Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Wifi :)

  • Please log in to reply
21 replies to this topic

#1 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 12:02 AM

I found something very interesting that might shed light on WiFi woes.  So many posts on wifi dropping.

 

After several sessions of flawless and strong wifi connections (dongle and built in), i finally hit a snag!  Finally, just to sympathize with many of you.

 

What changed?

 

Simple, it was my first night out with battery power for the AVX mount, connected via dongle.   Fully charged, no slewing at all.  All my previous sessions were on AC power to AVX.   Well, tonight AVX was powered by Celestron power cell pro (the bigger one).   Wifi or more specifically - CPWI kept erroring out - saying mount has stopped responding.   Laptop showed wifi connected, but CPWI had lost connection, and it was not possible to get it back without restarting.

 

I tried about 5 times, failed each time.   CPWI lost connection in a few mins.   Laptop showed connection.

 

So having proven battery power was some connected (pun intended) to the connection drop, I brought in an AC connection and boom!   No connection drop for 2 hours!  I was struggling to stay connected for 5 mins on battery.

 

So not sure what this is - but hoping people can pool their experience to help pinpoint.   Somehow on battery power, the connection is not holding over wifi -- I am not sure if its actually wifi dropping (because laptop still showed connection to Celestron) or the information being transmitted over wifi is getting stopped some how (I suspect the latter).

 

This means, on battery power I should use USB to controller and use wifi when i have AC power.   



#2 Forward Scatter

Forward Scatter

    Viking 1

  • -----
  • Posts: 775
  • Joined: 22 Jul 2018
  • Loc: Wandering the PNW

Posted 30 September 2020 - 01:14 AM

Hi Sanjeev
Its a good idea to check the amperage of both the output from the AC adapter and the battery while running the mount and wifi dingle. The battery output may not be sufficient; the AVX really is sensitive to low current. also, which wifi dingle version are you using? You can tell by the number of digits on the wifi ID displayed when connecting to it by your PC. The dongle version with 3 digits are the most stable. The older 2 digit flavors had major connectivity issues.

#3 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 01:32 AM

Hi Sanjeev
Its a good idea to check the amperage of both the output from the AC adapter and the battery while running the mount and wifi dingle. The battery output may not be sufficient; the AVX really is sensitive to low current. also, which wifi dingle version are you using? You can tell by the number of digits on the wifi ID displayed when connecting to it by your PC. The dongle version with 3 digits are the most stable. The older 2 digit flavors had major connectivity issues.

 

The dongle has 3 digits (902 I believe).   As mentioned, i have used it in 10+ long sessions with zero issues using AC power.   The AC adapter is Celestron, rated 12V and 2A.  I checked the voltage output to be sure -- Celestron has another 12V battery charger that actually outputs 16V.  This is definitely outputing 12V.

 

As far as the Celestron battery pack (Power tank Pro), its fully charged, brand new, rated 12V 5A.  So i have a hard time understanding why it would be the battery capacity itself.

 

But you got me thinking.

 

Now I am wondering if it is possible that the wifi action gets impacted if voltage drops.   I am very familiar with LiPo (these are LiFePo4) and one thing about LiPos -- if you have continuous draw (or a sudden draw when starting to slew for example), instantaneous voltage can drop a lot.   By even 2 to 3 volts, depending on the circumstance - this happens to fully charged LiPo's.   A nominal 24V battery can drop 3V when i instantly accelerate my RC plane or go vertical.   With a 20 pound scope to move, it is possible the instantaneous voltage could drop similarly.  Perhaps the Celestron tank has high capacity (it says can power scope for 17 hours) but not enough surge capacity.  In the RC world, we look at peak amps capacity - so a battery that is 6 cell 5000 mah 65C will be substantially better at dealing with current draw and maintaining voltage compared with a 6 cell 5000 mah 25C.

 

The voltage drop may not matter for most other operations, but the wifi part is likely dropping packets or getting switched off for an instant.

 

Possible?


Edited by SanjeevJoshi, 30 September 2020 - 01:35 AM.


