ThermoEngine: A disturbing change in direction

[Hoot]

If you feel compelled to drill it out close to the original, use a 1/2 inch bit and drill it down 1-1/4 inches. FWIW, I filled the cavity with pure silver and it did not make much difference <.5C in the performance. Now your idea of the copper rod sounds interesting from the standpoint of getting a thermally, more conductive material on top of the core. If you pursue that, let us know how it works out.

Hoot

 

[ManOfWar]

I was just at you website and why do you not mention this on your heatsink -page?? Warp Factor-E

 

[Tachyon]

Deleted

 

[Paladin06]

ThermoEngine: Holes Drilled-Now What?

Ok, I took the damn thing in the garage and got out my best drill bit. I drilled three holes to a depth of about 1/4 inch from the bottom of the heatsink. I reinstalled the fan on the heatsink and installed the HSF on my machine. Now everytime I try to start the damn thing I get all kind of error messages while in the DOS boot up.

If I reinstall my other HSF it boots ok. Anyone got any idea why? Could this damn thing be making contact with other areas of the CPU and causing this? I did notice that the CPU does not sit dead center of the heatsink. I could tell by the impression in the Arctic Silver II.

 

[Paladin06]

ThermoEngine: Holes Drilled-Now What?

Double post, sorry.

 

[Praetorian109]

Nagorak (May 01, 2001 06:09 p.m.):

How is this limited to Taiwan? I can name a fair number of U.S. companies that do the same thing. I'm sure there are Canadian and European companies that resort to these same tactics. That's not to say that all companies are bad, but quite a few of them-- maybe even the majority-- are into fleecing the customer.

This is not just a problem with Taiwan, this is an inherent problem with capitalism in general. I'm not sure that there is really anything that can be done about it, outside of people just refusing to buy from companies who employ these tactics. Buyer Beware, as always, appears to be the name of the game. :-(

Nagorak,

I thank you for the response, however, I didn't state specifically this problem is only occuring in Tawianese companies. I'm using Taiwan as an example becuase I've worked for, with, and at these either Taiwanese owned or HQ based in Taiwan companies, so I'm sharing what I know from my own experiences. I'm won't doubt that almost every company in every country has these policies but I've never personally experienced such things so I rather not say. Again I thank you for your response and I think the real reason is that hopefully this matter with Thermalsonic becomes resolved one way or the other.

Praetorian

 

[AbRASiON]

Well, now that we know they DEFINATELY have solid core ones, how long till we see a review of both side by side?

I would love to see this - on a t-bird @ 1500 (what the damn fan is "meant" to be rated for) and also say a duron @ 700 or something low end....

with stock fan and 7k fan - comparing the differences...

 

[Duck Hook]

If you are able to drill a hole of similar proportions, shouldn't you be able to achieve similar temps?

I just ordered a ThermoEngine 2 days ago, just before all the scam stuff broke, and am pretty pissed. My purchased was based largely on the superb review from HardOCP. If they knew about it, why didn't they say something.

It looks like I'll have another paper weight to use along with my Gorb, Coolmaster, and Vantec HSs. Acutally I think it will be a pretty neat looking paper weight, probably not worth the $30 though. Thank goodness I've got a Alpha that works great. Another case of if it ain't broke, don't fix it.

I'm assuming this thing is pretty soft aluminum so I'm going to take a drill to mine and see if I can't cut a hole similar to the one in the "bait and switch" models.

Any ideas what would happen if you leave the hole uncovered?

 

[Nevin]

It seems to me that a test on an Intel CPU with a calibrated internal diode or a synthetic test platform would be a much more valid test. The in-socket thermistors used to measure temperature in AMD based systems are subject to temperature change compression as a natural phenomenon of secondary flow path when there is "both" heat flow and thermal resistance from chip to ceramic surface. This is the normal/natural case and cannot be avoided.

Not only are in-socket thermistors in a secondary heat path, they also only measure an average of the CPU rear temperature (minor contact with thermistor surface) and the temperature of the air in the socket (major contact with thermistor surface) which is itself an average of the CPU rear temperature and the motherboard temperature.

With both of these factors contributing to temperature change compression, the differences between heatsinks are minimized. In most situations, this compression can be as severe as 4 to 1 so an actual 4 degree difference in core temperature is only measured by the in-socket as only a 1 degree difference.

