[AUG 4] *12 Radios Available* Updated Touch Pro CDMA Radio Archive Thread

[MACE]

Originally Posted by obamaftw08

You know, I can sum up everything you just said without needing a disclaimer about rudeness and/or being highly demeaning, in a much more refined form.

Hello fellow radio testers. I'm gonna try and put up a guide here soon to let you know about some other testing scenarios/guidelines in which I think would be helpful for everyone here to figure out what works best for each person individually. In the meantime let me find a link that will give everyone an idea of what I'm talking about.

See? Easy peesie. No thanks needed.

Unfortunately, that does not address the major concerns that were my reason for posting.


I wanted to cull the the spread of misinformation before it got out of hand, as it almost always does when discussing this topic.

I wanted to list the most common sources of inaccuracy, and explain *why* they were inaccurate, so that people reading would stop relying on them.

I also wanted to provide a non technical way that ppl could test the radios, without needing a degree in network engineering, and wireless physics, that would still provide a *VERY* meaningful, and helpful contribution to the community as a whole, as well as for themselves.


That being said, I re-read my post, and while I certainly did not intend to sound demeaning, and don't really think I was tbh, there are phrases I could have worded a bit more softly, and would have changed some of the wording in the second paragraph in particular.
For that, I apologize.

[MACE]

Originally Posted by psycho_maniac

if you had to answer the same type of questions i think you would be a little testy also

It was not questions I was worried about so much, as having many ppl being mislead by inaccurate and in many cases, nearly meaningless information. I don't mind answering questions, and actually like to help ppl if I can, as long as they actually put in a bit of effort and don't just expect to have their hand held.
But thanks for the support

Originally Posted by havox22

i want a guide & what radio do u use???

I am currently testing the new Altell one, but I dont have enough info yet to make any kind of determination, just *barely* flashed it.

[havox22]

i want a guide & what radio do u use???

Originally Posted by MACE

to long to really quote

[MACE]

I swear, sometimes the search function of these forums is so useless.

I've wasted way too much time looking for that guide I mentioned I had seen, I should have just taken the time to write my own
But now its almost 2:30 am here, and I was tired before I got started...

Here is one pertinent post I found however, which you can probably get by with until I can write a more comprehensive guide.

Originally Posted by snovvman

I don't know how many times this has been discussed, but I guess here it goes again.

The bars are not nearly accurate enough to make conclusions about signal quality.

Here is another twist, on my Diamond, in a relatively weak DO area but connected to an Airave (1x), the bars display the state of DO (therefore, few bars even though I'm connected to Airave with good 1x service. On my Touch Pro, on the other hand, the bars display the state of 1x (again, connected to Airave) even though the indicator shows "EV". That is to say, even though DO is weak and 1x is strong, the TP will show you "EV" while the bars are actually showing 1x signal strength. In simple terms, two phones, side-by-side, same coverage quality, one has many bars, the other nearly none.

In other words, the Diamond and Touch Pro shows you signal quality on the bars differently. Got that? Remember, there are both 1x and DO portions of the radio at play.

You MUST use ##DEBUG# (go to the dialer, enter ##33284#) to accurately determine signal strength. Set your Sample Rate to 1 second. If you want to see DO performance, choose Debug Info, then HDR. The bars do not report accurate information.

Also, the signal strengh can bounce A LOT, unless you have an Airave in the next room. You must take samples over time and average the results.

Next, I had both phones running Telus 1.04, and using the above method, side-by-side, they report very very close Rx Power and Rx Ec/Io figures (again, even though the bars are bouncing and don't agree). They are consistently within 1 to 2 points of each other both in 1x and HDR, dispite what the bars report.

I then flashed 1.11 on the Touch Pro and ran the field test again. Now, side-by-side, the Touch Pro ALWAYS show a 3-7 point weaker signal strengh when compared to 1.04. In HDR (DO), the performance is about the same. Mind you, this is when TP is showing full bars.

Although that 1.11 shows less quality connection compared to 1.04, I have not observed a degredation in battery life nor call quality. They seem very close to me in terms of performance. DO speed seems to slighly improve with 1.11.

Something very important to understand: the Rx signals are measured in decibel drop where the closer to zero, the better. -100dbm is bad!
Typically, with EVDO, you want to have less than -80dbm for decent throughput, with top end being reached somewhere below -70dbm and anything below that giving lower latency, and higher stability. In real world usage, -50dbm is about good as you are ever going to get, and in most cases, not bloody likely, lol. Although, I have actually been able to achieve -46 to -47 by standing at about a 30° angle from a tower, about 60 yrds away (because of the physical characteristics of typical radio towers, you will actually get *lower* signal strength directly underneath (or above, lol) due to the conical umbrella they broadcast, I got around -65-67 standing at the base of this particular tower, for example)
Also remember, even if you were to get say, -55dbm, which is pretty much incredible signal strength, (might as well plug your phone directly into the tower, ) your throughput can be limited by many other factors, such as the backhaul of the tower, and network load. So, while your phone might be able to communicate with the *tower* at 3.1mbit your network might not be able to provide that. Of course, if it doesn't, there is nothing really wrong with complaining to your provider about it, to give them a little prod to induce them to improve their infrastructure

[obamaftw08]

Originally Posted by MACE

I swear, sometimes the search function of these forums is so useless.

I've wasted way too much time looking for that guide I mentioned I had seen, I should have just taken the time to write my own
But now its almost 2:30 am here, and I was tired before I got started...