#4 Ittaku

Ittaku

    Ranger 4

  • -----
  • Posts: 354
  • Joined: 09 Aug 2020
  • Loc: Melbourne, Australia

Posted 30 September 2020 - 01:33 AM

Most batteries are notoriously unreliable at even maintaining 12V and often fall below it. This is almost certainly what's affecting your wifi functionality. What is and how big's your battery etc. and have you actually measured its voltage during usage? I use a 95Ah car battery...


Edited by Ittaku, 30 September 2020 - 01:34 AM.


#5 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 01:42 AM

Most batteries are notoriously unreliable at even maintaining 12V and often fall below it. This is almost certainly what's affecting your wifi functionality. What is and how big's your battery etc. and have you actually measured its voltage during usage? I use a 95Ah car battery...

Yep i agree with you, our posts crossed -- i was guessing the same thing based on my lipo experience with high performance RC planes (which i monitor realtime).

 

The battery is rated 158Wh, but if you look at my post above, there is something called a "C' rating for LiPo's - might be something similar for LiFePo4's.

 

Essentially - more than the overall capacity (eg. 95AH vs 200AH) what matters a lot more to maintain voltage is the C rating.   For RC planes we spend a lot more money to buy batteries with much higher C ratings.

 

Now if you are using car batteries - their "C" rating will likely be higher than what the scope needs, so no issue.   

 

I need to figure out how to make a quick step down for my 24V LiPos to 12V - that will answer this definitely and without investing in more batteries, since the C rating will be well above what the scope needs. 



#6 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 01:45 AM

thinking of using this to do my test - can deliver 20A at 12V.

 

https://www.amazon.c...ZAR4ZD08B1JCBJZ



#7 Noah4x4

Noah4x4

    Soyuz

  • *****
  • Posts: 3,796
  • Joined: 07 Apr 2016
  • Loc: Colchester UK

Posted 30 September 2020 - 01:46 AM

I think it is more complex than this and suspect it is not merely a question of power. I always use an AC input at home and can still report strange failings that I had to overcome...

 

Celestron WiFi whether first, second or third generation SkyPortal would work fine at a remote dark sky site, but at home has proved unreliable for me due to local urban 'network clutter'. I could turn around simply placing my body between scope and tablet and that would result in "dropped packets". Signal strength was consistently too weak to compete with other local networks that might overwhelm it. You don't notice this when (say) streaming a movie as your eye won't miss the odd frame, but a scope is sensitive to this as it loses direction (hence wild slews might result). Even when using Access Point mode I needed MESH signal boosters (which helped). But until I switched to the 5Ghz channel and employed a dual computer Windows Remote Desktop networking solution instead of Celestron WiFi  I suffered all manner of problems with wireless remote control. Even then I made discoveries like how to disable RemoteFX compression to remove restrictions on data flow over my WAN. But there are wider communication ssues that could be relevant...

 

For better reliability, I now connect HC to scope side computer using a serial adapter cable. But even if your HC has a USB output it is inherently serial/RS232. I have noticed that if I plug it into a USB3 enabled port on my Intel NUC it is flaky. But it works fine when plugged into a USB2 port, but the ports are supposed to be backwards compatible. I have experienced this with two different computers so it isn't a problem with the port itself. This is merely a nuisance, but it took some frustrating diagnosis to find a solution. Old technologies and USB3 don't always play nicely.

 

I also did try creating a WiFi bridge to the Intel NUC using a second Wireless adapter and this proved flaky despite the two being only a foot apart.  There is also a known problem of interference between USB3 devices and the 2.4Ghz channel which is used by Celestron WiFi. 

 

I suspect all of these issues are probably related. Serial/RS232, the 2.4Ghz channel and USB2 are progressively becoming  redundant technologies but in astronomy we cling onto them for backwards compatibility. My guess is that there are more users of Windows 7 laptops in astronomy compared to in any other hobby.  If Celestron and other manufacturers supported only Windows 10, plus only the 802.11ac WiFi standard, 5Ghz channel, USB3 and similar then I suspect things would work better. But the need for backwards compatibility is probably holding us back as regards reliability. 


Edited by Noah4x4, 30 September 2020 - 01:56 AM.