AMD CPUs are great processors and tremendous values, but given the measurement problems associated with in-socket thermistors, AMD based systems are pretty much worthless for making valid comparisons of heatsinks unless you go to some of the extraordinary measures that Joe has at this site.

Nevin

 

[Colin]

Kyle,

I saw your post on the front page. That's ok by me. Glad to see it there. Given the number of folks on various forums that imply that your reviews were a factor in their purchase decision and the continued presence of the ThermoEngine at your Site, I would think that any design change would be headline news as soon as you are aware of it. As you point out, I don't know what goes on behind the scenes at your Site, nor do I want to. As with any hardware oriented Site, I am interested in content.

This is not about who's Site is doing what, it is about truth in advertising. If the review samples are different in design and performance than the consumer models, it's not fair to the consumer. It’s the consumer that counts. It’s the consumers that buy the products and ultimately pay for the hardware oriented Sites. It appears that Thermosonic may have pulled a fast one on us. I hate to think of the overclocking world taking on the overtones of the proverbial used car lot but the Thermoengine controversy leaves me with that same disgusting feeling in my gut.

Colin

 

[Red_Eye]

Tachyon (May 01, 2001 10:06 p.m.):
To dispel any solid core "rumors"...here is a picture.

Do the solid core units have the tops printed to match the review units? Specifically the circle and url to their web site? I ordered mine on 4/11 from inflowdirect.com with a 30 CFM fan. Mine has the printing and here are my figures.

Currently I have been folding all day and my Celeron 600 running at [email protected] volts is sitting at a warm 39C/102F. The case temp according to my CUSL2 is a steamy 30C/86F (restrictive fan opening in my case). The room temp is at 76F. I bought the ThermoEngine because prior to this I was running a GORB at up to 109-120F under the same conditions. Twenty four seconds after killing my folding the ThermoEngine drops temp to 34C/93F with the case temp remaining the same.

Hell it works better than the GORB I had, but its no huge improvement.

 

[AbRASiON]

1'st of all.

I do NOT think drilling a LARGE hole in this heatsink is easy to do, OR a viable option for most people, plus we shouldnt have to "mod" a well reviewed heatsink.

Nevin:

I disagree with your post, the thermistor on most AMD boards is located UNDERNEATH in the CPU in a socket which has no airflow etc, plus the test should be done by someone who is in controlled conditions
ie. a SMART reviewer like Kyle
This means roughly same room temp
done within the same few hours.
not even removing the cpu from the socket, just the cooler
trying both fans.
etc
and multiple tests and working out averages - a thermistor inside the cpu isnt going to make that much difference - your post basically indicates that ANY amd roundup of HSF's is subject to errors due to the lack of an internal temp monitor, but as I say - controlled conditions solve this issue.

Kyle: thanks for the comment on your page - good luck with the testing - lets see if this new one performs as good as the last one (best heat sink with 7k rpm fan if i recall from the "roundup" recently)
If it's still in the top 2 or 3 positions, it's not a bad heatsink solid or not, but if it drops much more (2+c?) then I personally believe we have an issue.


Colin:

Couldnt agree more, it's basically false advertising they could be saving quite a bit of money from this - I mean it DOSENT look like a cheap heat sink to produce, but I dont see a price drop, yet they have obviously shaved costs.... not a good sign to me.


Red_eye 30CFM fan = ?
Is that a 7200 rpm like the one Kyle used in his review, because if that's the case as you can see from his roundup - you obviously wont get similar results, BUT you should get damned good results, or so his benchmarks show.

Have you weighed your unit and or checked if it has the stamp on top?

- Scott

 

[SP]

Well, did the company that makes these heatsinks actually claim that they contained a heatpipe inside them? Also in the ones that are hollow can anyone actually confirm that there is in fact a heatpipe or are they just hollow and contain no heatpipe?

If the company never claimed there was a heatpipe in the heatsink you guys may not have any legal claim against them. Furthermore, it may be that the hollow ones don't actually contain a heatpipe and are perhaps just hollow to reduce weight or something. If this is the case perhaps the problem here is not that the company promised more than they delivered. Perhaps the problem is that people made assumptions that this heatsink was more than it actually was. If that is the case then you all have noone to blame but yourselves for falling for all the hype and not getting the facts. Remember everytime somebody on the internet tells you something is the greatest product in the world it isn't always true and never assume that a product does more or has more features than it is advertised to.

