SignalCheck - Android app to monitor your Wi-Fi/2G/3G/4G LTE/5G-NR signal strengths

[jsage]

I've been having issues connecting to LTE on my GNex running CM10.1 as well.

 

Hi cgass,

 

How's the weather in Seattle today? I had the same connection problems on stock 4.2.1 also, so I switched to CM 10.1 to see if it would make a difference.

 

I have two theories, both of which are probably wrong:

 

1. The first-generation Samsung LTE chip in my GNex is inferior to the Qualcomm LTE chip in my wife's GS3, so she can get LTE when I can't

 

2. Since San Diego is in the midst of NV build-out, a lot of the sites that are popping up aren't really finished, so LTE is going to come and go at least until mid-April

 

That's my story and I'm sticking to it ;-)

 

 

Have fun,

John

[4ginnc]

What I find interesting on the screenshot with the odd LTE numbers is that both EVDO and LTE are showing signal strength numbers at the same time. I've noticed on my S3 (running stock, not rooted) is when a 4G connection happens the EVDO connection disappears.

 

And, from what I know and what I've digiblur and others say, is when the LTE connection is active the EVDO connection cannot be active at the same time since they use the same transmission path. Wonder why CM 10.1 is putting up both numbers when only one of the connections is active?

[jsage]

And, from what I know and what I've digiblur and others say, is when the LTE connection is active the EVDO connection cannot be active at the same time since they use the same transmission path.

 

Hi 4ginnc,

 

A couple of observations: my wife's GS3 shows LTE & EVDO strength at the same time, running stock Jelly Bean, and I believe the single transmission path refers specifically to HTC EVO LTE radio, not necessarily Samsung.

 

I could be wrong :-)

 

 

Have fun,

John

[4ginnc]

Hi John,

 

That's strange. I've watched mine as it switched. Shows no signal strength on 4G, and a decent signal on 3G. Then when it switches to a 4G connection, the 3G signal strength display disappears and the LTE signal strength starts showing up. Then vice versa when it switches back to 3G.

 

Interesting that they behave differently.

 

Mike

[mikejeep]

Thanks to all for their screenshots and feedback posted here or sent to me directly! It has been very helpful in narrowing down the problem. It appears to be isolated to those running nightly builds of CM10.1. There are known issues with LTE signal strength calculations on those builds. I am testing a fix that will resolve many of these issues in SignalCheck; I intend to have it released by tonight.

 

In the meantime, if you don't see "LTE" as your connection type, any LTE signal data you are seeing is probably invalid.

 

Is anyone experiencing any "odd" LTE data on a non-nightly CM10 build (like an M-release)? I have not had any such reports, but I know many people prefer the nightlies. The CM folks don't really accept bug reports on the nightlies because (understandably) they are essentially beta versions for testing purposes.

 

-Mike

[digiblur]

Hi 4ginnc,

 

A couple of observations: my wife's GS3 shows LTE & EVDO strength at the same time, running stock Jelly Bean, and I believe the single transmission path refers specifically to HTC EVO LTE radio, not necessarily Samsung.

 

I could be wrong :-)

 

 

Have fun,

John

 

You guys are probably seeing something software hanging on to those connection numbers. The S3 also shares the same baseband setup as the EVO LTE where EVDO & LTE are on one transmission path and 1X is on the other. This is how it is a SVDO device. The Note2 however is a different little animal as it is also a GSM international phone where the 1X and EVDO are on the same path and LTE is on another. Technically it could do both LTE and EVDO at the same time but I do not think this is allowed via firmware.

[jsage]

You guys are probably seeing something software hanging on to those connection numbers. The S3 also shares the same baseband setup as the EVO LTE where EVDO & LTE are on one transmission path and 1X is on the other. This is how it is a SVDO device.

 

Thanks digi, I appreciate the information. I guess it makes sense because the Sprint GS3 uses the Qualcomm all-in-one SoC. Do you think this applies to the Galaxy Nexus, which uses separate chips for LTE and CDMA/EVDO?