#8 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 01:51 AM

i can now virtually guarantee my wifi will not fail on AC power, inside or outside the house.  So at least in my case, it is most likely a voltage drop issue, and I will prove it.   Either use a step down converter for my high performance 24V LiPos or just get a car battery.   But i will report back once i fix it.



#9 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 30 September 2020 - 02:40 PM

Doing more testing at home.. some interesting findings and bad news!

 

1.   I can improve the performance slightly by using the cigarette lighter output from the power bank (rated at 10A) instead of the 12v pin out - but this improvement buys me just a little more time

2.   After #1 is done, at least at home, I can fix this issue by making sure EITHER the mount OR the laptop is connected to AC power.   Issue only exists when both are on battery

 

So it seems both the laptop AND the mount go into some kind of a lower power connection mode (not maximum) for wifi when they are on battery.  If it was just one of them, i would not be able to solve the issue by plugging AC into either.  This was an important find.

 

I went thru every possible setting in Windows 10 including command prompt actions, no luck.    For whatever reason, once both laptop and scope are on battery power, the wifi handshake data packet flow breaks, and this is independent of ALL possible settings in Windows 10 -- adapter, PCI, power plans, battery management, etc etc - including command line to fix the powerplan, adapter, and power saving mode.

 

Then i saw a tangential article regarding Windows 8 (I am on Windows 10 home) which said to change plan to "balanced" instead of any other.   So changed from "ASUS Recommended" to "Balanced" -- all settings still tweaked to prevent wifi adapter from sleeping.    

 

This almost worked.. i was able to go for roughly 15 mins (just a guess) before "mount stopped responding".   Compared with 2 to 3 mins max.

 

But unfortunately, even with "balanced power" plan in Windows, it cant work.

 

 

In summary:

 

1.   Issue only shows up for me when both mount and laptop on battery, never if either of them is on AC power.  Its not about the amperage but how the driver / adapter work with each other

2.   Based on MSFT, this is a driver compatibility issue -- the Realtek device might need a driver update from OEM to work better with Windows on battery

3.   "Balanced Power" almost works, but not quite

 

Until the driver is fixed, there is no workaround unless you have AC power.   Good old USB to USB mini.   :(



#10 Linwood

Linwood

    Viking 1

  • -----
  • Posts: 844
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 30 September 2020 - 02:55 PM

If Celestron and other manufacturers supported only Windows 10, plus only the 802.11ac WiFi standard, 5Ghz channel

Bear in mind 2.4ghz goes further and passes through walls, etc. better, so for people imaging in their own back yard it can be a good thing.  Remember when 900mhz phones (for landlines) would work down the block and in neighbor's houses.  High frequency carries a lot of data, but it is more localized. Which is one reason it's being pushed, to cut down neighbor-interference (also, well, bandwidth). 



#11 TelescopeGreg

TelescopeGreg

    Vanguard

  • -----
  • Posts: 2,168
  • Joined: 16 Jul 2018
  • Loc: Auburn, California, USA

Posted 30 September 2020 - 11:45 PM

Bear in mind 2.4ghz goes further and passes through walls, etc. better, so for people imaging in their own back yard it can be a good thing.  Remember when 900mhz phones (for landlines) would work down the block and in neighbor's houses.  High frequency carries a lot of data, but it is more localized. Which is one reason it's being pushed, to cut down neighbor-interference (also, well, bandwidth). 

In practice, 2.4 ghz can also be problematic, as that radio band is shared with Bluetooth and microwave ovens.  The latest problem is that USB3.0 uses a 2.4ghz modulation overlay in order to get the massive bandwidth, and when some of it leaks out of the cables, it can interfere with a WiFi connection.  We fortunately no longer have the 2.4 ghz "baby monitor" problem, and the 2.4 ghz cordless phones are gone.  Those were a kick.  We had one guy at work who set up his cubicle all very Feng Shui.  Monitor, phone, laptop, fern.  All fine, until he got onto a conference call.  Then the laptop would lose connection... 

 

So, it all depends on the environment.



#12 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 01 October 2020 - 11:56 AM

Continued testing to isolate the driver issue.   Tested battery to battery with Alienware (which is a power hog, blazing on).    Has the same Windows 10 Home build (1909).   