 

[T. Random]

From Hoot's original post:

Basically, the heat drawn off the cpu core was coaxed to run up the walls of the chamber, where the fins radiated from. This encouraged the heat to draw out into the fins, as opposed to storing in the mass that would have resulted from a solid core.

As far as I can see, this does not make much sense. An empty cavity inside the sink cannot "encourage" heat to move any particular way, and heat is certainly not "stored" in a solid core. On the contrary, the cavity decreases the cross sectional area of the metal through which heat is conducted from the CPU to the fins and should therefore be bad for cooling.

Now it seems that Thermosonic has officially acknowledged that the new design is indeed a bit inferior to the old one (at [h]ocp). They also say that they are surprised by the result. So am I. I'm sure that there is a logical explanation to the finding, and I'm interested to learn what it is. But whatever it is, I don't think it can be the one offered by Hoot.

 

[Hoot]

The beauty in the ThermoEngine starts with the Radial Fin concept, as opposed to the conventional Lateral Fin we've seen in HSFs for an eternity. The lateral fin design is based upon the ease of manufacturing being traded for performance. If you look at a thermogram of a lateral fin HS base, it's hot in the center and cooler towards the edges. It relies upon the heat at the center conducting outward, through the baseplate and then upward through the fins/pins. The radial approach relies upon the heat coming off the core conducting upward and then outward. If you model the two approaches in terms of R and C. (Resistance to heat flow and Capacitive storage of that heat), the two designs behave quite differently. I am not a ThermoDynamics Engineer. The chamber in the original ThermoEngine does not contain some exotic gas. It does not contain parafin. It contains nothing. The only unanswered question was whether it was a vacuum or simply full of air at standard pressure. Certainly, once the seal is broken, it is the latter. I stand by my original assessment that the hollow chamber was intended to "steer" the thermal current up the walls.

Hoot

 

[Nevin]

AbRASiON,

I don't believe I said that the thermistor measurement is influenced by air passing under the socket. I said that the temperature of the air in the socket influences the thermistor measurement. The air temperature in the socket is an average of the CPU rear temperature and the motherboard temperature. The motherboard temperature behind the socket is lowered by heatsinks with designs that blow air down around the socket (Thermoengine, Orb type).

Joe noted this effect here at Overclockers when he tested the Thermoengine and wrote:

Note also that the ABIT's socket thermistor reads very close to actual CPU temps; most other heatsinks will read 3-5C higher than actual. Expect to see other reviewers using motherboard thermistors going GAGA (erroneously) over the ThermoEngine.

Having one of the few systems that measures CPU temperature with a calibrated, imbedded thermal diode as well as with an in-socket thermistor, I have confirmed this in my own tests. This means that the user needs to add 3C to 5C to the measured Thermoengine test temperatures when comparing it to a conventional heatsink on an AMD system using the in-socket thermistor. Careful insulation of the socket could reduce this measurement discrepancy somewhat, but it is impossible to completely eliminate it without extraordinary means like covering the top of the motherboard to prevent the exhaust air from the ThermoEngine from actually contacting the motherboard’s top surface and the surrounding components.

That being said, the measurement compression that affects all in-socket thermistor measurements has been covered in numerous posts and articles here at Overclockers as well as other sites. Even in the latest heatsink round up at Anand's, heatsinks that measured over 3C apart in the synthetic portion of the tests measured EXACTLY the same with the in-socket thermistor.

So given that the two versions of the ThermoEngine could actually perform 3C or more different relative to CPU core temperatures yet both register the same temperature on the in-socket thermistor measurement, I again submit that testing the two versions using an in-socket thermistor is a waste of time

Nevin

 

[AbRASiON]

Nevin:

To an extent I do agree with your post...
However when comparing 2x thermo engine heatsinks on the same motherboard aligned the same way (fin wise and amount of air that can actually hit the motherboard) I really fail to see how a head to head comparison of ONLY these 2 heatsinks can "flawed" ?

