True Ultra High Speed Cable Service 50 - 100 Mbps

[jamiesensei]

Am in the process of getting a 100 Mbps cable internet connection from True as detailed here : http://www.trueonline.com/th/product/ultra_hispeed.php

I thought that the 50 Mbps and up internet connection deserved a new topic thread although there is already a very long thread on the True cable service here : http://www.thaivisa.com/forum/topic/427295-true-ultra-hi-speed-internet/page__hl__true+internet+cable It seems that 50 Mbps and up you get a DOCSIS 3.0 modem rather than the DOCSIS 2.0 provided for customers using 10 - 20 Mbps.

I have been provided with the Thomson TCW 770 modem as shown here : http://support2.truecorp.co.th/detail.aspx?document_id=317

The technicians have been round repeatedly and we have found that you need to disable the wifi access point in the Thomson TCW 770 to get higher speeds. The technicians actually managed to get 105 Mbps speed test results from speedtest.net but with the wifi access point enabled speeds from the lan cable drop to 20 Mbps despite having nothing connected on wifi. Have found it better to use my own wifi router plugged in to the lan ports of the TCW 770 modem and then I get speeds approaching 90% of what is available on the lan port over wifi, although the signal for my wifi n router does not seem to reach all over my large house.

Speedtest.net speeds have now dropped down from the 105 Mbps speed I was seeing when the technicians were here last yesterday to 30-50 Mbps and I see speeds of 10-17 Mbps as measured on dslreports.com to L.A.

Am trying to get True to give me a CISCO router instead of the Thomson one which I don't have much confidence in. If they would provide me with one of these CISCO EPC3010 shown on this page http://support2.truecorp.co.th/detail.aspx?document_id=316 without the wifi and then I bought my own good quality wifi n router or a router that used the power lines that would I imagine be much more reliable and might even squeeze an extra few Mbps out of the cable??

I wonder if I would get similar speeds if I dropped down to the 50 Mbps plan at 2,800B instead of the 100 Mbps plan. I think I might try that and then I can always upgrade if I want to.

The True technicians and True in general have been very helpful. They do not know there way around this technology very well and do seem to try various settings out in a rather random way to see if it improves any problem but they have said that I do not need to sign the contract until I am getting a satisfactory service. They seem to expect that I should be seeing 100 Mbps speeds on speedtest.net and apparently on the 10 Mbps service customers normally get near the maximum 10 Mbps speedtest.net results which would mean pretty high real speeds. I get the impression that they have not installed a lot of 100 Mbps connections as yet and are not really sure what kind of speeds to expect.

At True I was told that it is possible once you have signed your contract to upgrade speed at any time and to down grade after 6 months. I am not sure if this would mean restarting a new contract though. I presume the cable contract is like the ADSL one which I believe you can quit with no penalty after one year.

Will try to post follow ups on what kind of speeds and set up I end up with. Would be very interested in hearing of others experience particularly with 50 Mbps and up and also perhaps people could fill me on some of the areas above I am not sure of??

[Pib]

I'm on the True cable 20Mb/2Mb plan here in western Bangkok...been on the plan since mid June 11...been very pleased. I've also got some comments about faster plans and various Wifi routers.

I started off with the Cisco EPC 2325 as shown in your/this Link. But approx 3 weeks ago in doing a routine/quick speed test to a Bangkok server (I do a quick test almost daily just to make sure things are speeding along at full speed) I was only getting 1.3Mb Upload where I normally getting 2.2Mb. I waited a day to see if the problem would clear up; it didn't and I called True.

