True Ultra Hi-Speed Internet

[jfchandler]

BTW, here's the link to the Microsoft Online diagnostics site (which was the source for some of my screen caps above) that provides a lot of useful info, much more comprehensive than just the normal routine speed test site. And of course, it's based in the U.S., so it's giving you stats on your connection to the U.S. west coast.

http://speedtest.microsoftonline.com/

[Pib]

Here's a few pictures of the True Cable TV/internet trunk line and equipment being installed in my moobaan....almost done.

An Amp with Two Inputs and Two Outputs

Signal Tap Point & FeedThrough...The RG-6 Cable That Runs Into Your Residence Connects Here and Then Runs to Your Cable Modem

Signal Tap at End of Soi Trunk Line

Amp Feeding a Splitter Feeding a Tap Point

Amplifier & Splitter Right Outside My House. Amp input on right, amp output on left, feeds into a two way splitter, then the splitter sends the signal down two soi's into signal tap points along the way.

[jfchandler]

I did another round of speed test comparisons last night, 10:30 pm... The speed test results were mixed this time, moreso than the morning ones I previously posted. But the quality of service indicator continued to heavily favor the cable connection.

I should add...though I don't know if it makes a great difference... The cable internet measurements right now are coming from an older PC running Windows XP with all the latest updates and service packs, whereas the DSL results are coming from my main and newer desktop PC with Windows 7 and all the latest updates... I was reading something the other day that indicated the Windows 7 OS has a more efficient (better) TCIP stack compared to XP.... Enough to make any meaningful difference in this discussion, I don't know.

BTW, when the True guy very first connected the cable modem and tested it at my main television point, on a different PC (a new Windows 7 64 bit laptop), I got a 4 Mbps download speed reading that day about 4 pm on a quick test I ran via the DSL Reports L.A. server site... That was with the cable modem just hooked up to my original cable TV line, before the guy later installed a splitter at my house's entry point and ran one cable off that to the TV and a separate cable to my main desktop PC... Since that time... with the splitter installed, I haven't seen anything close to the 4 Mbps reading I got initially for U.S. connections....

True 10 Mb Cable Speedtest

True 8 Mb DSL Speedtest

True Cable via DSL Reports L.A. server

True DSL via DSL Reports L.A. server

True Cable via DSL Extreme L.A. server

True DSL via DSL Extreme L.A. server

True Cable via Microsoft Online Diagnostics

True DSL via Microsoft Online Diagnostics

PS - I can't account for the extremely slow download speed result shown above in MS diagnostics for the True Cable connection. I repeated that test twice and got the same result.... But, the other speed tests for cable didn't show results anywhere near that number... So I don't know what to make of that one reading.

Edited March 2, 2011 by jfchandler

[jfchandler]

OK.. next step is to cut in my Linksys N Wireless router and use the Motorola Surfboard cable modem internet to feed my main PC, disconnecting the True DSL cable modem...

Can anyone advise on how to correctly reconfigure the Linksys router settings to work with the cable modem, instead of the DSL modem?

I swapped all the Ethernet wires to the correct connections. And I went into the Linksys router settings, and changed them from PPPoE to Dynamic DHCP and enabled the DNS server setting... (I'm assuming that's the correct thing to do).

But the resulting connection seems a bit balky... I'm not sure I've done everything I need to do???

[jfchandler]

The first time around, I powered down and then powered back up first the cable modem, followed by the router...

The second time around, I did both of the above and restarted my mainPC... Seems to be connecting better now...

Checked my net connections for a separate wireless laptop and second Ethernet cabled PC feeding off the router... All working now..

Came across this general guideline for SB5101 setups with Linksys routers... How to Troubleshoot a Linksys Router and a Motorola SB5101

Another general resource:

How to Install a Linksys Wireless N Router With a Motorola Modem

And the ever popular... How to Troubleshoot Linksys Routers For Cable ISP?

Edited March 2, 2011 by jfchandler

[Pib]

JFC,

I have read in other blogs that a "cable modem" likes to really hold onto/memorize the last IP address of the device it was feeding into...say a computer, a router, etc. So, if moving the cable modem's output from a computer to router (or vice-versa) without powering down the cable modem you may not get an IP connection to the big, wide internet out there. A person needs to turn off the cable modem, swap the connections, and then turn the cable modem back on....then it's sets itself to the new device it's directly feeding. An ADSL modem don't seem to be as picky when swapping connections.

And I wouldn't be surprised if having to turn of and then turn back on the router may also be required. What I've found over the years when making modem/router/VOIP wiring changes or these device have seemed to hang-up is you need to power down both the modem and the router...then turn the modem on and let it sync to the ISP...and then turn on the router and let it sync to the modem. And if you had a VOIP device hooked to the router, you may need to power down and power up the VOIP device "after the modem and router have been powered-up and sync'ed." Kinda like a 1, 2, 3 sequence...turn the modem on and let it sync...then turn on the router and let it sync...and then turn on the VOIP device and let it sync.

Wondering if you could give us some "speedtest.net" results to Bangkok, San Fran, Atlanta, and London? Big Thanks.

Cheers,

Pib

[jfchandler]

Ya....turning off/unplugging everything, waiting 2-5 minutes while changing to the correct cable connections, and then powering up seems to be the protocol...

From the links I posted above, the suggested order for bringing back power to the devices seems to be:

1. Cable modem...giving it a chance to set up and have all lights steady

2. Router

3. PC

The part I'm still a bit unclear about is the router settings...

As I mentioned above, because I previously had been using the DSL modem, my router config was set for PPPoE with DHCP disabled... When I did the changeover to the cable modem, I manually changed those settings in the router to DHCP and Dynamic DNS server enabled (I hope I got that terminology right, or at least close).

But if I had just turned off all the power to everything, not manually changed anything...and then turned back on, would the router have self configured by recognizing the cable modem... and made those changes itself... Or if I hadn't done the changes manually, I wouldn't be online writing this now???

[Pib]

Assuming we are talking a separate modem and a separate router, DHCP is a later/simpler protocol which usually works fine between the "router and the modem." It's what I've always used over the years in my "router" setting whether having ADSL internet or Cable internet. Cable operators usually use the DHCP protocol, while ADSL operators usually use PPPoE or PPPoA between the "modem and the DSLAM." And maybe a cable DHCP modem don't like trying to talk to a separate router set to PPPoE....likes to talk in DHCP instead.

On my current TOT ADSL setup, my TOT-provided ADSL modem is set to PPPoE and it connects to my WRT54G router which is set to DHCP. My other router I use sometimes, a US Robotics router is set to DHCP also. I've never tried setting the WRT54G to PPPoE to see what happens. I may try that later today. So, this indicates to me that the ADSL modem talks to the DSLAM using PPPoE and then router and modem talk to each other using the DHCP protocol.

On a combo/intergrated ADSL modem and router which are common nowdays, which people usually refer to just as a router, I'm not sure if there is one place to set the WAN protocol (i.e., the modem-to-DSLAM link) to PPPoE and then another place to select another WAN/LAN protocol from the router-to-modem and to the clients to DHCP, PPPoE, PPPoA, etc. All these different protocols hurt my head when thinking about it....way too complicated for some one like me with a Chang Beer impaired brain.

From doing some googling it appears cable internet operators normally use DHCP. In fact, here's a cut and paste from my WRT54G basic setup page which shows my router setting on DHCP and another cut and paste from my ADSL modem showing it set to PPPoE. So, at least in my ADSL setup case which is now allowing me to connect and write this post, the modem is set to PPPoE and router is set to DHCP. Note the WRT54G even mentions DHCP is commonly used by Cable operators.

In closing, here's a link to a document that provides some a semi-layman explanation (don't hurt the head too much) of the differences between PPPoE and DHCP. http://s-tools1.juni...pers/200187.pdf

[jfchandler]

Pib, the Motorola Surfboard cable modem that True is handing out is, of course, a modem only device -- no router functionality included. That's why they go to great lengths to say that the cable modem, per se, only supports one PC.

In my case, I switched from using True's Hua Wei DSL modem and my own separate Linksys N router, to now using the Surfboard cable modem with my separate Linksys N router....

For all the time I was using the True DSL modem, my router was set to PPPoE.... When I swapped in the Motorola cable modem, I changed my router settings to DHCP and Dynamic DNS....

It's still not clear to me whether I NEEDED to make that change in the router in order for it to work with the True cable modem/service...

But the True cable modem clearly works fine with the Linksys router in DHCP mode.

Edited March 3, 2011 by jfchandler

[Pib]

In your DSL hookup with the True modem and Linksys router and with both set to PPPoE did you enter the ISP-provided User ID and Password in "both" the router and modem, or just in the modem?

In my ADSL setup with the modem and separate router, I only enter the ISP-provided User ID/ Password in the modem. When the router is set to DHCP it don't ask for any ISP User ID/Password. But if I change it to PPPoE it asks for a User ID/Password. Seems entering a User ID/Password in the router is not needed since the modem is the one using the ISP-provided User ID/Password to log onto the DSLAM/Wide Area Network. Seems this kind of connection for ADSL is pretty common among ISPs. If you did have the User ID/Password also entered in the router, maybe that works fine with an ADSL modem/setup, and the ISP-provided User ID/Password entered in both places doesn't cause any conflicts. But with a cable modem the story apparently changes...it likes using DHCP.

Additionally, if you were hooked into a non-DSL network like your work location network "and using the router to hook into the network" before hooking to your computer that would probably be a case for entering what ever User ID/Password is needed/provided to log onto that network.

But with a cable modem which does not use a User ID/Password like ADSL modem but instead uses different log-on credentials/different log-on method, then any User ID/Password in the router, like your True ADSL provided User ID/Password, may have been causing a conflict. Plus, it seems cable modems are built to usually use DHCP.

Then again, I know a person can also set a ADSL modem to "Bridge" mode which removes the User ID/Password fields in the modem and then the separate router needs to be set to PPPoE with the ISP provided User ID/Password entered in its User ID/Password fields in order to connect to the DSLAM.

So many connection options/setups possible with modems and routers....hurts my head.

[jfchandler]

I think the salient point here is, because True is only providing the cable modem, they're not going to provide any technical support or guidance for people who want to take the additional step of using their own self-provided router in terms of how to get the router to play well/best with the True cable modem.

That's why I've been trying to elicit a discussion about the best way to pair the True cable modem together with people's separate routers.... Because we're going to be on our own in making that connection and setup.

As I mentioned before, apparently, users of the True cable modem have no user ID or password to deal with for that account. Once True configures the account for the cable modem, it automatically connects to the True network when plugged in.That's of course different from the DSL service, where you absolutely need to use the user ID and password in order to connect.

In my prior DSL setup, I had configured the router with my DSL ID and password. And then I also had those same values entered in the dialer software in Windows. I never did anything in terms of directly accessing the firmware settings of the True provided DSL modem.

[tropo]

I think the salient point here is, because True is only providing the cable modem, they're not going to provide any technical support or guidance for people who want to take the additional step of using their own self-provided router in terms of how to get the router to play well/best with the True cable modem.

That's why I've been trying to elicit a discussion about the best way to pair the True cable modem together with people's separate routers.... Because we're going to be on our own in making that connection and setup.

As I mentioned before, apparently, users of the True cable modem have no user ID or password to deal with for that account. Once True configures the account for the cable modem, it automatically connects to the True network when plugged in.That's of course different from the DSL service, where you absolutely need to use the user ID and password in order to connect.

In my prior DSL setup, I had configured the router with my DSL ID and password. And then I also had those same values entered in the dialer software in Windows. I never did anything in terms of directly accessing the firmware settings of the True provided DSL modem.

I'm on my second day of Sohpon cable broadband. The ISP is CAT. A username and password is required.

I need to get a wireless router today - have an option of getting my own or buying one of theirs for 2500 baht. If I get my own and I need help setting it up I have to pay 500 baht, or its free if I use theirs. I didn't realise how expensive cable wireless routers are compared to their ADSL cousins.

[Pib]

JFC,

Wondering if you could give us some Cable internet "speedtest.net" results to Bangkok, San Fran, Atlanta, and London? Big Thanks.

Pib

[jfchandler]

JFC,

Wondering if you could give us some Cable internet "speedtest.net" results to Bangkok, San Fran, Atlanta, and London? Big Thanks.

Pib

Pib, I just did what you requested here at home... Using the Speedtest.net web site, all of the results showed basically 10 Mbps/1 Mbps on my True cable feed....but with ping times of 7 to 15 ms...

I'm surprised you want to know test results for that site... Because as we previously have discussed here at length in various threads, the Speedtest.net web site is worthless for any meaningful results from True... The results it produces either are being gamed or only measuring the first hop of the connection from my home to True... not the server 8,000 miles away.

Here's the San Francisco screen cap... I could post the others, but it would be redundant and meaningless...

Edited March 3, 2011 by jfchandler

[jfchandler]

Here's a more meaningful result from noon today... MS Diagnostics... one of the better speed numbers I've received from that site... since switching over to cable...

BTW, I did some video streaming last night, and the result seemed quite good... I was able to bump up my player from lowest speed stream to the medium speed stream setting, and it played well from the U.S. with very little stalling/freezing...

That seems to confirm the stats I've been getting that show the quality of service levels from the cable feed are better than the comparable QOS I was getting from True DSL... at least here at my particular location in BKK.

[tropo]

JFC,

Wondering if you could give us some Cable internet "speedtest.net" results to Bangkok, San Fran, Atlanta, and London? Big Thanks.

Pib

Pib, I just did what you requested here at home... Using the Speedtest.net web site, all of the results showed basically 10 Mbps/1 Mbps on my True cable feed....but with ping times of 7 to 15 ms...

I'm surprised you want to know test results for that site... Because as we previously have discussed here at length in various threads, the Speedtest.net web site is worthless for any meaningful results from True... The results it produces either are being gamed or only measuring the first hop of the connection from my home to True... not the server 8,000 miles away.

Here's the San Francisco screen cap... I could post the others, but it would be redundant and meaningless...

It would seem they are giving you exactly the results you want to see. This is interesting, because with my CAT I consistently see pings in the 200 - 300 ms range whereas you're getting a ping of 7ms. I would alway get ping results in the 30ms - 40ms range on 3BB, but never on CAT do I see such low numbers.

Edited March 3, 2011 by tropo

[jfchandler]

Tropo, I've only ever used True for home internet here, first DSL and now cable...

And ever since day one with True, whenever I'd use the Speedtest.net web site, I'd get those kinds of clearly bogus results...

Speedtest may give accurate results for other ISPs (I have no personal experience to offer an opinion of that)...

But if Speedtest is accurate with other ISPs, then that's a pretty clear sign that it is indeed True doing something themselves to game the results...

Indeed, if I'm True... I know I can't actually deliver a single stream 10 Mbps connection to the U.S. But I can at least make people THINK they're getting the rated speed of their package.

Though to be fair to True, for whatever your rated package is, they never claim you'll get that speed for international connections. They'll only claim that speed for local connections, and I don't think they even guarantee that... with terms and conditions language such as "subject to local network congestion" and such.

[Pib]

I think it's just a True thing that affects many (maybe all) True internet customers. I'm on TOT and when ever I use speedtest.net the speeds/pings match up with the bandwidth/speed monitor within my WRT-54G, I never get pings of less than 200ms to the US, etc. And when I go to other speedtest programs I get similar results to speedtest.net. It's probably just a True smoke and mirrors thing with Speedtest.net. Was hoping there wouldn't be any smoke and mirrors on the True cable internet. Oh well, will just use other sites/programs/ways to measure the real speed.

But we must remember that True uses 25th century warp speed, faster than light servers. If TOT, CAT, 3BB and other Thai ISPs would only buy a couple of True faster than light servers, internet speeds in Thailand would jump ten fold overnigt (according to speedtest.net results).

[jfchandler]

You're absolutely correct Pib... If only everyone signed up for True Internet and used Speedtest.net to measure the results, we'd all be getting 7 ms and 10 Mbps to the USA... Can't imagine why the other ISPs haven't jumped on True's bandwagon...

PS - I can't remember if I mentioned it originally, but let me mention now one of the outstanding habits of True's customer service...

The day of my cable install Monday, we had an 11 am appointment. I didn't know it, but True had called my wife that morning at work to say they'd be late, arriving about noon... No problem.. Noon passes, no tech and no call... Wait wait wait... call True again about 2 pm to complain.... Finally, the tech guy shows up about 2:30 pm -- 3.5 hours after the original appointment time...

Today, I had a supposed morning service appointment. The guy was supposed to call before noon and advise of a time for coming... It's now 2:35 pm... no call and no contact and no tech from True... Tried calling True Online's customer service number. All our staff busy, please leave your number.... That was an hour or more ago... No call back.

As I said before, their actual service technology works a whole lot better than their service staff.

I guess my point being: if you're planning a True cable internet install, to play on the safe side, better plan on spending the whole day at home. Might not take the whole day, but on the other hand, it might very well take the whole day between waiting and the actual service.

Edited March 3, 2011 by jfchandler

[Pib]

As FYI, in this Thai blog post there are some adslthailand.com speed test results for the True Cable Internet 20Mb/2Mb plan. With the exception of the ping times for the international sites and one or two of the speed results, the results look believable. http://www.adslthail...3667#post163667

Edited March 3, 2011 by Pib

[jfchandler]

Pib, every one of those screen caps the guy posted for the international destinations, including to the USA, showed reported ping times from Thailand of about 20 ms or less...

And actually, now that I look more, even shorter ping times to places like the UK, Singapore and HK than to servers also inside Thailand.

What ought to that tell you???

Edited March 3, 2011 by jfchandler

[tropo]

I think it's just a True thing that affects many (maybe all) True internet customers.

You must have missed my last post where I mentioned that I get pings in the 30 - 40 ms range to LA etc on Speedtest.net with 3BB.

[Pib]

Pib, every one of those screen caps the guy posted for the international destinations, including to the USA, showed reported ping times from Thailand of about 20 ms or less...

And actually, now that I look more, even shorter ping times to places like the UK, Singapore and HK than to servers also inside Thailand.

What ought to that tell you???

That's why I mentioned in my post the ping times where not right, but what seems different on the "speed" results was how they varied, were generally lower to international sites versus speedtest.net results which seemed to report full speed all the time especially for many True customers.

I don't know if the adslthailand.com underlying speedtest program is really speedtest.net in disguise and just being displayed differently. On the adslthailand.com speed test only about a half dozen sites come up for selection and a different speed test interface is shown. But on their same webpage there is also links that will take you to the speedtest.net website to run their test and it shows their interface.

Why don't you give the adslthailand.com speedtest a try. I won't be surprised if the ping times are bogus but if the "speed" results to the half dozen of so international sites, especially to the UK, then maybe they are really just not from an in-Thailand hop/True hidden server. Since it seems UK/Europe connections are the slowest when doing speedtests, I almost think the connection goes east out of Thailand to the US on the under-Pacific cables, across the US, and then over to the UK via under-Atlantic cables....almost like going two-thirds around the world versus going west which would be a lot closer. No, I didn't do any tracer action to see which way it really goes...maybe someone else reading this might want to do that...I just know speed tests to Europe are generally slower, at least for me on my TOT ADSL, even through Europe is much closer to Thailand than the US is.

Maybe give the adslthailand.com speed tests a try to see what type of results they give for your 10Mb/1Mb True cable package. Would only take a few minutes to run them....maybe you already have. I would just be interested in the results you get from that particular speed tester....but sure you don't go in to the speedtest.net link to run the tests. Cheers.

[Pib]

I think it's just a True thing that affects many (maybe all) True internet customers.

You must have missed my last post where I mentioned that I get pings in the 30 - 40 ms range to LA etc on Speedtest.net with 3BB.

No, I saw it. I seems from looking at a lot of 3BB related posts over the last year that most 3BB customers don't' get faster-than-light ping times...but most True customers seem to get the faster-than-light ping time results. I know for me I never got faster-than-light results on my TOT plan or previously when on a JINET plan. With TOT I would just be happy to get faster-than-sound results to international web sites (grin)....for within Thailand sites I get full speed of my package, but to international sites I usually around 25% of package speed. But my previous JINET package was extremely reliable...and TOT has been very reliable in terms of up-time....99% plus for both.. When on a JINET plan it was actually connecting over the TOT line to the TOT DSLAM, but then JINET took over for server/gateway support. Would have stayed with JINET except many of their plans are no longer price competitive for the home user. Never had the pleasure of having 3BB since I've been a captive audience to TOT or their ISP concessionaires, but I'm getting closer to experiencing more freedom, breaking TOT's bondage with my True cable package being hooked up around mid/late March....of course, that's a Thai time estimate, so I won't be surprised if it's not until April sometime. Cheers.

[negreanu]

The Ping time is because True "invisible" cache servers have the speedtest.net site cached in Bangkok due to the amount of access that site gets. So the ping and speed you are seeing is the ping to the True Caching Server and the download speed from that caching server.

Proper Ping results as stated many times before can be obtained by running the pingtest from the same website which is not cached by True as of yet.

[jfchandler]

Pib, ran the ADSL Thailand speedtest site to L.A., using my laptop and wireless N connection... didn't even bother about using my direct cable modem wired desktop...

As we might well have expected, the test showed 10/1 to Los Angeles with a ping time of 12 ms.... As I was saying before, more bogus BS....

I'm not saying the speedtest sites like this one or Speedtest.net are bad/inaccurate... though maybe they are... I have no way of knowing... Rather, they're just clearly giving inaccurate results, perhaps because the local ISPs like True here have configured their networks to handle calls to those addresses in a particular way...

Another clear sign, of course, is that these sites will give you basically the same speedtest result to any international destination you may select, and pretty much at any time of the day...regardless. And the result will be almost exactly the rate of whatever service plan you're subscribed to thru your ISP...

I couldn't help but noticing also, that True's own speedtest web page, Speedtest.net and ADSLThailand all appear to rely on some service called OOKLA... And visually, the appearance of True's own speedtest web page is almost identical to that of the ADSLThailand web page... So it's not particularly surprising that they're all producing the same results...

What I did find interesting, however, was that the Speedtest.net web site contains a link to a separate site for ping rates called PingTest.net, which certainly looks like it's being operated by the same people.. And when I used that site to run a ping test between here and L.A., look at the result it produced, 241 ms and line quality C... Comparing those results with the ones from Speedtest.net, I'd say that's more evidence that True is gaming the speed test results in some fashion...

PS - Thanks Negreanu for the technical explanation for what most of us clearly knew was occurring... It also explains why more sites than just Speedtest.net may be getting handled in that same fashion.

Edited March 4, 2011 by jfchandler

[Pib]

Thanks for running the tests. True gaming speedtest results via in-Thailand cache servers...never happen!!!!, except maybe Monday through Sunday.

But it sure eliminates about 99% of the service calls for low internet speed since I figure about 99% of True's customers take the test results as gospel as they really think the results are accurate. Nothing against True customers as I'm already a TrueVisions customer and soon to be a TrueOnline customer with the cable internet.

It does seem that the MicroSoft Online speed test/diagnostics give the most detailed/realistic results...results most people won't like seeing regarding their internet package. For those listening out there, here is the link: http://speedtest.microsoftonline.com/

And here's a link to my MicroSoft speedtest results run a few minutes ago on my TOT 6Mb plan...doesn't look too pretty on the download speed or "idle time"...the Forced Idle Time is the real killer in getting streaming video/audio to play without pausing due to all the hops to get from here to where ever in farang land. http://speedtest.mic...report?id=73153 But even with the low download speed of only around 0.5Mb, my browsing seems real snappy this morning...even my access to ThaiVisa whose server is in Singapore...but I know local cached servers could be helping a lot depending on what ever sites I'm surfing to.

[jfchandler]

Pib, as the one who introduced using the MS diagnostics info to this thread, I'd certainly agree that it provides some of the broadest and most useful set of information that I've seen from a single site... (though if anyone has any others out there that are comparable or better, please post them here. The more the merrier... ).

And the MS site certainly is far more informative than the basic download/upload number sites out there...

Did your TOT service really produce a 0% Quality of Service rating from the MS site??? My True DSL had been producing 7-20% QOS ratings from MS when I'd been checking it earlier in the week.... The True cable numbers for QOS were generally much higher, as I've noted previously.

Whenever I see an odd result like that, I always rerun the test....or try an alternate source for the same reading....

BTW, at this point, I've now unplugged my True DSL modem and cables, and running the household Internet entirely off True cable... Everything running nicely, thus far...

I'll probably keep the DSL service account for another month or so, just to make sure no unexpected surprises arise... But if everything cable wise continues along well, I'll probably look to cancel the DSL service in a month or so and perhaps look to upgrade to the 20 Mb cable plan at that time.

Not being sure about how the True cable would actually perform, having very little feedback from existing True cable users here, and knowing that I was locking myself into a one year service commitment (unless I wanted to pay close to $100 if I canceled early to reimburse True's installation charge), I decided to start out with the lowest/cheapest cable plan, just as a precaution.

Really, there's no harm for approaching it in that way, since to the best of my knowledge, True Online is more than happy to upgrade existing customers to higher speed and cost plans once you're already established with them. And, as we've learned elsewhere, quality of service levels can and do vary by geographic location even with the same ISP.

[Pib]

Did your TOT service really produce a 0% Quality of Service rating from the MS site??? My True DSL had been producing 7-20% QOS ratings from MS when I'd been checking it earlier in the week.... The True cable numbers for QOS were generally much higher, as I've noted previously.

It sure did. But I bet you if I was making a VOIP call the States the call would still have been clear as a bell, since only around 35Kb bandwidth is needed, at least with my VOIP service (8x8).

Just ran the test again and got a 48% QOS and 1.144Mb download speed. Forced idle of 64% ( a lot of marching in place, more stop than go) with a Round Trip Travel (TTT)(a.k.a., ping) of 280ms. See report: http://speedtest.mic...report?id=73178 My "first" run of the test since opening the browser.

I'm hoping there is more go than stop with True cable internet which should result in improved video streaming (less pausing), which you have been reporting appears to be better than your 8Mb ADSL package. I think you'll be happy with the cable 10Mb package since it costs less and is faster than your 8Mb ADSL package.

Ok, after writing above sentence I just ran the test again...got a 1% QOS, 257Kb download speed, Forced Idle of 65%, and a ping of 280ms. See report. http://speedtest.mic...report?id=73184

Ok, lets run it one more time...got a 53% QOS, 425Kb download speed, Forced Idle of 66%and a ping of 278ms. See report: http://speedtest.mic...report?id=73185

Regarding the "download" speed, as least for me on my TOT ADSL plan, I have noticed the "first" time I run a download speed test to an international server for most any speed test program my "first" run of the test shows a significantly higher download speed than subsequent runs (note: this rarely happens for in-Thailand sites, only the international sites). I can close my browser, go back to that speed test web site, such as speedtest.net, adslthailand.com, and maybe even the Microsoft speedtest site (need to do more testing on the MS site to confirm but on the surface it does appear to be happening)....and the great majority of time get a significantly higher download speed result on the "first" run. Even if clearing my cache before doing the first run, this higher result still happens whether I test with IE8 or Chrome. This is why many times I will ignore/throw-out the first run of the test and use the average for the subsequent tests run. Seems the Microsoft site is giving very consistent ping and Forced Idle results but download speed and QOS results varieties greatly...and maybe the Microsoft site is giving accurate results and it's just a case of the quality and speed of connection from Thailand can change easily from second to second, minute to minute, etc.

Yeap, take speed test results with a BIG grain of salt and try to figure out where any smoke and mirrors may be occurring...possibly taking only "portions" of certain speed test results and combining them with portions from other/different speed test programs may give a clearer picture...clear away a little more of the smoke....preaching to the choir I know. And having two separate internet plans hooked up like you do, at least right now, and running tests against both plans using the same test and at almost at the same time can really help to see which plan is giving the best overall results.

[jfchandler]

Pib, I concur about the MS QOS numbers... I've seen them bounce around quite a bit here on my new cable internet service... sometimes up above 80%... sometimes down in the 10 to 20% range... But in general, the cable MS QOS numbers for me do seem to be consistently better than those I was getting thru DSL...

I ran the MS test to the USA again today about 12:30 pm off my cable connection, and got one of the better speed results I've seen since switching over to cable... 3.19 Mbps down and 18% QOS... Like you, I will often re-run the same test a couple of times in succession, just to make sure whatever data I post here is reasonable accurate and reflecting of what's actually occurring..and not a one-off. But I've also had other times where the MS speed number for cable was down in the 300-400 Kbps range...though not often.

Meanwhile, I do have a bit of an update, courtesy of True... Today, finally, I had their service guys out for a bit of a follow-up check. They sent one guy to do my install last Monday, but today they sent 3 guys in a truck just for the check... Oh well... Anyway, one of them spoke passable English and I used the opportunity to try to elicit some technical info... as follows:

I had the guy use his monitors to measure both the signal coming off the cable TV/TrueVisions line feed, and then the separate coax line internet feed going to my desktop PC... both of which coming from the new two-line splitter that True installed Monday where their original cable enters my home.

According to the True guy:

For TrueVisions cable, they want the signal strength to measure above the 200 score level, and the signal to my TV measured 255.

For True cable internet, they want the signal to noise ratio to be at 30 dB or above, and mine measured 33 dB...

So the True guy pronounced that both my TV and internet feeds were operating fine... FWIW.

Edited March 4, 2011 by jfchandler