Help! 160Kbps at -46 dBm EVDO???

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

I just got ushered into another test drive by a very friendly Verizon store rep who wouldn't hear otherwise. As soon as he saw that the coverage map had changed in our favor, he had the modem on the counter and our old account on his screen. So I'm here with another activated USB720 and going nuts--partly because I've got a live modem to test with again, and partly because I'm getting such widely and wildly varied test results with it:

Speeds in the building (w/cartop antenna outside window) generally range between 40K and 500K download, with about 60K being by far the most common and lasting, and 10K to 250K upload--and a strange thing, upload speeds are higher than download speeds more than half the time, by a lot, often nearly 300%. The modem is also about half and half the time in EVDO and RTT, but these speeds are all in EVDO.

Signal strength ranges from -115 dBm to the upper -70s, and it fluctuates like crazy throughout that range, easily changing 10 or 20 dBm one moment to the next, depending on what, I don't know. There's no apparent pattern--it all happens anywhere at any old time.

But here's the scary part: We drove to one of the towers today, and where I parked at a nearby orchard, the RSSI was -46 dBm (I'm surprised the apples here don't ripen already cooked); I connected and did 3 speed tests, 1 at speakeasy and 2 at speedtest.com; all results were about the same, and all followed that strange upload-higher-than-download pattern: about 150Kbps down, and a startling 400Kbps up!

Now how can I be line-of-site a quarter mile from the tower and get 150K download in REV-A? And what's the story on these download/upload ratios and suchlike caddywampus occurrances back at the ranch? The early results of a step in coverage expansion? Extreme interference (my Field Test's ec/io doesn't function)? Tower modifications? Aliens?

My 3 computers, all Macintosh and all using VZAccess Manager, all see the same phenomena.

What's going on and what do I do about it? If I go ahead and get a strong antenna and an amp, am I simply going to reel in a whopping 150K? And will I not know one moment to the next whether I'll be connected at 50 or 500K? And...Or...?

Please, any input would be most appreciated. Thank you.

(From the tower today at -46 dBm:)

Peter B · Posted: Wed Nov 14, 2007 8:24 am Post subject:

dondrm:

I haven't got a lot to offer, but...

In one initial trial with the USB720, at a decent signal strength (about -60 dBm, I think), I was pulling in 600 Kbps to 1 Mbps peaks... roughly 3 to 5 miles from my local tower. But the speed reported by my browser during a large file download - the actual 'throughput', if you will - was only about 35 to 40 Kbps. (On a 24 Kbps dialup, throughput observed is maybe 2.5 - 2.75 K).

(The signal strength is from memory and uncertain, but reported speeds are accurate.)

So what exactly is SpeedTest testing? Were your observed page loads (and/or test downloads) appreciably faster than at home?

By the way, while testing in any environment, I found that VZ Access Manager's speed indicators fluctuated widely... and (as a guess only) I figured I was getting only an average of whatever range was showing on the graph. I also found - in weak signal areas - that reported signal strength could also vary considerably. Indoors I've seen -94dBm to -115 in Field Test... without moving the device.

For what little it's worth.

Peter B.

-----

Michael · Posted: Wed Nov 14, 2007 10:04 am Post subject:

I will also add....

If you are getting -46 dBm and you are right under the tower, adding an omni and amplifier is NOT GOING to improve your speeds.

Normally, adding an external omni antenna and/or amplifier will help your signal when you are not in a great signal area. Your signal does impact your speed, however, from what you are describing, you are not having a signal issue.

I would guess it is one (or many) of the following

- Tower Issue
- Tower Capacity issue
- Card issue
- Possible computer issue (yes, I know you tried 3 different ones, but I don't have all the details on what is installed or what isn't).

One easy way to eliminate the first 2 items on the list, is to drive to a different tower and test again.
_________________
EVDO :: EVDO News :: EVDO Antennas :: Buy Verizon :: Buy Sprint :: EVDO Amplifier

fordrhs41 · EVDO User Joined: 29 Jan 2007 Posts: 93

Much deleted
replies between the lines !!

Signal strength ranges from -115 dBm to the upper -70s, and it fluctuates like crazy throughout that range, easily changing 10 or 20 dBm one moment to the next, depending on what, I don't know. There's no apparent pattern--it all happens anywhere at any old time.

Same general thing here in the boondocks of oklahoma, signal all over the place using AT&T wireless broadband service. up and down results reversed
Plugged in my S-720 Sprint and all was well except for signal strength of -84dbm which gave a download of about 400 kbps, upload about 100.

But here's the scary part: We drove to one of the towers today, and where I parked at a nearby orchard
href="http://www.evdoinfo.com/rssi">rssi</a> was -46 dBm I connected and did 3 speed tests, 1 at speakeasy and 2 at speedtest.com; all results were about the same, and all followed that strange upload-higher-than-download pattern: about 150Kbps down, and a startling 400Kbps up!