Got very prompt/good service, but when the dust settled it turns out that apparently a recent upgrade to the True's cable/DOCSIS internet system had caused a firmware compatibility problem with some Cisco EPC 2325's. I watched over a couple of visits over a 3 day period by three different technicians to check/compare modem/routers, line to my house, etc., that the problem was only with the Cisco EPC 2325. They were telling me I was the first to report this upload speed problem, others had just begun to report it, and it seemed to be limited to certain Cisco models. I watched as they hooked up their test Cisco EPC 2325 and a new one out of the box....both couldn't achieve the 2Mb upload speed via ethernet port or Wifi connection...were limited to around 1.3 to 1.6Mb. They then hooked up a non-Wifi Cisco model and got the 2Mb upload no problem. The following day they brought out two more new EPC 2325s and had the same problem. They could have given me a Thomson TCW 750-4 Wifi router (apparently the Thomson Wifi router they now offer for plans up to 20Mb) on the first visit but 1 of the 2 technicians said some didn't feel the Wifi signal/connection was as good as the Cisco and if I wanted to wait a few days he could get me another Cisco 2325...but that was on the first technician visit and as mentioned over the coming days they determined that a recent system upgrade was causing a upload speed incompatibility problem with some Cisco EPC 2325's...maybe other Cisco's also.

During one visit a tech hooked up a non-wifi Cisco modem/router to check the line max speed. He got almost 105Mb download and 10.4Mb upload which showed the line could met the 100Mb/10Mb advertised speed. I wish at that time I had asked the tech if I could have run some speedtests to international web sites but I didn't. But I know I would have got the typical 3-6Mb download speeds like I get on my 20Mb plan and what I've seen others post on various blogs for 30, 50, and 100Mb plans...that is real world results versus the faster-than-light bogus results. To get real world results on True you have to use one of their proxy servers to prevent hidden cache servers skewing the speed test results on various speedtest programs...results like that far off server in FarangLand is just across the soi (when actually it's a cache server in Bangkok).

Anyway, when the dust settled I was given a Thomson 750-4 Wifi router, felt a little unsatisfied for a few days thinking a Cisco model would have been better, but after three weeks of use I'm no longer a little unsatisfied because the Thomson Wifi router gives me a better/more stable wifi connection than the Cisco Wifi router did...even before the firmware issue came up. By better, I mean I can "always" get a 20.4Mb down/2.2Mb up speeds with my Wifi router upstairs and my laptop downstairs where with the Cisco sometimes it would give me varying download speeds in the 15-20Mb ballpark...but I move the laptop close to the router and I get the 20Mb...and I would always get 20Mb if hooked directly to a Cisco ethernet port. I also like the Thomson menu interface a little better. The Thomson also runs cooler/less heat generated.

Now, with that story out of the way, when it comes to actual Wifi "throughput" just because say a Wifi router, say like an N router, says it has up to 100 to 300Mb throughput that is the theoretical and not real world throughput. I'll let you google for wifi router reviews on "throughput" but I think you will find that when dealing with internet plan speeds of over 20Mb that a Wifi routers real world "Wifi throughput" may not live up to it theoretical specs...ratings in a very controlled laboratory environment. Preaching to the choir I'm sure. Here's one Link that shows Wifi throughputs for various Wifi routers. Seems reaching a 100Mb Wifi throughput is pretty hard for any Wifi router to achieve unless real close to the router.

In closing, do you have a real need for 50-100Mb download speed or just an urge to have the fastest? Urges are fine but this urge may not be satisfied depending on how you plan to use the connection. For single-threaded type operations like browsing, emailing, video streaming you are still just going to get 3-6Mb "real world" speeds to US/European international web sites since True, like other ISPs, limit the bandwidth to international web sites regardless of your plan's speed (i.e., whether a True 10Mb plan or a 100Mb plan). And if DSLReports.com web site results are correct then you will only get 1-2Mb real world speeds to US/European international web sites. You can get higher bandwidth to some regional locations like Singapore. Now with multi-threaded operations like when using a download manager, torrents, etc., you will get higher download speeds...maybe all the way to 100Mb with a 100Mb plan...if that far off server allows full throttle/is capable of 100Mb output...many will limit their max output speed to something much lower. If your goal is get more bandwidth to international web site using single threaded operations it probably ain't going to happen. Cheers.

[jamiesensei]

I seem with my current configuration to get around 10-17 Mbps across wifi with my current 100 Mbps connection. I believe these are real world download speeds. Apparently the tests at http://dslreports.com, according to someone else's post I read on here, cannot be speeded up by a cache as each download is uniquely randomly generated.