Perhaps a heatsink which does not blow air down near the socket but rather sucks it up AND a different design may actually make the in socket thermistor a little warmer (although I still feel the difference may be more 0.3f not 3) these 2 identical heatsinks (besides the inner core) realistically should provide believable reliable results.

As for hoots (was that the name, sorry it's 12:30 here about to go to bed)

I do personally agree with your inner heatpipe argument

With a hollow inner chamber, the excess heat from the core of the CPU can NOT be "stuck" in this middle section - it's going to "hit" this point - find it hollow / quicker to travel through and therefore travel "up" higher alot faster as the centre simply wont retain any heat

similar to the old high school test of cooling a hot cup of coffee

the big flat pan far exceeds the cylindrical cup of coffee...

this principle is completely in effect here.

I am aware that unlike the coffee test, the heat is coming from a small central force, none the less internally this makes sense (to me)

EITHER WAY unfortunately I have unscrewed my fan to find I DO NOT have a cap on my newly purchased 48$ Australian heatsink, I am not impressed, considering I read a very favourable review at hardocp (no fault of kyles of course, no intention of throwing the blame)

I certainly do hope a head to head comparison will reveal some results, however my personal guess here (perhaps we should take bets?) would be approx 4-10f difference due to this discrepancy (if not MORE for the poor saps ME! who wanted it for it's quiet use mode (ie performs well with a basic fan.... apparently)



As for SP's comments:

I think you fail to realise we are not complaining because our "hype" was disapinted so to speak

We are complaining because a confirmed review of this (infact multiple, frostytech.com loved this heat sink (HOLLOW MODEL!) also)

the point is SP that if there IS a discrepancy, and this company claimed that there is some kind of "cool inner thing" (which they did but wouldnt tell us) inside the cpu then we shall believe them

PLUS they sent MULTIPLE units with hollow cores to reviewers to get excellent reviews and then cut costs

logically perhaps we can not sue (i was being somewhat arrogant, and certainly not serious when i mentioned it) but they HAVE done people wrong by creating a fale impression....

I cant wait to see the review of both (thanks kyle...) however my pessimistic side tells me bad things

Someone on the shack (shacknews.com) i discussed this with and i DIDNT make him aware of the issues with the hollow / solid models, he infact told me beforehand that he's unhappy with the performance of his model and his GORB performed better, he then checked after i told him the core, lo and behold, he had a solid...


Sure some people here can afford a new sink left right and centre AND to cut them up

others can not, I made an educated decision after research that this was an excellent product, only to be lied to by the manufacturer..... very very annoying indeed!

- bed time for me, let us hope my axia gives this fan mercy when i get my new board...

 

[AbRASiON]

(if not MORE for the poor saps ME! who wanted it for it's quiet use mode (ie performs well with a basic fan.... apparently)

sorry, just to confirm what i'm saying there

the descrepancy (if there is one) may be fore example 5% hotter with the solid core than the hollow core on a 7k rpm fan

however logically (what i was trying to say) with the standard more quiet fan, this figure will become larger and perhaps become 7-10% different.

very disapointing indeed.

 

[Nevin]

AbRASiON,

I agree that the airflow on the surface of the motherboard will not skew a test between 2 ThermoEngines, my point on that factor was that people should add 3C or so to the in-socket thermistor measurements of a ThermoEngine when comparing it to a conventional heatsink in a review or on their own systems.

The natural compression effect however, will still be a factor and could cause the measurements between the 2 ThermoEngine versions to not show any difference when there is actually a 3C or more difference in actual CPU core temperature.

Reading Kyle's post on his front page and looking at the Thermosonic letter reproduced there, I believe that he may have gotten the heatsinks backwards. It looks to me like the 'C' version is the solid core and that Thermosonic is saying that in their measurements, the solid core design is slightly better. Of course Kyle had access to the entire communication from Thermosonic, which I did not, so I could be wrong on this.

Nevin

 

[pii100]

sorry to leave for so long had to study for finals just made 82 woooooooooohooooooo but any way if a retailer makes false claims about a product they sell the could be on the reciving end of a product liability suit for instance if company XYZ sells the thermosonic and states that it uses a hollow core to lower cpu temps they are making an implied warranty and they are binding in most states