 

Alternatively it could be a Jelly Bean issue, although 4ginnc's GS3 doesn't behave that way. I'm going to have to borrow my wife to do some LTE war-driving today ;-) I want to see if her phone behaves like 4ginnc's.

 

Have a great day,

John

[mikejeep]

AJ, digiblur, or any of the many others reading this that are more knowledgeable than me.. From what I can find online, it looks like the valid range of LTE RSRQ is -19.5 to -3. Is that accurate?

 

Most of the false LTE reports I am seeing show -20 RSRQ, so I thought that could be a useful indicator of an invalid reading. However, on my stock EVO LTE, I'm seeing -20 RSRQ several times per minute while on an active LTE connection with decent (-90ish) RSRP...

 

-Mike

[WiWavelength]

Do you think this applies to the Galaxy Nexus, which uses separate chips for LTE and CDMA/EVDO?

 

For your sake, I am sorry to say this, but the Galaxy Nexus is kind of a turd.

 

As you know, it does use separate chipsets for CDMA2000 and LTE. This is what allows the Nexus to monitor both EV-DO and LTE simultaneously.

 

Unfortunately, the CDMA2000 chipset is an off brand part from VIA Telecom and is known for its poor performance. Additionally, the multiple, older chipsets for processor, CDMA2000, and LTE definitely exact a hit on the Nexus' battery life.

 

AJ

[jsage]

For your sake, I am sorry to say this, but the Galaxy Nexus is kind of a turd.

 

Hi AJ,

 

No doubt, compared to other phones. Also, Samsung is not my favorite brand because radio performance is generally poor - GPS, Bluetooth, CDMA.

 

On the other hand, it's a Nexus phone, which is my preferred platform. And it's the only one with LTE at the moment. Let's hope the next Nexus is made by Moto. My Photon had the best radio performance all-around, but was saddled with a locked bootloader. The Moto unlocker was released after I got rid of it :-(

 

Have a great day,

John

[WiWavelength]

AJ, digiblur, or any of the many others reading this that are more knowledgeable than me.. From what I can find online, it looks like the valid range of LTE RSRQ is -19.5 to -3. Is that accurate?

 

I believe the -3 dB RSRQ maximum is accurate for two reasons. One, I have never encountered anything greater than -3 dB. Two, my admittedly basic understanding of RSRQ is that it measures only one spatial channel. Thus, with 2x2 downlink MIMO, the unmeasured spatial channel puts the measured spatial channel at best at -3 dB (half power) in reference to the total power in both spatial channels .

 

AJ

[digiblur]

No doubt, compared to other phones. Also, Samsung is not my favorite brand because radio performance is generally poor - GPS, Bluetooth, CDMA.

 

I said the same thing until I saw the S3 where I was wow'd by the radio performance then wow'd once again by the Note2's performance, especially on LTE. I live in an area where a Sprint affiliate designed the system with probably the worst tower spacing out of many areas so I know fringe signals very well. By far the S3 and Note2 have been the top performers in my area. The HTC EVO LTE worked very poorly for me but did a very good job of burning up my 300 megabyte roaming cap every month.

[jsage]

I said the same thing until I saw the S3 where I was wow'd by the radio performance then wow'd once again by the Note2's performance, especially on LTE.

 

Thanks for the feedback; I'll have to play with my wife's GS3. I spend a lot of time on bluetooth, and the GNex's bluetooth audio is really starting to get on my nerves. I switched to the wired headset today, but I sound like I'm in a can and my ears hurt :-(

 

Perhaps my next platform of choice will be the S3 or Note2 with CM 10.1!

 

Have a great day,

John

[kckid]

Hi Lee, sorry for not replying sooner. I have the same device and never experienced that problem, nor have I received any reports of similar behavior. I have a zillion questions for you; I'll reply here for the benefit of everyone else, but feel free to PM me if you prefer.

 

What ROM, what version of Android, and what version of SignalCheck are you running?

 

What do you mean by "the new menu appears"?

 

Are you running any task-killer (aka "memory optimizer") apps or anything similar?

 

How long have you been having this problem?