[jamiesensei]

During one visit a tech hooked up a non-wifi Cisco modem/router to check the line max speed. He got almost 105Mb download and 10.4Mb upload which showed the line could met the 100Mb/10Mb advertised speed. I wish at that time I had asked the tech if I could have run some speedtests to international web sites but I didn't. But I know I would have got the typical 3-6Mb download speeds like I get on my 20Mb plan and what I've seen others post on various blogs for 30, 50, and 100Mb plans...that is real world results versus the faster-than-light bogus results.

I would also be very interested in what other people have to report about the speeds they see on http://dslreports.com I think I am no longer getting the maximum throughput on the line of 100 Mb/ 10 Mb although while the tech was here I did see repeatedly we reached 105 Mbps.

I would think if I am now getting 30- 50 Mbps on speedtest.net compared to 105 Mbps before then presumably it is possible that my real world speeds could also double if I can get the maximum out of the line.

To get real world results on True you have to use one of their proxy servers to prevent hidden cache servers skewing the speed test results on various speedtest programs...results like that far off server in FarangLand is just across the soi (when actually it's a cache server in Bangkok).

I am quite persuaded by the idea I have seen elsewhere that the download data from dslreports.com is uniquely generated for each speed test request and thus uncacheable.

[jamiesensei]

Anyway, when the dust settled I was given a Thomson 750-4 Wifi router, felt a little unsatisfied for a few days thinking a Cisco model would have been better, but after three weeks of use I'm no longer a little unsatisfied because the Thomson Wifi router gives me a better/more stable wifi connection than the Cisco Wifi router did...even before the firmware issue came up. By better, I mean I can "always" get a 20.4Mb down/2.2Mb up speeds with my Wifi router upstairs and my laptop downstairs where with the Cisco sometimes it would give me varying download speeds in the 15-20Mb ballpark...but I move the laptop close to the router and I get the 20Mb...and I would always get 20Mb if hooked directly to a Cisco ethernet port. I also like the Thomson menu interface a little better. The Thomson also runs cooler/less heat generated.

Interesting, thanks. Though as I said with the 770 model I have when you enable wifi the speed of the lan connection drops by more than half! With no wifi connections too. I believe their is a single processor in their used by both the wifi and modem that is getting overloaded?? Or there is interference from the wifi interfering with the modem?? Would like to try out some CISCO devices.

Now, with that story out of the way, when it comes to actual Wifi "throughput" just because say a Wifi router, say like an N router, says it has up to 100 to 300Mb throughput that is the theoretical and not real world throughput. I'll let you google for wifi router reviews on "throughput" but I think you will find that when dealing with internet plan speeds of over 20Mb that a Wifi routers real world "Wifi throughput" may not live up to it theoretical specs...ratings in a very controlled laboratory environment. Preaching to the choir I'm sure. Here's one Link that shows Wifi throughputs for various Wifi routers. Seems reaching a 100Mb Wifi throughput is pretty hard for any Wifi router to achieve unless real close to the router.

I was getting reported speeds of 40 Mbps through my wifi at the same time as 50 Mbps on my LAN reported on speedtest.net. Although these results are presumably cached it shows that wifi can perform when using a good device. I think if I am reaching real speeds of 30-60 Mbps high enough to reach the speed where wifi n becomes a bottle neck I will be doing very well.

In closing, do you have a real need for 50-100Mb download speed or just an urge to have the fastest? Urges are fine but this urge may not be satisfied depending on how you plan to use the connection. For single-threaded type operations like browsing, emailing, video streaming you are still just going to get 3-6Mb "real world" speeds to US/European international web sites since True, like other ISPs, limit the bandwidth to international web sites regardless of your plan's speed (i.e., whether a True 10Mb plan or a 100Mb plan). And if DSLReports.com web site results are correct then you will only get 1-2Mb real world speeds to US/European international web sites. You can get higher bandwidth to some regional locations like Singapore. Now with multi-threaded operations like when using a download manager, torrents, etc., you will get higher download speeds...maybe all the way to 100Mb with a 100Mb plan...if that far off server allows full throttle/is capable of 100Mb output...many will limit their max output speed to something much lower. If your goal is get more bandwidth to international web site using single threaded operations it probably ain't going to happen. Cheers.

In the time I have had this connection I have watched quite a bit of online video from ted.com and youtube.com. There is more and more web based instant access video content. It is really nice to have those videos play instantly all the time with no buffering issues. I work online and I get most of my entertainment and increasingly work related educational video content off the internet. I can do without a really fast connection but it is a nice luxury to have. Stability is probably more important than speed. I may plump for the 50 Mbps connection as I suspect the actual speeds delivered will be similar to the 100 Mbps at a bit more than half the price.

Thanks a lot for your detailed reply Pib!

[Pib]

I seem with my current configuration to get around 10-17 Mbps across wifi with my current 100 Mbps connection. I believe these are real world download speeds. Apparently the tests at http://dslreports.com, according to someone else's post I read on here, cannot be speeded up by a cache as each download is uniquely randomly generated.

Not bad...I think that is the highest True cable speed to L.A. I've seen. But ain't it a shame it's still only a little more than 10% of the advertised in-Thailand speed for the 100Mb plan. You would think based on the cost of the plan True would open up the international pipe wider. When looking at speed test results on www.khunwoody.com site in Phuket, only one person had reported any data for the True 100Mb plan and he reported on two tests with DSLReports download speeds to L.A. of only 1.814Mb and 1.371Mb. See this Link. Many folks have reported on the 10Mb and 20Mb True cable plans and the great majority seem to get in the 1.5 to 2.0Mb range to L.A. And fairly recently I exchanged a few posts with someone who had just got the 30Mb True cable plan and he reported the same international speeds as the 10Mb and 20Mb plans.

Since 10Mb is 10% of the 100Mb plan and 2Mb is 10% of 20Mb plan, it kinda helps to confirm that Thai ISPs seem to have a bandwidth throttle of something around 10% of in-Thailand speed and/or 2Mb ballpark if the DSLReports.com speed test is accurate.

I just ran the Flash version and got a 2.053Mb down/861Kb Up with 227ms ping. Almost idenitial when running the Java version. I only decided to run the Java version because I though I would be reporting a download speed of only around 1Mb which has always happened previously and then make a point about how the Flash and Java versions give me big time differences to the same server.....I've probably run the Java based version at least a half dozen times before today and always got a much lower number...but today it gives me almost an identical number...and I do run the latest version of Java....oh well, that speed tests for you.

It would be interesting to know what you get when using Speedtest.net, however, but, only if using one of the True proxy settings since the True proxy servers seem to prevent the skewing of Speedtest.net results to report a local cache version (i.e, the faster than light results). Proxy info is:

proxy.trueinternet.co.th port 8080

or

proxy.asianet.co.th port 8080

[jamiesensei]

I am now getting around 18 Mbps download speeds reported from dslreports.com and 70-105 Mbps on speedtest.net! Am not sure why.

But interestingly if I use either of the two proxy addresses you gave me the speeds come down to 1.8 Mbps on dslreports.com and 10 Mbps on speedtest.net

I am thinking that the proxy is slowing things down considerably rather than giving the accurate uncached speed at these speeds.

[jamiesensei]

Oh!

It seems things are fixed now. I am getting the full 105-108 Mbps from speedtest.net results, which are probably cached and not real download speeds.

I am getting 12-22 Mbps on dslreports.com

And I now find that I can turn on the wifi without it affecting the download speeds. Things seem to have fixed themselves. I think some has been tinkering with the system at the local exchange box or something before the technicians come again tomorrow.

[Pib]

I am now getting around 18 Mbps download speeds reported from dslreports.com and 70-105 Mbps on speedtest.net! Am not sure why.

But interestingly if I use either of the two proxy addresses you gave me the speeds come down to 1.8 Mbps on dslreports.com and 10 Mbps on speedtest.net

I am thinking that the proxy is slowing things down considerably rather than giving the accurate uncached speed at these speeds.

That 70-105Mb is a bogus/cached result. Interesting with the proxy turned on the speedtest.net results gave you a 10Mb speed which basically matched your earlier DSLReports speed of a little over 10Mb.

I don't think the proxy is slowing things down. I just run the DSLReports speedtest twice...1st time proxy on; 2nd time proxy off....got basically the same results of 1.8Mb. When I first got True cable I use to use the proxy setting the majority of the time because it made my browsing noticeably faster, eliminated a couple U.S. govt site access problems, and prevented speedtests from getting skewed. I stopped using the proxy settings, except when having site access issues, in late 2011 because I couldn't tell any difference in browser/surfing speed, those U.S. govt access problems just went away, and I rarely did international speed tests...only my almost daily speed test to a Bangkok server to ensure my in-Thailand connection was 20Mb/2Mb. DSLReports ain't perfect in its results....last year I could pull almost a 20Mb download speeds (bogus) from its Denver and Chicago servers but I can't now.

[Pib]

Oh!

It seems things are fixed now. I am getting the full 105-108 Mbps from speedtest.net results, which are probably cached and not real download speeds.

I am getting 12-22 Mbps on dslreports.com

And I now find that I can turn on the wifi without it affecting the download speeds. Things seem to have fixed themselves. I think some has been tinkering with the system at the local exchange box or something before the technicians come again tomorrow.

Cool....and I know I'm preaching to the choir in that a person needs to take many speed test results (Speedtest.net, DSLReports, etc.) with a BIG grain of salt since so many factors can cause bogus results.

[Pib]

You may want to give Microsoft Online Services speedtest a try...supposedly it's not skewed/fooled by cache servers. Go to this Link and about halfway down the page will have the Network Connectivity Analysis links where you can run the speedtests against a U.S. server (in LA I think), a Singpore server and a server in Ireland.

I'm most interested in what you get to the U.S. and Singapore. The Singapore test results should be significantly higher than the results to the U.S. since supposedly there are more direct fiber optics connections between Thailand and Singapore...and Singapore is also becoming a major location for content providers to host their content which greatly speeds up browsing/file downloads compared to having to reach out to a US/European server to get the content/files....this provides a real speed benefit to folks with higher speed plans like you have.

[jamiesensei]

I get 18 Mbps down and 9 Mbps to Singapore. And about 3Mbps down and up to Europe and US from the M$ speedtest..

[Pib]

Ain't it great how various speed test programs can give widely different results even when a person feels the results are not being affected by a cache server. Now you just need to try to figure out which speed test program gives the most accurate real world results without being biased toward picking the one that gives the highest/best results.

P.S. And I don't know the answer as to which speed test program is the most accurate.

[jamiesensei]

But interestingly if I use either of the two proxy addresses you gave me the speeds come down to 1.8 Mbps on dslreports.com and 10 Mbps on speedtest.net

Interesting with the proxy turned on the speedtest.net results gave you a 10Mb speed which basically matched your earlier DSLReports speed of a little over 10Mb.

Using the proxy also slows down dslreports.com's reported download speed from 10 Mbps to 1.8 Mbps.

I think all the speed tests are really testing different things. Speedtest.net is probably testing the speed and latency to True's internal caches and is helpful in it shows that now I am getting full capacity across my cable connection to True?? I think.

Then dslreports is measuring presumably uncached downloads to LA to a particular server and through all the bottle necks it meets on the way. My connection speed to the internet is not a constant thing of course across the whole internet. Then the MS speed test measures download speeds with another set of servers.

Just for the sake of experimentation, I am trying some multi thread downloading from multiple peers and get download speeds of 1.5 - 2.5 MBps ie. 12-20 Mbps. This is when downloading several files with a lot of connections.

[rojack45]

Are there any members with the True Cable Service located in Silom/Surawong area. I have moved there and have a bar/restaurant and wish to stream US sports channels etc. and I have had zero luck trying to find out if Cable Internet is available in my area. Surawong Road at Soi Patpong to be more precise. It took me 6 weeks just to get them to move my regular cable service to the new address. It is a Sat feed so not relevant for the Internet service.

