T-Mobile General

[gregorys]

This is a query. I think the problem is more widespread than we know, and I need to know if you've had the problem, or know someone who has.

The problem, is as follows:

The person on the TMobile network is NOT heard, but the other party is heard by the user on the TMobile network. (I've been on both ends of a call, both as the caller using the TMobile cell, and the "callee" using a regular land-line.)

The microphone cut-out on the Cell-phone is usually 2-4 seconds, and seems to have no periodicty (i.e. regular re-occurance, say every 30 seconds). Again, the caller on the Cell can hear the called party just fine, but the person on the other end of the call hears dead air.

It often seems to occur just between pauses or words, but not always. (The fact that it occurs in pauses causes it to be really pernicious since the called party has no cue to know they're missing something, other than dead-air.)

I've opened a trouble-ticket with TMobile and they have verbally confirmed to me that it's "a known problem." [Though they claim it is just my Model/Brand of phone, but I'm hearing the same symptoms on other phones too - G1's and Blackberries.]

If you're having the problem or know someone that is, could you please confirm on this thread?

-Greg



 

[adanziger]

There's a thread going on non-T-Mobile devices about this problem with Palms. For you, this could just be a bad connection or oversaturated spectrum, that can cause it, but it might also be an incompatible GSM device from a non T-Mobile source.

I've been doing a lot of reading and asking a lot of questions. There have been a great many changes made to what we just think of as GSM. This might bore you but read these following paragraphs and you'll see what I'm talking about:

In the 900 MHz band the uplink frequency band is 890–915 MHz, and the downlink frequency band is 935–960 MHz. This 25 MHz bandwidth is subdivided into 124 carrier frequency channels, each spaced 200 kHz apart. Time division multiplexing is used to allow eight full-rate or sixteen half-rate speech channels per radio frequency channel. There are eight radio timeslots (giving eight burst periods) grouped into what is called a TDMA frame. Half rate channels use alternate frames in the same timeslot. The channel data rate is 270.833 kbit/s, and the frame duration is 4.615 ms.

GSM has used a variety of voice codecs to squeeze 3.1 kHz audio into between 5.6 and 13 kbit/s. Originally, two codecs, named after the types of data channel they were allocated, were used, called Half Rate (5.6 kbit/s) and Full Rate (13 kbit/s). These used a system based upon linear predictive coding (LPC). In addition to being efficient with bitrates, these codecs also made it easier to identify more important parts of the audio, allowing the air interface layer to prioritize and better protect these parts of the signal. A codec is a device or program capable of encoding and/or decoding a digital data stream or signal. ... Half Rate or HR or GSM-HR is a speech encoding system for GSM developed in the early 1990s. ... Full Rate or FR or GSM-FR was the first digital speech coding standard used in GSM digital mobile phone system.

GSM was further enhanced in 1997 with the Enhanced Full Rate (EFR) codec, a 12.2 kbit/s codec that uses a full rate channel. Finally, with the development of UMTS, EFR was refactored into a variable-rate codec called AMR-Narrowband, which is high quality and robust against interference when used on full rate channels, and less robust but still relatively high quality when used in good radio conditions on half-rate channels.

I mention this so that you can understand that there are many different "implementations" of GSM. There are 124 channels available, there are 1/2 rate and full rate channels, and a bunch of different codecs. If T-Mobile has upgraded some of their towers to the new codecs and the phone your using doesn't support it, or if the tower, to avoid interference switches channels on you in the middle of a conversation, to a channel that your phone doesn't support, you will have silence on one end until it changes channels again to something that your phone supports. T-Mobile only supplies phones that it fully supports. Some of them causing trouble like Palm Centros and others may not have every codec or frequency available to it that T-Mobile has.

You're not going to be able to get help if this is your problem from Customer Care, it's going to take someone in their engineering department and even then, I'm not sure that they'll ever be able to help you.

[gregorys]

That thread is one I'm actively participating in...

It's not bad coverage, or non-open channels on a cell, etc. It happens every time, regardless of signal etc. It happens all over the country. The symptoms are very extraordinarily similar. It's not a "chance" event.

The problem is pervasive to virtually every call - even a 2-3 minute one.

 As for an incompatible "non-tmobile" handset - that suggestion really rankles me.