 

Success!  No glitches, at any distance thats practical (within 25 feet, in wifi congestion).

 

This means the driver or settings bug is solveable in Windows 10.

 

I am going to get a USB high power wifi for the Asus laptop to see if it can hold.   Its 20 dollars on Amazon, cheap research.    If the driver works with Windows 10 home, its the cheapest / easiest solve.  I really like the light Asus, also has great battery life (which will go down a bit with high power wifi), easily rechargeable in the field, etc.   Also like the ability to use HC with CPWI, really helps for visual with focuser and alignment.   Turning the HC into a boot loader screen display takes out the CPWI benefit for visual.

 

Will report back after that test in a few days (waiting for USB wifi to arrive).



#13 robbieg147

robbieg147

    Messenger

  • -----
  • Posts: 430
  • Joined: 23 Mar 2020
  • Loc: Kent, England

Posted 01 October 2020 - 04:41 PM

Although I would get some nights where WiFi would not drop out, it was always unreliable and I now use a USB hub connected by a long cat 7 cable no problems since.



#14 Maritime

Maritime

    Vostok 1

  • *****
  • Posts: 174
  • Joined: 19 Aug 2020

Posted 01 October 2020 - 04:59 PM

Continued testing to isolate the driver issue.   Tested battery to battery with Alienware (which is a power hog, blazing on).    Has the same Windows 10 Home build (1909).   

 

Success!  No glitches, at any distance thats practical (within 25 feet, in wifi congestion).

 

This means the driver or settings bug is solveable in Windows 10.

 

I am going to get a USB high power wifi for the Asus laptop to see if it can hold.   Its 20 dollars on Amazon, cheap research.    If the driver works with Windows 10 home, its the cheapest / easiest solve.  I really like the light Asus, also has great battery life (which will go down a bit with high power wifi), easily rechargeable in the field, etc.   Also like the ability to use HC with CPWI, really helps for visual with focuser and alignment.   Turning the HC into a boot loader screen display takes out the CPWI benefit for visual.

 

Will report back after that test in a few days (waiting for USB wifi to arrive).

God, if I only understood 1/10th of this...meanwhile, using only direct connect, my 60th, number 522, works great with SkyPortal and my iPhone. 



#15 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 01 October 2020 - 07:55 PM

God, if I only understood 1/10th of this...meanwhile, using only direct connect, my 60th, number 522, works great with SkyPortal and my iPhone. 

Yes, cant beat that for convenience for sure!



#16 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 01 October 2020 - 08:10 PM

SOLVED!   (after roughly 2 days of testing)

 

To summarize:

 

Problem:  Using my Windows 10 laptop, wifi would drop to AVX connected by wifi dongle.

 

Cause:   When both scope and computer are on battery -- the lowered power status for wifi caused the disconnect either quickly or after 15 mins.  Putting either one on AC power was fixing it for me, but this was not a voltage or amperage issue.  It was happening with no slewing or amp consumption in minutes.   I dont know what is happening on the wifi dongle / AVX side but on the laptop side, certain laptops / wifi adapters seem to struggle with battery power settings in Windows 10.  Regardless of deep changes to settings, the adapter would drop the connection.

 

Solution:  Use an external USB based wifi dongle, and suppressing the built in laptop wifi.  I used this Netgear A7000 (TP 1900) solution.  I dont need to use the stand for it but it puts less stress on the USB3 socket.  It is a bit of an overkill probably (a smaller one could do) but the 4 antennas in this are strong enough to do the job and more.

 

Even using this solution, i had to tweak a number of settings, and disable a USB service (this adapter app was calling on) to get the solution to work flawlessly.   Now it does work flawlessly and at any distance I choose.

 

Hope my experience helps others who might be struggling with a similar issue.

 

I so love working with a laptop (vs. phone or ipad), now i can have it all.   Laptop for pole master, stellarium, basic imaging and regstax, CPWI etc.  I do like using the HC for fine slewing and fine focus control, and this allows me to get the best of all options.

