, but have a read at this thread from back in 2017....
http://forum.pilotaware.com/index.php/topic,1091.msg12991.html#msg12991
Vic might be able to provide more up to date info. Hopefully not an issue with newer Lenovo Tabs - unless you find otherwise....
Regards
Peter
-
Thanks Peter
I did type "Lenovo" into the search box but it only came up with the post from Vic on this thread. >:(
Alan
-
Hi Alan
I had horrendous problems with at times constant disconnections with Skydemon on my Tab8-3 tablet to my PAW when UDP connection came in. The device seemed to be doing something 'far more important ' at times and 'ignoring' data coming in over wireless.
The eventual resolution was to create another user on the tablet (which is never used BTW) and the issue dissapeared. Why this resolved the issue is probably only known to Lenovo software design engineers who built and poorly supported the device >:(.
Vic
-
Thanks Vic
I read with interest all the problems you had on the thread that Peter linked to so I'm keeping my fingers crossed that they have resolved the issue in the Tab4-8 that I have just bought in order to become a Beta tester for the new EVFR4.
Alan
-
Well here is my current update. I still seemed to be experiencing issues, whereby PAW and SD would quite happily communicate for a significant time, then I would experience problems. From the messages that SD is putting up, it is looking like it is losing the GPS, and so therefore disconnects. Initially, it holds onto the wifi connection which is still present, but eventually because no GPS data is being received, it gives up completely. Yesterday I drove for about 3 hours with PAW on the dash of the car, with a good view of the sky, and it worked fine, up until about the third hour, when the GPS problem came up again. I didn't try to reconnect at this point, so left things till I got home. When home, I was getting no GPS data at all, with a problem showing up when scanning the Home IP address from a second device (sorry, can't remember what it said, it was getting late) but the message on SD said that it had zero satellites, and the green light on the GPS module wasn't coming on. Again wifi was all present and correct. Just as an experiment, I moved the GPS across to the adjacent USB port, and it all sprang back into life, and slightly more oddly, worked again on the original GPS port when I changed back to it. I was giving it loads of time to initialise the GPS each time. I am currently bench running everything, with the GPS plugged into the second USB port to see if my original problem duplicates after a period of time.
-
Hi Phil,
The PilotAware Team are at Popham this weekend (well most of us are anyway).
Intermittent / infrequent disconnects can be a real pain to diagnose, as these can be due to internal problems (with either PAW or the Nav System) or due to ‘external’ issues, such as the power supply, or even to external devices such as phones or tablets desperately trying to re-establish ‘full’ Open Internet Acccess via for example a shared Bluetooth or WiFi hotspot - which could well be on another device in the aircraft. This can occur for the strangest reasons. If the problem recurs, it is essential to try to take a screengrab of any messages and also the PAW Home and Logging Pages and post them on here to help us determine what is going on. A shot of the Configure Screen can also be helpful. (Obviously much easier if testing on the ground or if you have a friend in the plane with you).
Looking forward to hearing from you.
Regards
Peter
-
Hi Phil
The message from SD is confusing
SD continues to communicate when the GPS data is nonexistent.
To prove this, try removing the gps dongle whilst SD is connected and running, it will continue but report loss of satellite
The issue is that the wifi connection is broken
So either Pilotaware dropped the wifi or the ipad droped the wifi
Pilotaware will sometimes drop the wifi IF it has poor power source
Ipad will drop the connection if it thinks there is a better available
An acid test is to remove all network auto joins on the ipad
Also, when you see the disconnect, please take a look if your ipad has selected an alternate network
And give us a pist of Pilotaware home page
Thx
Lee
-
Ok, been running for ages today, not been watching the device, so not sure when the issues started. I came in to find the 'not received GPS data in some time' message, and checked the wifi connection to the tablet, which seems solid, with no indication of dropping. I pressed 'Go Flying' again, and selected PAW, and now I get an intermitant connection, with it flicking from 'waiting for device' and it showing my location. Observing Wifi for a while, again, it remains connected, and seems stable. It has been running on the same power source throughout
-
And just got the following message, wifi still connected.
-
I’m sure it won’t be this, but I had a similar issue a while back - PAW would work fine and then occasional issue (drops etc) after working fine for an hour etc.
