-
After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.
Root [ALL] FIX your 3G Speed NOW! [PLEASE READ]
- Thread starter g60madman
- Start date
-
- Tags
- motorola triumph
HSPA?
buzzcon
Android Enthusiast
I see everyone uses the Speedtest.net app. It seems to be all over the place in regards to actual measurements. Possibly due to the fact that it is ad supported? I tried the FCC Mobile Broadband Test and it seems to give a more consistent reading, and I believe a more accurate reading. It appears to be based off of the same test as Speedtest.net (Ookla).
Teklogikal
Newbie
Metro Detroit area, running Sharp 2.2 rev6,
Before-
First test
Ping:158ms
DL:80 kpbs
UL:35 kbps
Second test
Ping: 170ms
DL:25 kbps
UL:32 kbps
After-
First Test
Ping:278ms
DL:277 kbps
UL:368 kbps
Second test
Ping:316ms
DL:284 kbps
UL:294 Kpbs
Before-
First test
Ping:158ms
DL:80 kpbs
UL:35 kbps
Second test
Ping: 170ms
DL:25 kbps
UL:32 kbps
After-
First Test
Ping:278ms
DL:277 kbps
UL:368 kbps
Second test
Ping:316ms
DL:284 kbps
UL:294 Kpbs
I applied the fix, but living in the backwoods with no signal anyway I won't be able to tell a difference until I go to town later. But if it really works I'll definitely know, I usually don't get good signal until I'm in or near downtown. I'll update in a while, thanks
Edit: Noticed a difference soon as I got out had better signal in places that I usually don't and Pandora streams much better, thanks g60madman awesome work!!
Edit: Noticed a difference soon as I got out had better signal in places that I usually don't and Pandora streams much better, thanks g60madman awesome work!!
Grumpy Zombie
Newbie
Does this fix work on any of the CM9 roms?
G
g60madman
Guest
As far as I know the answer is no. However if you want to try it and post back it would be useful. So far Froyo Sharp, and CM7 ROMS have been solved.
G
g60madman
Guest
Really? Did you download the new ROM I released with the 3g fix?
Has anyone actually looked at what these properties are used for?
ro.telephony.call_ring.multiple is a property for vendor RIL's that send multiple UNSOL_CALL_RING notifications. It likely defaults to false so removing it won't change anything. Even if it did, it wouldn't affect 3g at all.
persist.sys.use_dithering is a property that affects image quality/rendering. In short, it doesn't remotely affect 3g.
To me, ro.telephony.ril.v3 looks like voice call settings but I guess it is possible it could affect 3g.
Take it for what its worth- I personally don't think it would fix 3G speeds but it shouldn't hurt to try
ro.telephony.call_ring.multiple is a property for vendor RIL's that send multiple UNSOL_CALL_RING notifications. It likely defaults to false so removing it won't change anything. Even if it did, it wouldn't affect 3g at all.
persist.sys.use_dithering is a property that affects image quality/rendering. In short, it doesn't remotely affect 3g.
To me, ro.telephony.ril.v3 looks like voice call settings but I guess it is possible it could affect 3g.
Take it for what its worth- I personally don't think it would fix 3G speeds but it shouldn't hurt to try
G
g60madman
Guest
You maybe right, of course I have seen a huge increase 2 days now using the fix. The whole thing came about that a users said he was getting screaming speeds with the TG ROM and said with g60 ROM he was getting slower speeds. So I dug into it and TG was using those settings I recommend. Issac changed the settings for what ever reason. I didn't take the time to look through his billion commits but after doing multiple tests with a friends MT phone who hasn't applied the fixes. I was getting better speeds than him each time we ran speed test for about 10 minutes at lunch.
Yeah, those settings are necessary in cm9 for the cm9 code to be able to talk to our froyo ril drivers. Basically, they're legacy calls that are used by legacy ril drivers. That's why you see most cm9 roms for older phones using those lines.
On the surface, what isn't related to 3G would unlikely affect 3G. But this is a smartphone. Everything is controlled by a combination of software/firmware and hardware. If by chance, a certain low level function that has a higher priority than the functions tasked to handle the funneling of data received by the radio to the memory - is not acting appropriately, then the 3G data could be hindered.
bobboberts
Newbie
An average implies that there is a standard speed you are trying to measure despite inaccurate measurements. But your measurements are accurate. It is the network capacity that is fluctuating. An average is pointless in this situation. You want to distinguish the effects of this mod from the effects of fluctuating network traffic. You want the maximum your device can download irrespective of the network traffic. If this mod works, it should increase that speed. The mod can't unclog the network. If all other network traffic were to be removed, what speed could your device attain? That is the question to be answered. And unless the PRL connects you to a tower closer or further away from you, it will have zero effect.
Think about as if you are changing the engine of a car. Do you measure the increase in performance by finding the average speed the car travels through rush hour traffic? No, you take the car out on the open road and floor it.
These speed tests are pointless because you don't know if you are in traffic or on the open road. You need many data points to figure it out.
activeservo
Member
I'm running Sharp v4 and wanted to try this for myself. I couldn't see those lines in the build.prop so I added them where they looked like they would belong. I tested 5 times twice and these are my results. I'm in Los Angeles. I rebooted in between changes and let it sit for 3 minutes. I'm on the original PRL.