GSM is the standard. These same Palm devices work the world over - on AT&T's network, and elsewhere without any problems. So, it's (IMO) a cheap-shot to claim that they kinda' shouldn't work on TMobiles network, or that they (TMobile) really shouldn't be expected to really work to solve the problem.

I'd say there's nothing forcing them to do so, but that doesn't mean they *can't* or that it would simply be too difficult.

I'd guess that some concerted effort in working with the base-station hardware/software vendors they use would result in some *deterministic* finding. Something like - there's a specific problem X with specific hardware Y - we have an updated software rev and you should see a fix within 2 weeks etc. Or, while I'd like it less - perhaps we know there's a flaw or incompatibility X with handsets Y, and while it's fixable, we don't see it as financially viable to address it. But since that costs you in having to switch handsets, we'll be glad to assist you with "Z" (a rebate, handset discount, credit to the bill etc.)

But none of these things are forthcoming.

My ticket's been in more than a week, and I keep getting the "well we're waiting on Palm."

As far as I can tell, they didn't get any assist from Apple on the iPhone issues, but they still fixed those.

IMO, if TMobile wanted to do so, they could address this issue and a straight-forward manner. But for some reason they're not doing so. My guess is they want to avoid responsibility - so they shift the blame, they claim it's not their handset etc. It's too hard. All that technology stuff, it's just too complicated. We couldn't be expected to solve a problem on a *world standard* GSM network with a *world standard* GSM phone. Etc, etc.

Again, IMO, that's the weasel way out.

The right way is:

A) Figure out what's wrong, and let me know when it's going to be fixed,

OR

B) Say..."we know what's wrong, we could fix it, but we aren't going to."

'nuff said.

Message Edited by gregorys on 09-28-2009 09:04 PM

[adanziger]

Please understand that I'm not making excuses for anything, I'm just telling you what the problem is related to. I don't work for T-Mobile or Palm, I've just researched your problem extensively and this is what I believe is causing it.

GSM is a standard and like all standards, is evolving all the time. Just because Palm continues to sell these phones or has not updated the firmware doesn't mean that the phone is compliant with all of the newest standards. It might be more than a firmware update, the phone may not even be capable of receiving all of the channels, it might not have the electronics to do so. Maybe the phone works on everyone else's network the world over and maybe it doesn't, neither of us knows because neither of us has traveled the world over to test it. And if it works today, it might not work tomorrow due to ever changing standards.

Regardless of whether it does or doesn't it, we do know that it doesn't work reliably for you on T-Mobile's network where you live and work. Assuming that moving isn't one of your options, that leaves you with two realistic choices; switch phones or switch carriers.

If, and this is a big if, it's T-Mobile's problem, and after all of my reading I don't believe it is, then yes, they should really try to fix it, but what if T-Mobile has installed the very newest equipment where you live, say to relieve frequency congestion, and the new channels aren't all supported on your phone, what then? Should they not upgrade their system, resulting in gradual deteriorating connections for everyone else as the frequency gets oversaturated from too many users?

All of your comments and complaints are making the assumption that it's T-Mobile's fault, I can't be certain, but the further I dig into this the more inclined I am to believe that it's not. I believe that Palm is selling a few phones that don't have the latest revisions to the GSM codecs or channels, and that brings you back to those two choices that I mentioned a few paragraphs back.

[gregorys]

But you are making excuses for Tmobile, IMO.

Unless TMo has dropped GSM support (which it hasn't, or a host of older phones would fall over) then the GSM standards are all backwards compatible. GSM-97 and GSM-99 are both compatible with the original GSM standard.

I have no idea why you think somehow the phone is negotiating to use a channel it can't use. (It wouldn't) - but I'd guess that any old phone from 10 years ago will function fine on a GSM network today. It might not be able to handle newer features like text messaging, or MMS etc - but it will handle standard voice communications fine.

That aside...

What IS clear is that my Palm phone didn't change in last August. There were no firmware updates etc. The only possible change is on the TMo side. 

TMo could fix the problem if they wanted. It doesn't yet appear they have any real desire to do so. Perhaps that's because it's not profitable, or because engineering hasn't been told there's a problem, or because all the senior executives got abducted by aliens. But the facts are they have not fixed it, and haven't communicated any useful info to me or anyone else about what the problem is and  whether they intend to do something about it.

Lastly, I have at least two first-hand accounts of the problem occurring with other TMobile provided handsets. So, these symptoms aren't just limited to Palm devices. 

-Greg

Message Edited by gregorys on 09-29-2009 10:39 AM

[jurgy1]

I'm new to these forums, but not new to TMo.

Been using this service since Aerial days.

I have had several Palms/Treos over the years - 180/270/600/650/680.

Most recently was using 3 680s on my plan - me - spouse and one kid.

My business partner also had 680 and some guys in our boy scout troop had 680s.

Within ONE WEEK - all of the above mentioned phones starting having the same problem as discussed here.

It has been interesting reading these topics as I have learned almost the same info from independent research.

Our research has lead to the following - told, in part, to  my business partner by a TMo rep. when he

went in to look for a new phone.

The newest upgrade to 3G status has caused an upgrade in some frequency work.

TMo did not include the portion of that upgrade that would effect Palm/Treo phones 

as the company had a falling out with Palm about 5 years ago when Palm and AT&T 

released the 650 without TMo being invovled. 

It is MY OPINION - no technology behind this - just my opinion - that TMo does now 

consider Palm hardware to be EOL (at least as far as TMo is concerned) and that 

since TMo has not officially sold a Palm device since the 600 about 5 years ago,

that they have no obligation to update the equipment/software that operates such a legacy device.

So unless TMo changes its mind and decides to upgrade the service to include whatever

is missing, the Palm/Treo/Centro devices will continue to have intermittent failures.

Maybe Palm has  a way to fix that, but I don't know and couldn't wait.

It bugged me enough that I had to switch devices - I use an HTC now - not a problem in the world (except for

crappy WinMo OS)

My business partner went with BlackBerry - not a problem (except for crappy OS, too).

If TMo fixes it - I will go back to my 680 - since it was/is far superior.

But until then - I am sticking with TMo dropping the ball and letting me down.

[dessa]

The other thread referenced in this thread is:

http://forums.t-mobile.com/tmbl/board/message?board.id=08&message.id=1494#M1494

[gregorys]

 Cross posting this here since it's obviously happening on other handsets, [that's confirmed] and it's frankly quite interesting...

---

Ok, I now understand the problem. I spoke with an RF Engineer about it.

There was a BTS upgrade to the system - perhaps not in all markets - but in mine, and I would assume everywhere where the problem exists...that upgrade was applied by TMobile about five weeks ago. [This also explains why some users report that the problem ceases when they travel outside of their "normal" use area. Those areas may not have base-stations by Nokia / Siemens, or perhaps don't have the same software revision that exposes the problem.]

The real issue is due to a "feature" called DTX - or Discontinue Transmission. In essense it allows either end to "discontinue" transmitting when there's a silence. In short, the phone recognizes that you're not talking and enters DTX mode. This means it doesn't have to transmit on the radio during the time DTX is enabled, thus saving power. The phone is supposed to recognize when you start talking again, and signal the base station to discontinue DTX mode.

For some reason, DTX mode isn't discontinued when you start talking again and the uplink radio traffic is either lost or doesn't occur. Since DTX isn't new, one would assume that it worked properly before, and that some interaction in the new base station upgrade has revealed some interoperability problems.

Normally the carrier could release updates to a phones firmware - because they would have access through the vendor. This gets into a grey area, I expect, as to what's possible etc. But suffice it to say that I expect TMo isn't exactly excited about spending time and money on a fix to the handset they never sold or promised to support - even if Palm would allow them to modify the firmware (which Palm might well not agree to).

Palm doesn't want to spend any resources themselves on Palm OS when it's dead and gone and they're trying to get WebOS going and keep the company alive. [And I've been told as much by Palm.]

However, as I said before, this DTX problem does occur on some other handsets - ones TMo sold. Thus, they'll want to address these handsets.

The wrinkle is this. They can address the interoperability problem in either of two ways.

1) They can update the software in the handset so it doesn't end up in a faulty DTX mode in relation to the base-station.

2) They can pressure Nokia / Siemens (that page link above is now dead – go figure) to re-work the software in the base-station, also to improve the interoperability problem.