After a lot of mucking about, I took my case apart and found that the PAW bridge was coming away from the main board - the pins were visible. I think the turbulent air and constant connection and disconnection of the PAW antenna had forced the items to slowly part. I pushed the bridge back in to the connector on the main board and then prevented it from parting by putting something inside against the case so when it was screwed back together, there was no means of anything parting company. This was the PAW classic - I have the Rosetta now and no way of it happening.
After that fix, never had an issue.
-
Hi Phil
Please note the message throttled 0x50005
This indicates a power issue, if the power reduces the wifi will drop out, this seems to be the issue
What power source are you using ?
Also I have a galaxy tab, and I recall it has a weird issue where after a while it becomes slower to respond to wifi data, and after about a day it would disconnect
Thx
Lee
-
Thanks for the feedback. When the fault arose I was running a rechargeable battery being charged while used (LiFePo4), with a commercial four way USB power board, rated at 5 Amps across the four outputs, with only PAW being powered off the board. Power consumption had settled to about the 600 milli amp level. What is puzzling about this if it IS a power problem, is that it will run for a significant tme before the fault arises, and there doesn't seem to be either a surge in consumption, nor any problems with the power source.
-
Average current is quite low at about 800-900ma
However spikes occur (during transmit etc), and in a perfect
Storm a spike plus heavy processing can cause the power
Rail to dip, then you gave an issue
This warning is generated at about 4.7v
Thx
Lee
-
Phil
I had similar problems as you. I have LifePO4 (7 Amp Hour) and it powered Radio and PAW with iPhone 6.
It would drop PAW consistently. When 8.33 became mandated I upgraded to Yaesu 550 but wired that separate from Battery with a 9V supply provided by 12V - 9V Board with regulation.
The iPhone 6 and PAW are now powered via a Charge 2 Unit and the problem has completely gone. Another aircraft in my Hangar chose a Charge 4 after PAW inconsistency.
That too is working 100% reliably. Power/Voltage needs to be adequately above lowest limits of RPi.
Regards
Gerry
-
Hi Phil,
I have just got back to the Forum after another busy day at the Popham Rally on Sunday, then a night in my tent before a (very) long day dodging rain and snow yesterday flying back home to East Fortune.
I’m glad to see that you have now posted screenshots we asked for and that they have borne fruit and that we are now getting to the crux of your problems.
As Lee and Gerry have advised, power levels are critical for effective operation of the RPi. This can be a major source of frustration when fault-finding remotely - especially when the Nav System messages point you to the ‘GPS Device’ when it really means the ‘Data Source’ - hence why we ask for the screenshots. SkyDemon has BTW now changed their messages in the iOS versions which makes this clearer, so I presume Android will follow suit.
For the benefit of future readers of this thread, please let us know how you resolve the problem.
(Personally, I have been using a ‘Charge 4’ in my flexwing for about 3 years now. It provides reliable power to my Rosetta while also maintaining 2 x iPad Minis at full charge - provided they are both pretty much fully charged before starting out.)
Best Regards
Peter
-
also maintaining 2 x iPad Minis at full charge - provided they are both pretty much fully charged before starting out.)
Best Regards
Peter
Hmm, should be able to charge an iPad mini to 100% even if full backlight is on and it's running something like SkyDemon. Have you put new USB cables on these at all recently?
Note that ALL usb cables will degrade over time, as they are made of many tiny strands to ensure they are flexible. After time these start to break dropping the effective current carrying capability of the cable.
For for Apple Lighting I've also seen the connectors on the device get a bit 'fluffy' preventing good contact and charging issues. A cocktail stick (as it is wood, so non conductive) can be used to GENTLY clean the socket of embedded fluff. It's worse for phones kept in pockets, you'd be amazed at how much you can get in there.
It's a common problem, reported as broken device, charger broken, the supplier does this on software to make me buy a new device, etc. when all they need is a new cable and/or a de-fluff.
-
Hi Jeremy,
Point taken. I was being cautious! We often get PAW users trying to run tablets and PilotAware from the same cigarette-lighter adaptor and then complaining when the tablet takes all the power and throttles the processor in the PAW, causing all sorts of issues, so normally advise against trying to charge tablets from a low charge state in planes because of the high likelihood of interference or other issues.
My Charge 4 habitually runs PilotAware plus the main iPad Mini ‘in the plane’ even on 100% screen brightness and often other devices such as my GoPro as well and the tablet charge state is always 100% by the end of any flight - using a ‘fitted’ JuicEBitz 20AWG USB to 90 degree Lightning cable.
For the trip down to Popham I also ran a second iPad Mini on a kneeboard as a backup and to allow comparisons using the Beta Mode-S/3D software. This pad was connected to the Charge 4 using a domestic standard ‘Apple Type’ lightning lead (which might not even be ‘Apple’). It was fine all the way down to Popham, even though I used both tablets ‘out of the plane’ for a while on the ground during a couple of weather diverts, but on the way back I forgot to plug the charging lead back in to the second tablet when I left Sherburn-in-Elmet. I only noticed this when I got a ‘Low Battery - 10% Remaining’ Warning on screen. I immediately plugged the charging cable back in but to my surprise the pad never got above 11% during the remainder of the trip.
I did notice that the ‘battery charge light’ in the plane wasn’t showing a particularly healthy state for most of the flight from that point on but to be fair I was running PilotAware, Transponder with ModeS-ES, FlarmMouse, Strobes and both iPads and also had the landing light ‘on’ most of the way from that point due to poor viz. I guess that was asking too much of the aircraft alternator. Strangely, however, the main iPad Mini retained 100% charge all the way home!
Regards
Peter
-
Ok, so I bit the bullet and bought an Anker, PowerCore+ 20100, and being back home today after a few away, got to try running PAW on it. Well, long and short, after about 5 hours, back to dropouts and the 'Home' screen showing 'throttled=0x5005' as before. Nothing else is running off the Anker, the devices are in close proximity, the tablet seems to think WiFi is still present. Got any more ideas, as I am starting to feel a little frustrated?
-
Hi Phil
We will be in touch to arrange a return
A fully charged anker 20100 should give more than 5 hrs
How many charge leds are showing on the anker pack ?
Thx
Lee
-
By the time I had a problem, it had only dropped two 'notches' on the Anker display.
-
That is a bigger charge status drop than I would have expected in 5hrs. My 22000mAh power pack lasts more than 24hrs.
I can see no mention of the USB lead. Is it still the supplied 1m juicebitz cable? I have had two of these cause power issues. I now only use a 20cm juicebitz lead and have had no further issues. When PAW moved from supplying a 0.5m lead to 1.0m, I was a bit surprised.
Note: my battery is piggyback with the PAW so it is more convient to use a short lead as well.
-
Hi Phil,
I was going to say that it is looking like a faulty cable or could be a faulty power socket on the RPi. Lee’s offer to arrange a return sounds good to me and will hopefully resolve the issue.
Regards
Peter
-
The fault has been reproduced on several USB cables, including the original supplied one and a second JuiceBitz one that I purchased.
-
Hi Phil
Can I get one of the track files from when a disconnect occured ?
I will pm my email address
Thx
Lee
-
Lee, by all means, if you tell me where I need to obtain it; However, recent development. Was reading through other peoples woes, and came across a thread where they said they had had a very similar issue, except that their SD and PAW wouldn't communicate from the get go, even though the wifi was showing as active (pretty much the same symptom as mine, except mine would drop out after a variable period of time). They resolved it by changing the 'external device' setting in SD from PAW to Flarm, and then connecting via that. Well today, I have had PAW up and running for over nine hours by following this same route without a glitch. The log file is still showing a 'throttled 50005' return, but no falling over. I have been unconvinced that this has been a power issue, and this 'fix' makes it look more so. To be honest, I would be concerned if power performance was so marginal that a change of USB cable made the whole system fail.
-
Phil,
In an interesting development, SD has just released a beta test update this afternoon for iOS, in which...
The FLARM connectivity code has been rewritten in this build to be more resilient. if the AirConnect device drops the connection, or anything else occurs to the connection, SkyDemon will relentlessly keep re-attempting the connection forever. It will never kick you out of navigation mode.
Not sure how this relates to dropouts when using the ‘PilotAware’ UDP Connectivity option though. Perhaps Lee can advise.
WRT the track files, these are automatically recorded by date and start of flight time (e.g. 20190514:10.34) under the PilotAware ‘Tracks’ Tab (accessed via 192.168.1.1) and can be downloaded to a USB stick using the controls on the ‘Tracks’ screen - usually by using one of the ‘empty’ ports on the Pi for the USB stick, or where the ‘spare’ ports have been used for ‘FLARM-In’ or ADSB-Out’, by temporarily removing the GPS and using that port. (Attempting to download tracks or upload updates to a port which has been configured for one of the aforementioned uses can corrupt both the PAW Track files and the content of the USB stick).