Same general result using the AT&T card and -50dbm signal level

Now how can I be line-of-site a quarter mile from the tower and get 150K download in REV-A? And what's the story on these download/upload ratios and suchlike caddywampus occurrances back at the ranch? The early results of a step in coverage expansion? Extreme interference (my Field Test's ec/io doesn't function)? Tower modifications? Aliens?

My 3 computers, all Macintosh and all using VZAccess Manager, all see the same phenomena.

Using Windows laptop here for tests on both cards, one AT&T and one Sprint
I use the S-720 at a location 30 miles away and with a solid -72dbm, get 1200 to 1300 down from the same testing stations he used.

What's going on and what do I do about it? If I go ahead and get a strong antenna and an amp, am I simply going to reel in a whopping 150K? And will I not know one moment to the next whether I'll be connected at 50 or 500K? And...Or...?

I was baffled by the varying signal level and reversed download/upload results and so was AT&T support

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

n6gn · Posted: Wed Nov 14, 2007 1:14 pm Post subject:

Just a thought.

Since your upload performance is (sometimes at least) so good, the RF/antenna/path area is pretty well ruled out. If you can get full rate to the base, it sure can get full rate to you, everything else equal.

What you describe could be accounted for by competition for the download capacity of the segment/site. Are you sure there isn't a heavy duty user pulling pretty hard on the tower at the same time you are? 400k-600k down is only for the cases with one to a very few (2-4 *maybe* with rev A) users sharing a segment/site. As I've described elsewhere, once there are very many users sharing an EVDL segment, per-user throughput drops quickly.

n6gn

Michael · Posted: Wed Nov 14, 2007 1:43 pm Post subject:

The only issue - is how would one user *know* if there is another heavy duty user, is using the tower?

There is an internal Sprint tool called GAMMA and we have been trying to gain access to it for a long time, it would show how many users are connected, bandwidth in and out of the tower and the range of the tower.

P.S. If anyone at Sprint is reading and they can pull some strings to get us GAMMA access, I would be EXTREMELY grateful!
_________________
EVDO :: EVDO News :: EVDO Antennas :: Buy Verizon :: Buy Sprint :: EVDO Amplifier

n6gn · Posted: Wed Nov 14, 2007 3:19 pm Post subject:

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

Michael, it seems from the email I just received that you and the others at 3G didn't notice my response to your response earlier in this thread. In case that's the case, I'm letting you know now that it's there a few posts up from here, and that it explains all of those points of the email I just received--and if it's not the case, and my response has been taken into account in 3G's recent email to me, then, well, never mind. ;-)

Jim_in_VA · Posted: Wed Nov 14, 2007 9:11 pm Post subject:

What is all that info when you type ##DEBUG?

Michael · Posted: Wed Nov 14, 2007 11:01 pm Post subject:

Your support ticket going back and forth has had 4 different people involved at different times and any information you post up here on EVDOforums.com, does not automatically get added to your support/sales ticket.

Someone will get back to you tomorrow and they will read this thread before replying.

Thanks!
_________________
EVDO :: EVDO News :: EVDO Antennas :: Buy Verizon :: Buy Sprint :: EVDO Amplifier

Hp_EVDO · EVDO Fledgling Joined: 06 Apr 2007 Posts: 13

Ive got software called QXDM which is pretty much Qaulcomms software for communicating all the inner workings of their chipsets... Including signal levels, modulation and such. There is so many graphs and numbers and things i cant even identify! Maybe that would be of some use in this situation?

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

3 speedtests at home before going out to the towers--wow:

At about -100 dBm:

A couple of minutes later at about -105 dBm:

And about 10 minutes later at -95 dBm. Yowsuh!

dondrm · EVDO User Joined: 06 Sep 2007 Posts: 64

Got to only one of the two towers left to test yesterday, the one with the most powerful signal at home. Also swung by the first one again, and got the same 150ishDL/400ishUL results as last time.

This second tower showed more of what you might expect in the -40's dBm range, but also a lot of everything else--a roller coaster, in fact. Below are 4 exemplary speedtests out of about 10, all at the low -40s and even upper -30s dBm.

It would seem that the instability is understandable in the light of the usage thing pointed out earlier, since Verizon would naturally allocate a minimum of hardware to cover an area where half the homes don't even have a computer, and each one of the few users involved would thus be a good percentage of all the users the hardware was designed to accommodate. So if some aspect of the hardware/sub-area is near capacity with 3 users, and one of them--me for example--comes on and does a speedtest, then 33% of the users have suddenly come online and started downloading. I was laughing to myself thinking, "Here goes my speedtest, slowing itself down 33%"! lol