 

Did you recently upgrade any device firmware/software/etc?

 

If you reboot your device, does the problem come back?

 

Thanks,

-Mike

 

Mike,

 

I think I'm having phone issues with the touch in that area of the menu. Let me do somemore testing before having you spend time troubleshooting a non issue.

 

Thanks

 

Lee

[jsage]

As you know, it does use separate chipsets for CDMA2000 and LTE. This is what allows the Nexus to monitor both EV-DO and LTE simultaneously.

 

I had a chance to do some driving with my wife and her GS3. As AJ noted, my GNex kept both EVDO and LTE signals. And as 4ginnc noted, the GNex dropped EVDO when it picked up LTE.

 

Galaxy Nexus:

 

 

Galaxy S3:

 

 

I was very much mistaken when I said my wife's GS3 showed both signals at the same time. Screenshots or it didn't happen :-) In any case, I hope Mike can use the numbers for fine-tuning his app.

 

 

Have fun,

John

[MReiser4670]

I have an issue where I'm showing a false LTE signal, but I think I understand what may be happening. Take a look:

 

 

This has happened now 3 days in a row in the same general vicinity. Here in Philadelphia, we're seeing new LTE towers popping up every day so it's kind of like a scavenger hunt. At first I thought it might be just too weak of a signal, but the LTE tower that might cause that issue is currently off.

 

Since this is coming up again & again in the same area, I have to assume that there's more to this than just a bug. Also, the signal strength varies each time, and even though I'm connected to eHRPD in this screenshot, my phone will sometimes take a minute or longer to make the connection. What I think might be happening is the phone is recognizing a blocked LTE signal. Does that sound like a reasonable assumption?

 

Also, in areas like this, when I hit the "Reset Data Connection" button, I get the chimes before the cellular radio goes into disabled mode and won't reset without hitting the reset button again. This doesn't happen unless I'm in an area known to currently have (or have had) an LTE signal coming from somewhere.

 

I'm starting to learn how these little quirks are actually clues to where I will pick up LTE. I'm not sure if any of this is new information, but I thought I'd share for the benefit of everyone.

[koiulpoi]

Ooh! Ooh! Found a bug!

 

So, I set my Note II (running 4.2 AOSP-based) to LTE-only mode, to see if I could get it to hold on to weak signals further. I wanted to use signalcheck as my "bars", as the phone itself shows "no service" when in this mode. CDMA Field Test shows that it is connected:

https://www.dropbox.com/sc/tr0lxyfgwur1zws/hw_jHqBBck

but SignalCheck shows absolutely nothing:

https://www.dropbox.com/sc/nssubgytcy40yzi/IH2SEQiHNJ

 

Is this something that can be fixed? Or is it simply that Android itself isn't reporting properly?

[jefbal99]

Flipped airplane mode and now see this from my cdma galaxy note 2...

 

 

 

 

 

 

[ATTACH]2945[/ATTACH]

 

Sent from my SPH-L900 using Forum Runner

[mikejeep]

I have not had any time to finish up the next release or reply to every support issue over the past week.. sorry! Life gets in the way sometimes. Just need a little more time. Still testing the CM10.1 bugfixes, hopefully I can polish that off over the weekend. I always make sure to keep up with the thread, even if I don't have time to respond.. I will be in touch with each of you when I have an update or troubleshooting questions.

 

-Mike

[jefbal99]

Flipped airplane mode and now see this from my cdma galaxy note 2...

 

 

 

 

 

 

[ATTACH]2945[/ATTACH]

 

Sent from my SPH-L900 using Forum Runner

 

Ignore this as it is not an app issue, but a phone/rom issue

[mikejeep]

Ooh! Ooh! Found a bug!

 

So, I set my Note II (running 4.2 AOSP-based) to LTE-only mode, to see if I could get it to hold on to weak signals further. I wanted to use signalcheck as my "bars", as the phone itself shows "no service" when in this mode. CDMA Field Test shows that it is connected:

https://www.dropbox....1zws/hw_jHqBBck

but SignalCheck shows absolutely nothing:

https://www.dropbox....0yzi/IH2SEQiHNJ

 

Is this something that can be fixed? Or is it simply that Android itself isn't reporting properly?