 

 

 

 

 

 

 

Attached Thumbnails

  • IMG_1676.jpg
  • IMG_1677.jpg


#17 Linwood

Linwood

    Viking 1

  • -----
  • Posts: 844
  • Joined: 09 Jul 2020
  • Loc: Ft Myers, Florida, USA

Posted 01 October 2020 - 08:14 PM

Did you try going into the power options in the control panel, work your way down to the advanced settings, and turn off all the power saving settings that might relate to the wifi/network/usb adapters? 



#18 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 01 October 2020 - 10:37 PM

Did you try going into the power options in the control panel, work your way down to the advanced settings, and turn off all the power saving settings that might relate to the wifi/network/usb adapters? 

Ofcourse.  Here is the summary / run down.

 

The behavior is very different between adapters, there is no specific pattern to it.

 

1.  For example -- my heavy duty Alienware has a Qualcomm adapter, and with no setting changes at all, except setting performance to "better" on battery saving plan, it gets the job done.  Will not drop Wifi for hours at distances up to 30 feet (thats the limit of my testing).  On Windows 10 Home build 1909, same as case # 2 below

 

2.  My Asus Vivobook 15 has a Realtek card (combo wifi and bluetooth)  - and 10+ setting changes ranging from the very basic network adapter (dont allow pc to shut this off) to things like "roaming aggressiveness", windows default power plans, device manager menus, USB root hub settings, command line settings, etc... --> no change.   I never thought i would learn so much about network adapter settings - from networking center to device manager to power plans to command line sequences.   Nada, no luck.  Same Windows 10 Home build as case # 1 above

 

3.  Same laptop as case #2 above, switched off Realtek adapter; then a small adventure with Netgear A7000 (TP 1900).    Installed then deleted app, saved drivers, installed drivers.   Adjusted adapter setting, adjusted power plan setting (not much), disabled a USB service that some say was problematic, modified settings for USB root hubs.  Boom.  Now you cannot break this wifi unless we had an EMF lol.

 

Now take my laptop findings and extend to Android devices, which tend to be a lot less controlled than Apple devices.  You will have a cacaphony of irreconcilable results of cases 1, 2, and 3.

 

It is conceivable anyone with case #1 would never have a wifi issue, wonder what the fuss is all about.    Those with #2 would just give up frustrated unless they were skilled enough to do # 3.

 

And there is at least one person who did # 3, because he is looking forward to a trip to a dark site with his family someday when the smoke clears :)



#19 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 04 October 2020 - 08:56 AM

Final update.     An impromptu planet viewing session in my backyard - Jupiter, Saturn, Uranus, Neptune, Mars and moon.  I simulated a trip to the field by running everything on battery sources.   Laptop on built in battery and scope on Celestron Power Tank Pro (LiFePO4).  WiFi dongle, CPWI, Starsense, Polemaster, NextImage, etc.

 

Absolutely rock solid WiFi now that the cause is addressed.

 

Separately it was a good check on power consumption (granted on a pleasant night).  The light laptop had 65% after 4.5 hours (visual, basic images), Celestron scope batt was at least 3/4 full, maybe a bit more.

 

And the go to’s continue to impress.  I now keep a 6 mm or 9 mm focused from one object to the next based on seeing conditions and the objects are either perfect center or very close to it.



#20 Maritime

Maritime

    Vostok 1

  • *****
  • Posts: 174
  • Joined: 19 Aug 2020

Posted 04 October 2020 - 09:29 AM

Can you provide a one line summary of the cure in lay terms?  Thanks



#21 SanjeevJoshi

SanjeevJoshi

    Vostok 1

  • -----
  • topic starter
  • Posts: 191
  • Joined: 07 Sep 2020
  • Loc: Los Altos CA

Posted 04 October 2020 - 10:01 AM

Can you provide a one line summary of the cure in lay terms?  Thanks

In my case -on battery power, to switch off the built in WiFi adapter in my laptop, plug in an external one, and tweak settings to make it perfect.



#22 Maritime

Maritime

    Vostok 1

  • *****
  • Posts: 174
  • Joined: 19 Aug 2020

Posted 04 October 2020 - 10:20 AM

Thanks!




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics