dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
2503
share rss forum feed


ctceo
Premium
join:2001-04-26
South Bend, IN
Reviews:
·Virgin Mobile Br..

Problems with 4G outage

Is anyone else experiencing strange connectivity issues with Sprint's 4G network in the South Bend, IN area?

My 3G is getting damn near 130ms ping, 4mb down and .65mb up which is double my 4g connectivity speed on a normal day as well. What gives my 3G ping speed, stability and dl/up speeds are REAL sucky.



dib22

join:2002-01-27
Kansas City, MO

said by ctceo:

130ms ping

That's pretty much a normal ping for me on EVDO.

chris92

join:2008-09-20
Coal Valley, IL
Reviews:
·Clear Wireless
·Sprint Mobile Br..
·Mediacom
reply to ctceo

3G spec is only 3.1Mbit down. Depending on the tower, if you are one that is heavily utilized you may only get 0.5-1.3 down, even with Network Vision upgrades. An underutilized tower can exceed 2 down. Also not every tower has 4G, so you may be pulling 4G from far away that degrades a lot worse than Verizon's.


criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom

Also I think it's helpful to keep in mind Sprint's minimum specs for these networks. Their 3G minimum is dl/ul 600/350, their WiMax minimum is 3mbps/.5mbps and their LTE is 6mbps/2mbps. Knowing those minimum specs at least gives you firmer footing when you call tech support if you're in a so-called "good" area which ain't acting so good.



compuguybna

join:2009-06-17
Nashville, TN
Reviews:
·Millenicom
·HughesNet Satell..
·ooma
·Virgin Mobile Br..
·Charter
reply to chris92

I think theoretical max for EVDO/rev A is 3.1mbps down....HOWEVER, if you were originally connected to a Sprint 4G LTE tower, and you travel outside the range, you CAN drop to eHPRD.......which is faster than regular EVDO...

Its the same on GSM networks.....LTE can drop to HSPA+, then down to HSPA (3G)....

said by chris92:

3G spec is only 3.1Mbit down. Depending on the tower, if you are one that is heavily utilized you may only get 0.5-1.3 down, even with Network Vision upgrades. An underutilized tower can exceed 2 down. Also not every tower has 4G, so you may be pulling 4G from far away that degrades a lot worse than Verizon's.


chris92

join:2008-09-20
Coal Valley, IL
Reviews:
·Clear Wireless
·Sprint Mobile Br..
·Mediacom

eHPRD is EVDO that is routed through an LTE core. eHPRD is just a software layer on the EVDO radio that allows the handoff for EVDO to LTE to be seamless. eHPRD can be faster than EVDO since it is being routed through a completely different core router than EVDO that is usually closer placed. For example, the one here in Davenport, IA is routed through Chicago(175 miles) instead of Kansas City(350 miles), so ping times are now below 100ms. eHRPD can't go over 3.1MBit either since it is EVDO.

This article helped me understand this: »s4gru.com/index.php?/topic/1819-···try52976


chris92

join:2008-09-20
Coal Valley, IL
Reviews:
·Clear Wireless
·Sprint Mobile Br..
·Mediacom
reply to criggs

Adding to criggs's post, they will also want an exact address of where you are getting bad service. If you are in "Fair" coverage, they don't guarantee the speeds.

In my experience, I had a device that was failing, and one of the symptoms was extremely slow speeds (0.1Mbit down vs ~1Mbit). The first thing the rep. did was check the address, and said that there is nothing they can do about the speed. Needless to say, it took over an hour with support to get a replacement device.



compuguybna

join:2009-06-17
Nashville, TN
Reviews:
·Millenicom
·HughesNet Satell..
·ooma
·Virgin Mobile Br..
·Charter
reply to chris92

I guess my Galaxy Tab 3 on Sprint is lying about its speed on eHRPD then...

said by chris92:

eHPRD is EVDO that is routed through an LTE core. eHPRD is just a software layer on the EVDO radio that allows the handoff for EVDO to LTE to be seamless. eHPRD can be faster than EVDO since it is being routed through a completely different core router than EVDO that is usually closer placed. For example, the one here in Davenport, IA is routed through Chicago(175 miles) instead of Kansas City(350 miles), so ping times are now below 100ms. eHRPD can't go over 3.1MBit either since it is EVDO.

This article helped me understand this: »s4gru.com/index.php?/topic/1819-···try52976


chris92

join:2008-09-20
Coal Valley, IL

That could very well be...I've noticed a couple times the Speedtest.net app will have wierd spikes to speeds that don't seem right. I usually have it spike to 25+Mbit for upload on LTE when it should be 10-15 Max



compuguybna

join:2009-06-17
Nashville, TN
Reviews:
·Millenicom
·HughesNet Satell..
·ooma
·Virgin Mobile Br..
·Charter

Most speedtests are pretty inaccurate for the most part. My tablet does seem a bit faster though for streaming music when its in eHRPD:13...doesn't buffer as much.

said by chris92:

That could very well be...I've noticed a couple times the Speedtest.net app will have wierd spikes to speeds that don't seem right. I usually have it spike to 25+Mbit for upload on LTE when it should be 10-15 Max



dib22

join:2002-01-27
Kansas City, MO
reply to chris92

said by chris92:

That could very well be...I've noticed a couple times the Speedtest.net app will have wierd spikes to speeds that don't seem right. I usually have it spike to 25+Mbit for upload on LTE when it should be 10-15 Max

Usually caused by compression or cache engines.

criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom
reply to chris92

said by chris92:

Adding to criggs's post, they will also want an exact address of where you are getting bad service. If you are in "Fair" coverage, they don't guarantee the speeds.

That's odd. That's not what I was told when I called in about that back in September. They told me the minimum LTE Spec for "Fair" areas was 4 down/1 up.


GadgetsRme
RIP lilhurricane
Premium
join:2002-01-30
Canon City, CO

2 edits
reply to ctceo

For a more accurate test try:
»www.speedof.me
Or:
»www.testmy.net
--
Gadgets



Jim_in_VA

join:2004-07-11
Cobbs Creek, VA
kudos:4

pretty cool site for testing speeds



GadgetsRme
RIP lilhurricane
Premium
join:2002-01-30
Canon City, CO

I just added a second site. These are the two I use anymore. No more Java or Flash tests.
--
Gadgets


criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom
reply to GadgetsRme

said by GadgetsRme:

For a more accurate test try:
»www.speedof.me

I got pretty odd results, in comparison to what I get with other sites.

On Speedtest.net, here's what I got:




On Speakeasy.net, here's what I got:

Your Speed Result:
Download Speed: 51023 kbps (6377.9 KB/sec transfer rate)
Upload Speed: 34290 kbps (4286.3 KB/sec transfer rate)

And on ISPGeeks.com, here's what I got:

Download speed 57587 Kbps
Upload speed 38454 Kbps
D/load COS 97 %
U/load COS 93 %

But on speedof.me here's what I got:

Download Speed: 17 Mb/s
Upload Speed: 6.54 Mb/s

Now if those are megabits, not MegaBytes, then they're dramatically slower than I'm getting on the other three speed test site. On the other hand, if those are MegaBytes, then I'm getting 136 mbps d/l and 52 mbps u/l, which would seem to be extraordinarily fast.

This was even worse. In fact, testmy.net was shocked, and told me flat out I might have a problem since I was so far down in comparison to what I should be getting on Verizon FIOS. Here were the results:

Download :: 8 Mbps 1 MB/s up arrow Upload :: 6.6 Mbps 830 kB/s

I can certainly explore the possibility that my connection is in trouble. But that conclusion does not seem supported by the Speedtest, Speakeasy and ISP Geeks results (and those are the sites that I've trusted for years). Hmmm.


GadgetsRme
RIP lilhurricane
Premium
join:2002-01-30
Canon City, CO

When you get a chance read the how it works link on the »www.speedof.me site it explains how it works and why they feel it is more accurate.
--
Gadgets


criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom

Well, I just did, and I have to say I'm not impressed. I set a lot of store on real world experiences too. One of the most useful is downloading and uploading files from my ftp site. On my Windows 7 64-bit computer, I normally get an upload speed of around 20 mbps and I normally get a download speed of around 40 mbps. In my opinion, that's just about as real world as it gets. By those standards, the most accurate results I got from all those snapshot speed tests I just ran above were from Speakeasy. Actually that doesn't surprise me. I once went through all thirty-odd speed test sites I had bookmarked and got an average from all of them, on the assumption that a derived average would be accurate because the individual speed tests tended to be accurate (i.e. were not biased for too much speed or too little speed). And Speakeasy's results were closest to the average of all thirty tests.

But I always prefer real world tests. Another superb test is a real live video stream. Those were very valuable until connection speeds over 15 mbps started to become available. For example, there is one excellent 1280 x 720 HD live stream from Czechoslovakia at »server3.streaming.cesnet.cz/hdtv···inning=0 that runs at 2.2 mbps which I used to use all the time to test my connection . But currently those streams are of less value than they used to be because there are no high speed video streams that play much above 10 mbps, so they no longer are capable of testing the limits of the speedier modern connections. Perhaps, down the road, as 4K becomes the new video standard and live video streams need to play back at much faster speeds, video streams will once again become useful real world tests for connection speed and stability.



Jim_in_VA

join:2004-07-11
Cobbs Creek, VA
kudos:4
Reviews:
·Metrocast Commun..

1 edit

The defacto measure of internet speed is »www.speedtest.net/ whether you agree or not with the method, it is the standard that the vast majority of websites and ISP's use as the benchmark.
--
... need help? »evdo-tips.com/


criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom

In general, I'm O.K. with Speedtest.net, though, like I say, Speakeasy seems to run slightly closer to my real world experiences.

However neither of those are seriously helpful, in my opinion, in gauging the reliability or stability of a connection, to a not inconsiderable extent because of their relatively short test durations. That's one of the reasons why I also like the ISP Geeks test. That test at least takes a stab at evaluating stability and reliability as well as speed.

Are there other good tests for quality of service? For example, is the jitter test at pingtest.net doing pretty much the same thing as ISP Geeks? Or is that a much narrower parameter? Is the ISP Geeks test more global?

Years ago, someone (was it espaeth?), who I believe worked on ISP Geeks, posted a few detailed discussions of the ISP Geeks tests here on this site. Unfortunately I don't recall much detail from those postings and I can't seem to find them now.



Jim_in_VA

join:2004-07-11
Cobbs Creek, VA
kudos:4

you are entitled to create your own world
--
... need help? »evdo-tips.com/


criggs

join:2000-07-14
New York, NY

1 recommendation

Only on the quantum level.



Jim_in_VA

join:2004-07-11
Cobbs Creek, VA
kudos:4

so you are good to go



dib22

join:2002-01-27
Kansas City, MO
reply to criggs

Don't be branching off to your own universe and leaving us behind.


criggs

join:2000-07-14
New York, NY

Yep, can't be in two places at the same time; it's the principle of the thing. Well, actually, I'm...uncertain about that...



Goliath2k
Premium
join:2013-12-28
united state
reply to criggs

I like testmy.net . It's quicker than speedtest.net and appears to be more accurate.


criggs

join:2000-07-14
New York, NY

On what basis did you reach the conclusion that it is more accurate than speedtest.net?



Goliath2k
Premium
join:2013-12-28
united state

The speeds that it gives appear to be closer to my real-world download speeds.


criggs

join:2000-07-14
New York, NY
Reviews:
·Millenicom

Well, I decided to give testmy.net another go. But I had the same problems as those I reported above.

Then I poked around the site and found the multithread option and decided to give that a whirl.

BINGO! Suddenly I started getting download results that were comfortably in line with those from the other speed test sites, 48 megs, 50 megs, and so on.

But then I bumped into a brick wall on the upload test. I was unable to figure out how to run an upload test with multithreading. It would be stupid for the site to offer multithreading and then limit it to download tests, so I'm assuming I'm just being stupid. Could you please explain how I can perform an upload speed test on that site with multithreading? Thanks!



Goliath2k
Premium
join:2013-12-28
united state

Glad to hear that you liked testmy.net! Unfortunately, I don't know how to do a multithreaded upload test...I've never even tried an upload test.

I'm generally not too concerned with our upload speed, as long as it's not something like 0.08 Mbps (which I believe we were getting during our AT&T Home Fusion Trial).