 

Sorry for taking so long to get back to you! When you are seeing that, are any of the LTE values changing in CDMA Field Test? Or are they holding steady?

 

I just tested your scenario out by setting my EVO to LTE-only mode, grabbing a live LTE signal, then moving out of the signal area.. once the LTE signal is lost, it appears that CDMA Field Test does not clear out the stale data -- the last valid LTE signal information is "frozen" on the screen until another valid reading is obtained. It does not reflect when you are disconnected.

 

So, it looks like you did find a bug... in CDMA Field Test, not SignalCheck!

 

-Mike

[mikejeep]

I have an issue where I'm showing a false LTE signal, but I think I understand what may be happening. Take a look:

 

 

This has happened now 3 days in a row in the same general vicinity. Here in Philadelphia, we're seeing new LTE towers popping up every day so it's kind of like a scavenger hunt. At first I thought it might be just too weak of a signal, but the LTE tower that might cause that issue is currently off.

 

Since this is coming up again & again in the same area, I have to assume that there's more to this than just a bug. Also, the signal strength varies each time, and even though I'm connected to eHRPD in this screenshot, my phone will sometimes take a minute or longer to make the connection. What I think might be happening is the phone is recognizing a blocked LTE signal. Does that sound like a reasonable assumption?

 

It could be that, but which device do you have... looks like an EVO LTE on a stock ROM? If it's something else, let me know which ROM and what version of Android it is running. There have been wacky LTE readings on devices running CM10.1. You say the strength varies each time, but once it connects, does the strength ever fluctuate before disappearing?

 

I'm intrigued by the fact that you're getting a Serving Cell; that leads me to believe that you're hitting an actual tower.. Your device wouldn't just make that up; if SignalCheck doesn't "see" a Serving Cell, it displays nothing. Does the Serving Cell change as you move around? If you know where a tower is, try driving around it.. see if the last two digits change from 01 to 02 to 03. Any gurus know if the "01A1B103" in the screenshot is a valid LTE cell ID, whether in a live market or possibly set for testing?

 

It has been stated several times (including just a few days ago, a few posts up) by knowledgeable folks on here that the EVO cannot be connected to EV-DO and LTE at the same time.. but I wonder if it can "see" both at the same time if it's being prevented from connecting to one of them?

 

Also, in areas like this, when I hit the "Reset Data Connection" button, I get the chimes before the cellular radio goes into disabled mode and won't reset without hitting the reset button again. This doesn't happen unless I'm in an area known to currently have (or have had) an LTE signal coming from somewhere.

 

This happens to me too sometimes. My EVO reports the connection type as LTE when it first loses its connection. Very odd, but it's not SignalCheck-related... Android briefly reports a (bogus) LTE connection right before Airplane Mode kicks in.

 

-Mike

[MReiser4670]

It could be that, but which device do you have... looks like an EVO LTE on a stock ROM? If it's something else, let me know which ROM and what version of Android it is running. There have been wacky LTE readings on devices running CM10.1. You say the strength varies each time, but once it connects, does the strength ever fluctuate before disappearing?

 

Yes, completely stock EVO LTE. I mentioned that the signal strength varies because sometimes the signal is as strong as -84 dBm while it's not connecting. I have another screen shot for you from today I was able to capture before it failed:

 

 

The LTE signal is plenty strong @ -89 dBm, but the RSRQ is -20 dB & the SNR is 0.1 dB so that's a ton of interference. That's why I thought it might be a blocked signal? And once the connection fails & I revert to eHRPD, the LTE values that toggle on & off do not change. LTE will look like this off & on all day regardless of where I go:

 

 

I'm intrigued by the fact that you're getting a Serving Cell; that leads me to believe that you're hitting an actual tower.. Your device wouldn't just make that up; if SignalCheck doesn't "see" a Serving Cell, it displays nothing. Does the Serving Cell change as you move around? If you know where a tower is, try driving around it.. see if the last two digits change from 01 to 02 to 03. Any gurus know if the "01A1B103" in the screenshot is a valid LTE cell ID, whether in a live market or possibly set for testing?

 

As of today, I know this tower actually is transmitting a valid LTE signal as I picked up 01A1B102 this morning right outside my house about 2 miles away from this spot. As I moved closer to Horsham, I traded towers to 01A1902 but dropped LTE soon after. I've learned these towers by location thanks to your app & the Sponsor Maps...the Serving Cell ID doesn't directly match any data in the maps, but they do indeed work to identify the tower.

 

When I drive to work every morning on the PA Turnpike, my connections go like this:

 

01A16A03 --> 01A16A02 --> 01A17A03 --> 01A17A02 --> 01A01D03 --> 01A02901 (these 3 toggle 2-5 times within a minute right near my exit)

 

I know the location of all these towers except for 01A01D03 by watching when they switch sectors. As you can see your app is a valuable tool for HTC users to track towers.

 

It has been stated several times (including just a few days ago, a few posts up) by knowledgeable folks on here that the EVO cannot be connected to EV-DO and LTE at the same time.. but I wonder if it can "see" both at the same time if it's being prevented from connecting to one of them?

 

I *think* SVDO makes that impossible but I'm certainly no expert. Before I started using your app, I would have several instances of 1-2 minutes without data in certain areas where an LTE approved tower was in the vicinity. I just had no way of diagnosing it. This situation is the likely culprit; the phone desperately wants that LTE signal but is somehow blocked from connecting. Eventually either you move out of range of that tower or the phone eventually defaults to EV-DO. One spot where it was happening constantly is now where I pick up tower 01A16A03 going the other direction which is EXACTLY adjacent to a non-LTE NV tower. My phone would try to connect to LTE for a good 2 minutes but getting that non-LTE tower in between my phone and the LTE tower broke the cycle.

 

This happens to me too sometimes. My EVO reports the connection type as LTE when it first loses its connection. Very odd, but it's not SignalCheck-related... Android briefly reports a (bogus) LTE connection right before Airplane Mode kicks in.

 

-Mike

 

That makes sense...it just seemed interesting that it only appears to happen in areas that have actual LTE signal. When I was in Delaware this past weekend, I didn't have this happen once in 3 days. I was cycling pretty consistently too. As soon as I cycled data across the PA border close to an area with LTE active, it started to happen again. I don't think it's a SignalCheck bug at all, but the evidence suggests that an actual LTE signal might play a part in that brief report.

 

 

Sorry for the really long post...just trying to get all the information out there. Hopefully there are 1 or 2 points brought up which help with the SignalCheck experience.

[WiWavelength]

The LTE signal is plenty strong @ -89 dBm, but the RSRQ is -20 dB & the SNR is 0.1 dB so that's a ton of interference. That's why I thought it might be a blocked signal?

 

I agree it could be a live but barred LTE site. Or it could be a fault of the EVO LTE.

 

I live in an area of basically 80 percent complete LTE deployment. Some days, the EVO LTE performs well. Some days, it falls flat on its face in the same locations. One possible pattern that I have noticed is when the EVO LTE struggles, RSRQ and SNR are unusually low. This could indicate that the EVO LTE is creating some sort of internal RF interference with itself or that its signal metrics go haywire occasionally, causing the EVO LTE to determine falsely that the LTE signal is unusable.

 

AJ

[MReiser4670]

I agree it could be a live but barred LTE site. Or it could be a fault of the EVO LTE.

 

I live in an area of basically 80 percent complete LTE deployment. Some days, the EVO LTE performs well. Some days, it falls flat on its face in the same locations. One possible pattern that I have noticed is when the EVO LTE struggles, RSRQ and SNR are unusually low. This could indicate that the EVO LTE is creating some sort of internal RF interference with itself or that its signal metrics go haywire occasionally, causing the EVO LTE to determine falsely that the LTE signal is unusable.

 

AJ

 

Well then it's perfect timing for me to receive my new Note 2 today so I can compare the two devices side by side. I will report back in the next few days.