For those of us with Treos and Centros – the first solution will leave us out in the cold. The second has a good chance of fixing it for us.

But I don’t know how urgent TMo feels the problem is, and how they’re leaning to fix the problem. If they only have a few handsets with the problem – and they can get the problem to go away by making people upgrade etc – they’ll probably choose option 1. It would be a lot cheaper and easier I expect.

If they get a boat-load of problems, or Nokia/Siemens has a rash of problems with the base-station software – we may get option 2.

If I were a betting man, I think I’d put the odds right now of the fixes as 60-70% likelihood on option 1, and only 30-40% on option 2.

---

Also, to get the problem to go away – pause in taking a bit, and start talking again. The transition into and out of DTX mode on the handset again will likely cause it to reset the DTX state properly – and they’ll be able to hear you again.

---

The whole issue is rather fascinating and a rather stark lesson in interoperability issues.

I’m glad to follow-up as you have questions!

-Greg

Message Edited by gregorys on 10-02-2009 10:41 AM

[MaxPower]

Does anyone know if cycling "mute" then "unmute" bumps the Palm handset out of DTX mode (and thus re-establishing the voice channel from the handset)?

I'm probably grasping at straws, but if this has a positive impact on the problem, it would be possible to write a small app that "unmutes" every few seconds while calls are active via the Palm OS TelSndMute() API call. It's an ugly imperfect approach, but it's an ugly imperfect situation...

[barry99]

Seattle tacoma area. Blackberry 7290 (t-mobile branded). Same problem and it is driving me nuts. started about the end of august. i have replaced with a used t-mobile 7290 (t-mobile)  and it still is occurring. Move the sim into an old traditional nokia phone an dit works just fine.

[gregorys]

barry99

Thanks for the report.

[rickp315]

I have had a similar issue with a Motorola Renew.  It happens when I talk to people on land lines or other cellular devices (regardless of carrier).

Just recently I bought a used Samsung R225M, and oddly enough it seems to have fixed my problem.  But I still do not know if it was device related issue, or simply older phones are better transceivers.

[adanziger]

It's a device related issue. Some usually older devices, or devices that have older technology (even though they might currently be sold by different manufacturers), have some known incompatibilities with some of the recent upgrades made by T-Mobile at some of their towers. This happens most frequently when they upgrade their 3G service in an area. There appears to be no solution coming. T-Mobile doesn't guarantee backwards compatibility as technology advances and the affected hardware manufacturers appear reluctant to do anything about their products that are nearing the end of their life cycles. Many if not most of these devices are made by Palm as well as some older Blackberries.

[gregorys]

adanziger

I have to take issue. It is NOT a backwards compatibility issue. 

In theory, any GSM handset should be able to operate properly - even one that's 7-10 years old.

In this case, BB's and Palms and G1's (these are the ones I've heard specifically) are having the problem. I haven't seen a technical description of it, but I suspect it's that the tower doesn't realize that the handset has come back out of DTX mode. As I said before, the same feature DTX is, was and  has been available for a long time as part of the GSM standard.

It's not changed at all - so to say it's a backwards compatibility issue is simply dead wrong.

If you want to say there probably an ambiguity in the GSM standards about how DTX communications is handled and that a new software revision doesn't handle it in exactly the same way, (yet still adheres to the GSM standard) and this interoperability is the problem - that I'll buy.

 In rickp315's case, it is the NEWER handset with the problem, not the older one. The renew is brand new (q1, 2009), and the Samsung appears to have been released Q4, 2001 - that's around 8 years ago.

We'll see about a fix. If brand-new handsets (just at the start of a manufacturing cycle) have the problem, I'd bet they'll have to address it via a basestation upgrade/patch.

-Greg

Message Edited by gregorys on 10-04-2009 07:07 PM

[adanziger]

It seems that most if not all of the problems are occurring in areas where T-mobile is performing upgrades to their service. Specifically I believe that a lot of it is related to the BTW-S/W upgrade which is a router update for 3G upgrades now and 4G down the road. It's for improved efficiency in switching, data carrying and energy usage. Here's a link to more information than you ever wanted to know about the BTW-S/W upgrade. Lots more is available with a quick Google search. It may also indeed be a DTX issue with some handsets, though I would describe that as a compatibility issue as well. But I guess I shouldn't call it backwards compatibility, though some newer handsets may not fully support the current or newest standards.

http://www.nokiasiemensnetworks.com/NR/rdonlyres/9DD8041D-DD13-424E-BDA2-6F034787D5D8/0/Flexi_multir...