If anyone is using the service and is located anywhere near me, could they either post their general co-ordinates or send me a pm

Thank you

Rojack

[jamiesensei]

I have moved there and have a bar/restaurant and wish to stream US sports channels etc. and I have had zero luck trying to find out if Cable Internet is available in my area.

We found here in Chiang Mai that the staff at the True shop were able to tell us if it was possible to hook up cable internet to any address in Chiang Mai. They had a detailed map of CM on a computer which showed were all the cables went. We asked about 3 separate addresses which were fairly centrally located and they all could get cable, we took along a detailed map of where the houses were located exactly. You may need to take a Thai speaker along to make communicating work smoothly, a map of the sois around you and any significant landmarks might help and your address. Would have thought they would be eager to sign you up, especially if you want one of the 50 Mbps and up packages. Or maybe things are different in BKK?

[muratremix]

is there fair use policy in true ultra packages?

[Pib]

is there fair use policy in true ultra packages?

Yes, see the chart at this Link. For example, on my 20Mb plan it's 5.4GB/Hour.

[mikewet88]

is there fair use policy in true ultra packages?

Yes, see the chart at this Link. For example, on my 20Mb plan it's 5.4GB/Hour.

It is a fair quote. I hardly hit any quote on my 10mb/1mb link..haha

[dr_lucas]

Just a few clarifications, to prevent confusions.

For the less techies among us, let's say that an internet connection's bandwidth is similar to water pipes.

Just because you paid and have a wider pipe, doesn't mean that you are going to get more or stronger water pressure.

You do get a pipe that can theoretically contain the max declared water if all conditions are optimal, and there are no bottlenecks along the way.

To find out if you are getting the speed/pipe you paid for, simply go to:

http://speedtest.trueinternet.co.th/

For the past 3 months I was on the 100mbit DOCSIS3 package and got 105 mbit and about 11 mbit download/upload respectively. (just a note: A few days ago dropped to the 50/5 mbit DOCSIS3 package, as I never really needed to max out my 100mbit package).

This is using Cisco's EPC3010 modem, connected to the amazing ASUS (Dual band 2.4/5 GHz) 450mbit RT-N66U router [HIGHLY recommended]

I can get max speeds over wifi too, when using Trendnet's 450mbit TEW-684UB [also HIGHLY recommended], located a few meters away, and operating using the Dual band mode.

This means that I do get the "pipe width" or bandwidth I was promised. This does NOT mean that I will be able to connect to every server at the top speed, especially not on a single thread connection and especially not servers abroad which do not use a close-by CDN. I was able to max out the 100mbit connection when connecting to several very healthy torrents at the same time, when all the seeders were out of Thailand. This is normal.

There are many factors affecting connection speeds to local/domestic net and even more affecting connections to international net.

The bottleneck which slows down your connection to a server in Thailand or abroad can be sometimes "skipped/overriden" by using a good local or overseas proxy for which you have good routing to and has a good routing to the servers you are trying to reach. This is a workaround, not a permanent solution, as sometimes those proxy servers become too congested by themselves and/or get bad/slow routing or suffer the same or similar bottlenecks you got when you tried to connect directly.

There is also the total ISPs bandwidth congestion. During rush hours, most of the ISPs' international bandwidth become very congested, thus slowing down your connection. Most ISPs offer a 10:1 usage ratio per home-user-line. Corporate users get around 3:1 usage ratio, thus have higher chance of getting better speeds and dedicated line users (very expensive) normally get 1:1 usage ratio from their modem to the final router the ISP owns, which generally provides better routing.

Even the most expensive 1:1 ratio account, may get the same download speeds and connections speeds that a 10:1 home user will get sometimes, depends on routing, the server connecting to and other issues.

That's true everywhere, even in Singapore, Sweden, Japan, S. Korea and the USA, not only in Thailand. (no pun intended)..