OMG!!!! Thermaltake wants Mikewarrior to close his site?!?!?!?

[Maverick2002]

I agree. We do not and will not put up with this. There will always be critics, and is Thermaltake tries to silence them it will only make things worse. Once again Mike, let me know of any recent developments, I'm quite interested in seeing where this will go.

 

[Mikewarrior2]

Just to chime in, this is my 2 lines to which I refer to thermaltake.

From looking at the graph, something very serious is going on with the Thermaltake sinks. I sincerely hope that Thermaltake isn't relying on socket-thermistor tests to "lie" about their own heatsink/fan performance.

if anyone can find anything slanderous in that statement, please let me know. This "supposed" Bias I guess has founds its way into two sentences .



Mike

 

[Liquidity]

Can someone please explain to me how "requesting that a web page be removed" is a "strongarm tactic"? From what we've heard so far, they didn't threaten to sue, so I hardly feel like TT are doing anything more than any other company would do.

 

[Antisocial Virge]

I personally think you should finish the page. Every experiment I ever did in University had to have an explanation for the results. I personally can't see any reason for the greater variance in temp readings with the orb on the core. I was thinking about your idea of the airflow affecting the socket thermistor and I don't see how thats possible. The airflow comming off a orb is weak at best and it in a more horizontal plane than any other heatsink. I can't see how that would affect the socket thermistor. Like I said though, a simple sealing of the socket would remove that variable.

I see thermaltake is listed as one of your sponsors... I on the other hand, am sponsored by nobody. Some bias, perhaps?

If that was the case then all review sites would be biased. The fact that the news of pending shutdown of your site was posted on Millisec, who does not sell ORBS, could be used to call you bias also.

 

[Mikewarrior2]

But LIke I have said several times on this thread, I did not initiate the posting. I cannot control what others do.

The fact remains that something happens wtih orbs and socket-thermsitor readings. yes, more work needs to be done.

In an ideal world, a socket-thermistor would have compressed, but "relative ranking" of heatsinks would be intact. In the case of the orbs, the relative ranking is skewed. I have not tried sealing the socket, and will do so soon. I'm in the middle of doing some other testing as well.

And Like I said, I can't explain it yet. the next step is to insulate the thermsitor in the cpu back... this would likely cause all temps to go up, but I'm expecting a higher-"climb" for the Orbs.


Mike
EDIT: Again, I do not know why the orb results are skewed... They continue to be skewed in the same fashion, now that testing with 2 S-orbs and 4 C-orbs has been done. DId I think this was gonna happen when I made the page? no. I honestly expected socket-thermistor temps to be "relative" but compressed.

 

[DaddyG]

I believe that I was the first to post an E-Mail response from AMD regarding the Golden Orb. This was not the Socket 370 version, but the first Socket A. AMD's response to a request as to why a 900 BIRD failed was that the GORB provided insufficient cooling. Some Thermaltake big wig continued to state that the GORB was tested to 1 gig OK. The Chrome ORB was supposed to resolve all the issues that AMD had presented. After months, its still not AMD approved.

Keep up the good work MikeW.

 

[Mikewarrior2]

As far as some of you calling the Anandtech people sheep, consider the following:

1) You people are new
2) You have not seen the many months of "work" I have done posting this information over and over and over and over again
3) I would guess that almost everyone rejected the "theory that socket-thermistors are wrong" when I first brought it up. It is only after repetition that people have started to "back" me and I still have many, many detractors.
4) I've have learned as much as I can in the past 6 months as to why socket-thermsitor readings are incorrect. That is no reason for people to act as "sheep". As I am still learning more, I am doing this for people. I get no financial gain, and no real personal gain doing this. Every day I learn about a new reason why socket-thermistor's are wrong. I have thoroughly explained the compression phenomenon, which is perhaps the most compelling reason why these tests are wrong.

5) you're not giving people enough credit. Yes, there are sheep all over every BBS, but I guarantee for a good number of people on this thread that I had to pound this idea into them.


Mike

 

[Eug]

Good thread. The funny thing for me is I have absolutely no idea what my temps are because my mobo can't read the Celeron's diode, yet I continue to gleefully overclock to the max. (Not an AMD 50 W chip though, heheh, and I've got an Alpha PEP66.)