[gregorys]

I *know* it's a DTX issue since I've talked with an RF engineer at TMobile who discussed the current issue at some length with me. Obviously as evidenced above, it doesn't have to do with how "old" or new the handset is. It has to do with how they communicate DTX state.

It's hard to get more "from the horses mouth" than that, IMO.

I was also told that locations that don't use the Nokia/Siemens base-stations, or which haven't had the software upgrade, operate fine and don't have the DTX state problem.

I'll leave it there. I just find it grating to have you claim it's "bad" or "old" handsets and essentially give TMo a pass because, it seems you imply, they would have such a hard time supporting all those handsets properly.

It happens to very new handsets as well as some older ones. It happens to both TMo sold handset as well as not. 

The standard didn't change. The Treo and all other handsets are following the same standard [DTX] they were before. The standard wasn't new, it didn't change. Implimentation changed at the base-stationin a particular rev. of the software.

For TMo to change that implimentation wouldn't in the least impact their ability to handle new handsets, or 3/4G data. It would simply be a rework, or revert to the old implimentation method.

It's the equivalent of me digging a hole. I used to throw the dirt over my left shoulder, so you stood on my right side. But suddenly I started throwing the dirt over the right, and covered you in dirt. It wasn't your fault. Perhaps it wasn't mine - since I didn't realize it would cause a problem. But there's no reason they couldn't go back to throwing the dirt over the left shoulder again. It worked fine before, and it would work fine again.

Perhaps they just don't want to, but there's no technical reason they couldn't do so again. Obviously a number of handsets are impacted by the sudden change. A change that isn't evident on any other networks out there other than TMo, to my knowledge - and they're already all supporting 3G and fast pushing the data envelope to 4G [LTE/WiMax].

-Greg

Message Edited by gregorys on 10-04-2009 08:16 PM

[hisacs]

opening up an old thread...

Are people still experiencing this problem.

I've had this issues for last 3 months and initially I thought it was the microphone so I ended up replacing that. But now it seems like it could be broder than that.

Device is :Eten Glofiish X600, touchscreen WinMo6.5.

Location: in and around seattle

I also switched to BT headset but the issues it still there only with that device. sometimes I see it cut-off within 2-3 minutes and sometimes I can talk for up to 20 minutes.

Now I'm afraid of buying any non-t-mobile devices if it could happen on those? anyway to find out which devices it'll happen to or due to any particular chip? ro will it get sorted out after any other planned upgrades by tmobile?

thanks.

[dessa]

As noted in the other thread (look for my post above to a link to it), my wife and I both had this problem consistently with our PalmOS devices (GSM unlocked Centro and Treo680) from late August 2009. Since moving to Android (MyTouch 3G and Nexus One) at the beginning of the year, we have not experienced the problem.

[gregorys]

I've just gotten word that the new Siemens-Nokia base station firmware has been approved for application to T-Mobile's network. This update should fix the DTX state mismatch issues that have been seen on PalmOS phones and some other handsets.

I've been told to expect roll-out across the whole T-Mobile US infrastructure in approximately the next 30-45 days.

For more details (I've documented most of this pretty throughly as I've been through it all...)


http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

http://forums.t-mobile.com/t5/Non-T-Mobile-Phones/Palm-Centro-intermittent-voice-with-T-Mobile-servi...

So, now the DTX problem should cease to exist in just about a month - at least till a Siemens-Nokia software engineering programmer goes on a bender and as Wally from Dilbert fame says "I'm gonna code me a mini-van. [See: http://i18.tinypic.com/34461wi.jpg]

Cheers!
-Greg

[hisacs]

Thanks gregorys.. I'll keep an eye for that.. Hoping it'll be fixed for my phone as well (ETEN Glofiish X600 WinMo)... its there a specific chip which I can identify on my phone to confirm that the issue will be fixed.. ?