Here is one pertinent post I found however, which you can probably get by with until I can write a more comprehensive guide.
Something very important to understand: the Rx signals are measured in decibel drop where the closer to zero, the better. -100dbm is bad!
Typically, with EVDO, you want to have less than -80dbm for decent throughput, with top end being reached somewhere below -70dbm and anything below that giving lower latency, and higher stability. In real world usage, -50dbm is about good as you are ever going to get, and in most cases, not bloody likely, lol. Although, I have actually been able to achieve -46 to -47 by standing at about a 30° angle from a tower, about 60 yrds away (because of the physical characteristics of typical radio towers, you will actually get *lower* signal strength directly underneath (or above, lol) due to the conical umbrella they broadcast, I got around -65-67 standing at the base of this particular tower, for example)
Also remember, even if you were to get say, -55dbm, which is pretty much incredible signal strength, (might as well plug your phone directly into the tower, ) your throughput can be limited by many other factors, such as the backhaul of the tower, and network load. So, while your phone might be able to communicate with the *tower* at 3.1mbit your network might not be able to provide that. Of course, if it doesn't, there is nothing really wrong with complaining to your provider about it, to give them a little prod to induce them to improve their infrastructure

Now that is something I can work with. Thanks.

[t0mmyr]

i know online speedtests cant be too accurate but they give a rough estimate how well something is performing at a given time, i put a disclaimer in my post mentioning that in no way were these scientific tests....but u can clearly see that the 2 new radios are far faster than the stock rom

[MACE]

Originally Posted by t0mmyr

i know online speedtests cant be too accurate but they give a rough estimate how well something is performing at a given time, i put a disclaimer in my post mentioning that in no way were these scientific tests....but u can clearly see that the 2 new radios are far faster than the stock rom

I was not directing my comments at anyone in particular at all, you see ppl quoting speedtests as *proof* of the superiority of a certain radio, rom, or device, or whatever, all over the place. Sorry if it seemed like I was picking on you, this was not the case at all. TBH, at least you were trying something that was a "real world" test, sort of

As far as the meaningfulness of your particular tests, even though they were done as close to back to back as you could (considering you had to flash in between each) I'm sorry to have say this, but, unfortunately, due to the currently very low capacity of most wireless networks backhaul, as well as a host of other contributing factors, even back to back tests can vary by rather significant amounts, even if you have an *excellent* signal, and everything on your end is working perfectly.
For example, there have been a number of times where I have seen my bandwidth go from 2.5-3Mbps to 1mbps, or even less than 1, and then back to 2mbps or so in a matter of 10 minutes or less. And it is fairly normal for your average bandwidth to vary by at least .1 to .3mbps especially if you are in a metropolitan area.

On perhaps a more helpful note, I did notice your latency was *extremely* high. You should check to make sure it is nothing on your end, (the most common cause of this is spyware) and if not, call your provider, and tell them about it. If it is on their end, they *definitely* need to fix it, 300+ ms on EVDO is pretty bad, and 600+ is pretty much intolerable, and almost guaranteed, if your latency is that high, whatever problem is causing it is most likely limiting your bandwidth significantly as well. I would not be surprised to see up to even a 1mbps increase, if you get that cleared up, unless it is just fuxxed network settings on your machine.

I hope that this helps.

[Edit]
I just noticed that in your sig, that you had over 500 ms latency in one of your Road Runner tests. If this was done any time reasonably recently, it *strongly* suggests a problem on your end, so I would recommend trying a hardcore malware test first, if you come to the conclusion that is not likely, then check your network settings for a ridiculously huge mtu size, this is something that some tweak programs do for a slight boost in download speed, at a large cost in latency, but I have never seen it account for more than maybe 100ms lag at the most tbh, the only thing that I have seen nuke a connection that badly (from 20-30ms to over 500) is pretty much spyware. Your low upload speeds are also indicative of this.

Sorry if this is the case, but hopefully it at least helps you figure it out.

[havox22]

ok i did the ##33284# so which is more important 1x or hdr & what is hdr?

[t0mmyr]

Originally Posted by MACE

[Edit]
I just noticed that in your sig, that you had over 500 ms latency in one of your Road Runner tests. If this was done any time reasonably recently, it *strongly* suggests a problem on your end, so I would recommend trying a hardcore malware test first, if you come to the conclusion that is not likely, then check your network settings for a ridiculously huge mtu size, this is something that some tweak programs do for a slight boost in download speed, at a large cost in latency, but I have never seen it account for more than maybe 100ms lag at the most tbh, the only thing that I have seen nuke a connection that badly (from 20-30ms to over 500) is pretty much spyware. Your low upload speeds are also indicative of this.

Sorry if this is the case, but hopefully it at least helps you figure it out.

im sort of located in the desert far off from Los Angeles but not close to San Diego too, so when i pick a location i usually get a better ping from San Diego:

but sometimes it seems the Los Angeles servers are faster coming my way, even with an unsually high ping
speed not being the case today.

same when i tether my phone, sometimes speedtest will think i'm in Kansas (I assume this is the main network hub for Sprint since I think their corporate offices are here too) and Palm Springs to Kansas is a long way for a cellphone data packet to travel so the ping gets up there pretty good, usually if I choose Los Angeles it will be under 100ms. I pay for 15down/2up roadrunner so my upload has always been maxed right around 1.93-1.94mb which i think is on par for what i pay for, the download usually surprises me at times, things fly by my screen when downloading like nothing i don't think i have spyware on either my pc or laptop as the speedtest usually average out about the same, 1 running vista ultimate the other running windows 7, both using eset nod32 64bit av.

[Maxx134]

One of the things I notice may be a good test is how battery life is with a particular radio.
Only problem with that is that the rom you are using has a bigger impact on battery, so you should not change rom OS when testing.

That and "dead zones" is prob a good test.
I'll never 4get the funny post in an old thread where the dudes testing location was on his toilet seat.
If it past that test, then it was good.