Oscillating channel/band = lower speed?

tod

My speed has dropped recently by more than half. One change that I've noticed in field test mode is 'Channel' oscillating between 425 and 589. Tower/sector (active set) isn't changing. Is there some sort of switching between cell & PCS bands on the same tower that could cause a slowdown?

(I've got a dual-band panel antenna aimed LOS at nearest tower. RSSI in VZAM of ~-60dBm. USB720 / CTR350)

If band-switching is a problem, can I put a filter in? Should I get a single-band antenna? If 800MHz and 1900MHz are both used, which is preferable?

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

425 is a 850 channel as well as a PSC channel.
PCS B Band 425 1951.250 1871.250
Cellular B Band425 882.75 837.75

Lets say it is a PCS channel...

Yes and No. Cross band handoffs can happen on overlay sites due to congestion to try to load balance (this setup is rare due to complications), or going from a PCS site to 850 site. Check your PN to see which one is happening. This should only happen during a dormant stage. The handoff parameters are controlled by EC/IO not RSSI, and also through distance based handoffs. With a -60dbm RSSI I would expect to see around a -3dmb EC/IO. I wouldn't expect a handoff to happen until EC/IO gets in the -8dmb range.

If this is happening during dormancy, then it is normal. A Dormant AT is actually not in a call. So it will scan different channels to see if it can find a better signal. If it is happening during a connected state, then you are in a cross band hard handoff. And yes this will destroy your speeds. It is also possible during a dormant state to see it scan the voice carrier. If you have a voice phone, you can check this. It is the voice carrier that assigns you to a DO channel. So in dormancy it scans back to mama to see where else you might go. I would bet money this is what you are seeing.

If you want to go Cell or PCS only, you can install a filter. I would go the free rout, and get a hold of QPST and use the service programmer to change your AT to cell only. Or I guess you could use a custom PRL, but that might be hard to get a hold of.

Troubleshooting DO with VZAccess Manager is like troubleshooting Electricity with your tongue. Best of Luck.

tedsaid · EVDO Newbie Joined: 26 Apr 2009 Posts: 4

How would I go about changing the AT on a pantech UM175? I have the program and it sees the card but I don't know where to go from there.

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

Hope this helps.

tod · Posted: Mon May 04, 2009 5:54 am Post subject:

Looks to me like 'PN offset' is staying the same, and 'Channel' is only toggling when 'Dormant State' is 1.

I set 'Cell Only' in QPST and tried that, but it made no difference in speed or channel behavior that I could tell.

My Ec/Io values in field test are in the -8 to -10 range.

Now I'm scratching my head on the step change in speed. I've checked and re-checked my setup - sure seems like something has changed on Verizon's end.

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

That sounds normal. The channel swapping is normal dormant behavior. You are swapping to the 425 Cellular band. One of the two is your voice carrier. Being that you are on the same PN, the two channels reside on the same sector.

The -10EC/IO is most likely the problem. Would be nice to know this measurement before the problem. At a -10 EC/IO you are really close to dropping the call. I have noticed framing starts around there, and when it gets close to -12 you are in real trouble. The minimum quality for origination is -12. Once in a call, you should be able to hang around -12 to -14 for a little bit, then it boots you.

If you have a -60rssi, with a -10ec/io, that is defiantly a FW path noise problem. At -60 you should see more like a -3ec/io. I don't know how to go about telling you how to fix that. Just make sure it is not you causing it, by having a rogue amplifier, or a bad antenna system.

If you are certain it is not you, then you can call customer service, and tell them that you have slow DO speeds with a high ec/io (pronounced E C over eye not). The Performance guy will come out and find the source of the noise.

Hope this helps.

tod · Posted: Tue May 05, 2009 10:41 am Post subject:

thanks for your help, falsedragon.

looks like before the speed trouble, I had
RSSI: -60dBm
Ec/Io: -2.5

now I've got:
RSSI: -64dBm
Ec/Io: -8 to -10

RSSI values are from VZAM (hold mouse over bars)
Ec/Io are from field test

Does the change in Ec/Io relative to the change in RSSI make sense?

I haven't changed anything in my system. I checked the antenna aiming. Still have line-of-sight to the tower. I suppose the antenna itself could have damage from moisture/temperature cycles/... Would have to get another antenna to know for sure. I could try my omni trucker to see if that offers any clues.

Could Verizon have changed something on that tower (or another) that would cause this?

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

See if i can explain this without sounding too stupid.

Lets say we are in a room. I am the AT, and you are the Cell. We are the only ones in the room. As we talk. Your volume is just fine. around a -60rssi. And I can understand every word you say. a -3ec/io. Then 3 more people walk into the room and start talking to me. They are talking a little louder then you. I still hear your volume fine. A -60rssi. But i am having a hard time understanding what you are saying, because i am getting confused with the other guys talking. now you are at a -10ec/io. Loud, but not clear.

So. Something is creating noise. Being that it is FW path, i would guess it was something in your system, or around you. IF possible, take a laptop to the location of your antenna, and run with only the device. All we are looking for is the noise floor to go down. a -85dbm -4ec/io would be a much better signal then your -60, -10. At that point, if the ec/io is still high, then it is outside noise.

Turn off the cordless phones, wifi routers, ect. Anything that can transmit.
I have seen 2.4g cordless phones kill a city block worth of 850 voice.

If you get to the point where you are nowhere near your normal environment, and still see high ec/io then it is out of your hands. Someone from VZW will need to drive test the area and find the source of the noise.

I really doubt it is the tower. When the Cell detects internal power (it samples multiple times) it goes into power control overload.

It is also possible that there is a new tower in the area, and your antenna is picking up the RF from that other site, causing the noise floor to go up. The cells are not designed for people to have 20db of gain in antennas. Thats why you are seeing no phones with ext. jacks anymore.

tod · Posted: Tue May 05, 2009 8:18 pm Post subject:

that helps - you can't dumb things down too much for me.

I will start trying to isolate noise sources in my house. The neighbors always have lots of fun watching me balance 20 ft up on the ladder with my laptop. I bought an old LG Verizon phone with antenna jack so I could use the field test mode for debugging, but I don't have the right cable adapters yet.

Before the speed problem, I would only see the sector I was aiming at in the field test 'Finger Info PN Offsets' field. Now I see several other towers in there. Is that simply a result of my bad Ec/Io, or does is indicate a possible cause?

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

Bert65 · Posted: Tue May 05, 2009 9:24 pm Post subject:

tod
Here is a link to a blog provided by Jim_in_VA.

http://evdotips.blogspot.com/2008/04/cell-tower-signal-patterns.html

This is about cell towers and signal patterns but there is a lot of other information on almost anything to do with EVDO. Might be worth a read.

My service is with Altell and shortly after the 1st of the year they started using PRLs that allowed roaming on Verizion towers, I would assume Verizion did the same in reverse. Is it possible that a new PRL was pushed to your modem even though it was in the router and now you have two towers competing for your signal and this is causing the high ec/io. Where my modem is I get RSSI about -70 and ec/io of -3. If the modem is moved 6 inches the RSSI is still very good at about -75 but the ec/io jumps to -7 -8 or -9 and a 2nd P/N will intermittely appear in the Active P/N field.

Hope this helps but mainly check out the blog by Jim_in_VA he has a wealth of EVDO infomation.

falsedragon · EVDO User Joined: 13 Oct 2008 Posts: 45

Adding a second PN into the active set will defiantly cause higher ec/io, but you shouldn't notice much of a speed hit. With a -60rssi and -10ec/io you should not be competing with any other cell without seeing it in your active set. Unless that other cell is not in the neighbor list. Also, with that hot of a signal, I would really doubt you would be seeing a minor lobe.

The PRL does not determine what channels you use when in a call. Most of the time the PRL isn't used. But this gets too complicated to fully explain. It only comes into play in origination when you first need to talk to a cell, and you have exhausted your last known used channels. At the point when the AT talks to the cell, the cell will re-direct you to where you need to go. The only way you can have another PN go in and out of the active set, is to have that PN programmed into the neighbor list.

kocoman · EVDO User Joined: 07 Jun 2007 Posts: 54

Its better to use tools such as QXDM/CAIT to figure out the "channel things". the Field Test update is too slow (ie: slow down your system, and updates maybe every 5 seconds) for any good use.