(Original)
DL UL
212 834
781 792
311 720
395 668
363 764
(Lines Added)
DL UL
903 828
1466 842
1024 813
1182 831
1123 833
---------------------
Round 2
(Original)
DL UL
683 729
637 660
836 797
889 748
775 829
(Lines Added)
DL UL
511 812
1301 667
828 834
1500 853
1302 835
Conclusion: The results still varied quite a bit but overall I hit much higher numbers with the modified build.prop and I'm going to keep it unless my battery life gets worse or something. I don't think it will hurt anything so why not?
(Original)
DL UL
212 834
781 792
311 720
395 668
363 764
(Lines Added)
DL UL
903 828
1466 842
1024 813
1182 831
1123 833
---------------------
Round 2
(Original)
DL UL
683 729
637 660
836 797
889 748
775 829
(Lines Added)
DL UL
511 812
1301 667
828 834
1500 853
1302 835
Conclusion: The results still varied quite a bit but overall I hit much higher numbers with the modified build.prop and I'm going to keep it unless my battery life gets worse or something. I don't think it will hurt anything so why not?
No not at all, if this was the case the whole phone would be slow not just 3G. Even if there was something that was hogging cpu cycles the OS would still schedule cycles for 3G (as it would for other things). 3G requires a minuscule amount of time from the processor.
The fact is that data is inconsistent and any possible fix is going to yield results whether it does anything or not. I could tell you to take the battery out of your phone spin around 3 times and put it back in and your data would be forever fixed and if people believed me and tried it, they would come back with results.
I'm not saying this fix definitively doesn't work, I was just sharing the knowledge I obtained just quickly researching what those changes actually do. Again just take it for what its worth.
makosinski
Newbie
Something to think about. Virgin Mobile's network is Sprint's network throttled so that Sprint users get priority on the towers. I've noticed in my area (Orlando) that when there are more people here (potentially more Sprint users) my throughput goes into the bucket. During periods when there are less people here my throughput increases. I believe that no matter what you do the network itself is going to control what you get. It would be interesting to note the time of day people are getting higher throughput. During low demand hours I would venture a guess that throughput would be higher, during high demand periods lower. I don't think there is anything that could be done legally to circumvent the carrier throttling...
Think about what you wrote... the 3G requires a minuscule amount of time... so any process that could interrupt the 3G could also be miniscule and would not necessarily bog the phone down. But it may not even be a function that is bogging the cpu down, it could be a service that corrupts the data or makes the 3G think the data is corrupt and causes it to be thrown out. Data needs to be resent and thus throughput is lowered...
Probably not the answer, but just how things could happen.
Probably not the answer, but just how things could happen.
SyberTiger
Newbie
I put my stock ROM back on so I could update the PRL. Just as a point of reference I'm posting my before and after 3G speeds with the stock ROM before and after the PRL update. I'll be trying out the OP's 3G fix soon on a custom ROM to see how the 3G speeds change.
I ran the SpeedTest 6 times before and after the PRL change then averaged. I really did not expect there would be a difference here at the house (Metro Orlando). As you can see my download speed near tripled. I have no explanation for this but the results are the results.
PRL = 61007 (old stock PRL)
Down = 540kbps Up = 495kbps Ping = 344ms TIme of Day: 11AM
PRL = 60685 (a more up-to-date PRL)
Down = 1537kbps Up = 666kbps Ping = 349ms Time of Day: 2PM
* NOTE: After the PRL update I turned the phone off and pulled the battery. Then waited 10 minutes before rebooting and running SpeedTest.
I ran the SpeedTest 6 times before and after the PRL change then averaged. I really did not expect there would be a difference here at the house (Metro Orlando). As you can see my download speed near tripled. I have no explanation for this but the results are the results.
PRL = 61007 (old stock PRL)
Down = 540kbps Up = 495kbps Ping = 344ms TIme of Day: 11AM
PRL = 60685 (a more up-to-date PRL)
Down = 1537kbps Up = 666kbps Ping = 349ms Time of Day: 2PM
* NOTE: After the PRL update I turned the phone off and pulled the battery. Then waited 10 minutes before rebooting and running SpeedTest.
I've never tried to change my PRL, if it was easy as flashing I'd be more interested in doing it
Think about what I wrote? You obviously have some misunderstanding about scheduling and the incredible speeds at which processors work. If an interrupting process is miniscule then the processor would handle it and could then return and no human being could recognize the difference. Your computer and smart phone do this all the time- they switch between tasks at incredible rates. Its not like its possible for a process to only interrupt and starve the 3G process and nothing else.
You were saying that a misbehaving function that that would bog down 3G would bog down the entire phone. I am merely saying it doesn't have to be a process 'hog' to cause 3G to slow down. (I am not talking about cpu capacity).
Think of a basic function that somehow had a memory overflow or flipped a bit in the data encoding, that packet has to be thrown out and resent.
Think of a basic function that somehow had a memory overflow or flipped a bit in the data encoding, that packet has to be thrown out and resent.
Could you please share the new PRL? thanks!