Best Regards
Peter
-
Am I missing something on the above post?
No need for messing about with USB sticks just use the browser on your phone to log on to the PAW and download the track and just email to yourself if you want to access the data on say a laptop later, or send in an email to somebody else.
-
No - just me showing a memory glitch! :(
When I did the post yesterday, I assumed Phil P would be using the standard Public Release Software Version 20180520, which I didn’t think had the Direct Download option in it (One of the perils of running multiple Beta versions on test over the years), so I described the ‘old but still commonly used’ USB method of downloading track files. I just dug out and powered up one of my old units to check and found to my surprise that ‘Direct Download’ was in the 20180520 Release - my doesn’t time fly - Sorry for the confusion!
Which means of course that either method can be used, depending on your preference.
Thanks for the reminder :-[
Best Regards
Peter
-
Phil
I have set this to SOLVED as using the FLARM connection works for you.
I think we have a solution for the PilotAware/UDP version as well which will be in the release at the end of May/2019.
I can give you early access to test if you wish ?
Thx
Lee
-
Just to let folks know, the PAW/SD connection seems much more stable when connected via the 'FLARM' connection option in SD External devices.. That said, it has been agreed that my hardware will be returned to the company for more testing, just to make sure there isn't an issue with that (a replacement being provided) PAW are doing this despite me doing a little modification work along the way, trying my own solutions to the problem, so I would really like to thank the team for their help in all of this, they have been going above and beyond with stuff happening via email that hasn't been seen on the forum. Definite hallmarks of a first rate outfit.
-
I'm really interested to hear if the team find any problems. I am experiencing exactly the same issue as the OP and will now try the FLARM setting in SD as a get around. I'm using a juicyBitz lead (i've tried 2 new ones) and a PowerADD 30000mah battery pack with nothing else plugged into it and still getting the 0x50000 fault showing historical power drop . The cpu temperature was at 66oC (a bit hot I think).
I have now ordered the (Keith Vinnings) recommended battery pack and will test again.
-
iOS or Android.
We have solved a nasty bug with Android recently, but difficult to say which Android devices are affected as they are all different. This fix is in the imminent release. 66degC is not really an issue, these are rated to 85DegC and will throttle if they get too hot.
I have forced a Pi3B to 104DegC - reported throttling, but continued to work correctly
Thx
Lee
-
I'm using a Samsung Tab3 -8 T315 on Android and the JuicEBitz cables are 1mtr.
-
A very recent update to Skydemon last week...
3.12.7
29 May 2019
"Improvements to the resilience of FLARM connections, which are now automatically reconnected if the connection drops "
--------------------------------------------o----------------------------------
SimonM, That's exactly the same tablet I have which gave me endless problems.
see http://forum.pilotaware.com/index.php/topic,1091.msg12991.html#msg12991
Create another user on the tablet (no need to ever use it) and your problems will disappear!
-
Thanks Vic, I'll give it a try along with the FLARM setting.
Interesting this wasn't an issue when i initially used Rosetta but has been a P.I.T.A for the last 2 or 3 months.
-
Vic, Simon
If you are able to take an early release - I would be very interested in your findings with the Android devices ?
I have spent a lot of time trying to get this working properly, on my Samsung it is now flawless.
The problem was sending data which no serviced were listening, so for example I would send UDP data for PilotAware and GDL90, this really upset Android. I think there is some kind of denial of service software which is registering that no process is listening for the UDP data, and somehow judges it to be a threat.
anyway new upcoming version (certainly for my Android devices) does not have this problem.
Thx
Lee
-
Vic / Simon,
Slight confusion I think - Vic you’re problems were with a Lenovo Tab 3 - Simon is using a Samsung (Galaxy) Tab 3 as I read it.
Both Android devices though, so Lee’s ‘fix’ should be valid for both.
Regards
Peter
-
I'll give it a try Lee :)
-
Quick question... Could my drop-outs have anything to do with the bluetooth connectivity with my Bose A20 headset? If so will Lee's upgraded software address this issue?
-
Hi Simon,
Ihavent given you a pointer yet have I.
I will PM directly, do you have a second MicroSD ?
Thx
Lee
-
yep, got a 2nd MicroSD