Anyways, how severe was this email? So far all I've seen is something on Millisec, and when I read it, I thought it was a bit inappropriate to post it there, especially since they would have a good reason to have an anti-TT bias. MW2, it would be common sense that the orbs don't perform so hot (err... cool) vs. the Alphas and GlobalWins, and while I appreciate your efforts, the review did seem unusually biased against TT, especially when you used the word "lie". Plus, as you freely admit, there is a lot of conjecture and theory there. Sure, it makes sense the thermistor readings are not going to be reliable, but I find it hard to completely accept your theoretical numbers either.

If the reports of TT's response are true though, that is completely unacceptable. It would have seemed better to just to have tried to improve that supposedly crapola clip, because clearly the Gorbs work at least OK on the Intel side. (Again, I know these are not the 50 W chips, etc.)

Just trying to be objective here with the limited information I have on the subject...

 

[Maverick2002]

MIKE:
If I were you, I'de keep the page up but remove any objectionable material, such as the posted "lie about the readings" clause. But whatever you do, keep the graphs, if the coolers suck then they suck and there's nothing anyone can do about it

 

[Mikewarrior2]

Eug,

It was a "shut down your site" e-mail... I think it was from a PR person. He wasn't that threatening, and I'm thinking about whether he did it alone or with the company's consent.

As far as the "lie"..> I said I hope Thermaltake is not relying on on socket-thermistor tests to "lie" about their own heatsink performance. I'm not saying htey are laying, but its clear that socket-A boards are lying when reporting temp.

If hope they aren't pointing people to "socket-a based reviews" to show how well their performance is. Because these "reviews" are very misleading.

I am unforgiving of their performance. For all the reviews out there that claim a 1-2C difference from an alpha, lots of people have purchased orbs thinking they were low-cost great performers. When in fact, they are low-cost low performers. That isn't bias, its the truth. If it sounds like i'm going against htem all out, it isn't... it just so happens they are a very popular heatsink(not only in "reviews" but also in use).

As far as theoretical temps... that's what they are theoretical. But there is no way, when looking at heatsink testing, diode versus thermsistor, and seeing the 10+C gap between a FOP32-1 and a C-orb can you in any way envision the 1 or 2C difference that the anandtech review "claims". These tests were on a CPU that had at best 60% of the cpu W of the t-bird that Anandtech used. One can definately assume with safe measure that the differences would increase with a higher W cpu, not decrease.

On a cpu of 34W, you get a 10C difference.. on a 55W cpu, you get a 2C Difference? I simply do not understand how that makes any sense.


Mike
Maverick, I am not calling them liars... I'm just hoping they are using socket-thermistor reviews to incorrectly reflect their performance... I think i'll reword it like that

 

[Mikewarrior2]

Thermaltake News Room.

Hey, thought you guys would want to read this stuff...

The best cooling system for the next generational Microprocessor- NEW AMD DURON? PROCESSOR. . Link points to Chrome Orb.

Other interesting stuff abound on this page, too.

Like a link to the review at HardOCP where the S-orb magically ties the PAl6035

Mike

 

[office boy]

Funny thing is, if you look here, seems like TT should be getting the same results as Mikewarrior2.

BTW Mike, do you have any pictures of your testing? I'd be interesting to see how you do your thermocouple installation, and what it looks like when you?re done.

 

[Mikewarrior2]

OfficeBoy,

YGM.



Mike

 

[Maverick2002]

looking for that link?
here it is

 

[Eug]

Playing the devil's advocate here, although I in no way endorse the actions of ThermalTake.

First, is it true that TT has threatened to sue? That's what the first message in this thread states.

Second, I see no problem in TT's linking HardOCP's reviews to their site. Is Asus guilty of misleading information by putting AnandTech's recommendations on their products? Nah, it's part of the business.

I think before everyone freaks out, what we really should do is see what is important here:

1) TT coolers for AMD's may not be all that.
2) You should take AnandTech's and HardOCP's cooler reviews with a grain of salt.
3) TT's email was uncalled for.

The other stuff is no big deal to me.

 

[Mikewarrior2]

It isn't a big deal to me... I didn't post this stuff 2 weeks ago, and I could have.

Linking to a review that essentially "lies" and completely goes against their own recommendations for measuring temp is funny to me... THey post their method of testing, and i'm sure that Thermaltake, as a company, understands that socket-thermistor testing is inaccurate and compressed.

As far as taking the reviews with a grain of salt, I would rather see the review quality go up. For the limited audience my page gets, many more are influenced by the malfed up resuts on Anandtech, Hardocp, tomshardware and etc "reviews".

Just look at the amount of c-orbs that people recommend all over the 'net. Heck, even some people in this htread are die-hard Orb fans.


Mike

 

[extro]

<< From looking at the graph, something very serious is going on with the Thermaltake sinks. I sincerely hope that Thermaltake isn't relying on socket-thermistor tests to &quot;lie&quot; about their own heatsink/fan performance. >>




I can see why Thermaltake takes objection to this sentence. It implies that they are lying about their product's performance. It is a poor choice of words. It does not give them the benefit of the doubt that perhaps they made an honest error (if we were to assume that MikeWarrior's tests are valid), but jumps to the conclusion that they are being intentionally deceitful.

As far as Mike's tests are concerned; until they are duplicated by a third party, they have to be viewed with skepticism. Pons and Fleischman thought they had discovered cold fusion. Their test procedures and results were stringently well-documented and published in scientific journals. They were certain of their results. To date, nobody has been able to reproduce their test results.

 

[Mikewarrior2]

Extro,

The wording has been changed.

Also... The results have been correllated between 2 parties that are (one hte same page) working on this. Soon to have third adn 4th parties, and when that is done, all the results will be posted.

Preliminary testing shows that while cpu temp changes between each parties setup, the &quot;differences between heatsinks&quot; remains the same.


Mike

 

[natedog]

Now this is just getting kind of sad. I posted a well reasoned and thought out post (when I could have been going to bed) and mike comes back and tells me to F*ck off. What's up with that? I'm not a PR mole whether you like it or not. You want me to give you my ip to prove it? I would, but then I don't know what you'd do with it.

Ok everyone, Mike has had his 5 min of fame. Now I think we should all get a life.

 

[Mikewarrior2]

What polite words? And if you think i'm doing this for fame, then f-off on that too.

And I'm telling you to f-off because you refuse believe science, and now tests that prove that socket-thermistor readings are flat out wrong.


Mike

 

[Maverick2002]

So natedog, will you admit to EVERYONE here that mike's results are more accurate and prove TT's heatsinks are worse than they say they are? However you respond, it will explain a lot. And mike isn't doing this for fame either, as a matter of fact I started this post so it looks like you got the wrong man....

 

[necromancyr]

I just wanted to put my two cents into this conversation.

The Golden Orb, if you use the Burning Issues C/W (0.515) for the heatsink/fan, is adequate for CERTAIN applications. These applications are not 'hardcore' or 'super' overclocking applications in the least.

I'm a prime example...I have a GOrb, that I'm decently happy with.

I'm running a Cel566@850 (1.7V). Now, according to the equations, etc., my top temps. should be 12C OVER ambient under full load. This is EXACTLY what they are.

The funny part of all of this is the TT website says the C/W for the orbs is 0.98 - which is horrendous. According to THEIR graphs I should be 30-40C above ambient at the level I'm running.

I'm not taking sides here, just saying - their are applications where CERTAIN orbs are useful and others where their crap. (Personally, I think their crap for AMD solutions at the current time)

Hopefully this saved the person mildly OC'ing their Celeron from spending 50$ on a crazy ass heatsink they dont need .

 

[Mikewarrior2]

Like I said here and in other threads, orbs for p3s are fine.

Part of the reason the g-orb rates higher than thermaltakes ratings is the superb clip(IMO opinion, anyways). Socket-A orbs do not have nearly as good a clip.


Mike

 

[Insane3D]

This issue with the Orbs is strictly regarding Socket A applications...not Intel CPU's...

 

[necromancyr]

I was just saying. Didnt have time to read 200+ posts . What I am wondering though is this...the Gorbs...if you reach their 'theoretical' limit according to the C/W, will a diff. thermal interface material or lapping cause any decrease in temp?