OP
- Joined
- Feb 13, 2001
- Location
- Twin Cities
- Thread Starter
- #101
That's the best news I've heard since this thread started! Thanks for the update Anarchy.
Hoot
Hoot
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
ThermoEngine: A disturbing change in direction
- Thread starter Hoot
- Start date
Glad to help Hoot. Actually it's been on [H]ard|OCP since yesterday... just no-one bothered to click the image
lolSO... since you seem to be happy about it, can you tell the viewers just what the hell it means? Under-educated formerly public school victims like me only read the last sentence. I would like a translation of the numbers
Thanks!
Well...I just bought a thermoengine a little more than a week ago and was surprised that I was not seeing the huge performance difference I was expecting to see from replacing my imitation gorb. Of course, I got the solid. No cap on top, whole top is machined. Time to do some testing.
Celeron 566@850 1.85 v (1.82 reported in MBM5). Abit slotket iii and abit BF6. Average temperature starting from 1 hour after starting Prime95 Torture test to 1 hour afterward.
Thermoengine stock - 43
Imitation GORB - 42
Might be able to try measuring performance in an Asus CUSL2 for more accurate reading, but I think this is sufficient to see that this is definitely not worth it. >P
I hope I can return this thing at the computer show this weekend... >P
Sklathill
Celeron 566@850 1.85 v (1.82 reported in MBM5). Abit slotket iii and abit BF6. Average temperature starting from 1 hour after starting Prime95 Torture test to 1 hour afterward.
Thermoengine stock - 43
Imitation GORB - 42
Might be able to try measuring performance in an Asus CUSL2 for more accurate reading, but I think this is sufficient to see that this is definitely not worth it. >P
I hope I can return this thing at the computer show this weekend... >P
Sklathill
Bro, you're using the stock fan, thermoengine is supposed to shine only with a better fan. Even the sites that praised the thermoengine spoke of the crap stock fan. Suits me, I just want quiet... with the option to upgrade to a noisy bastard later, should 1.33ghz become too slow... but if you want the performance level that has been praised, you might want to stick a big noisy vornado fan on that thing.
Anarchy69 (May 02, 2001 11:52 p.m.):
Glad to help Hoot. Actually it's been on [H]ard|OCP since yesterday... just no-one bothered to click the image
SO... since you seem to be happy about it, can you tell the viewers just what the hell it means? Under-educated formerly public school victims like me only read the last sentence. I would like a translation of the numbers
Thanks!
It says that they tested the solid core to have less thermal resistance than the hollow core. Less thermal resistance is always good. For example, the numbers for the golden orbs have a thermal resistance 0.91 c/w (c/w I believe is the rise in degrees celsius per watt applied to it.)
Here's the problem, however. Look at how much wattage they are testing the chip with. Under 30 watts! >P Puny, puny processors like the celeron put out wattage like that. And AMD Athlon 1 gig puts out 52 W. Athlon 1.30 and 1.33 puts out more than _70_ watts!!! I guess in low wattage situations, the two heatsinks perform similarly, but the curve at higher wattage is very different between the two. The newer solid core heatsink just can't handle higher wattage jobs like the old hollow one can. That curve must be VERY messed up if it performs worse on my oc'd Celeron than an imitation GORB.
Hope that helps...
Sklathill
Just to give you something to weigh against, here's my system specs.
1333 TB (Not OCed yet)
ThermoEngine Solid w/stock fan
30c System Temp
CPU Temp is Idling at 44c, and under load hits 49c. This is comparable to test results I have seen with a hollow thermoengine with a stock fan.
1333 TB (Not OCed yet)
ThermoEngine Solid w/stock fan
30c System Temp
CPU Temp is Idling at 44c, and under load hits 49c. This is comparable to test results I have seen with a hollow thermoengine with a stock fan.
Neil:
Yes, but irrelevant. The whole story of heat "accumulating" in the heat sink core is layman's logic and irrelevant to the problem in question.
Sorry, but no. That's simply wrong. If the surface geometry and airflow are identical in the two sinks, then the fin temperature will be exactly the same in both. What matters to the cpu temperature is the temperature gradient inside the sink when moving from the fins towards the cpu. And that gradient is the steeper the less cross-sectional area there is for the heat flux. Steep gradient -> hot cpu.
The heat transferred to the fins is equal to the heat emitted by the processor and has nothing to do with the core being hollow or solid.
Try making a heatsink out of a large aluminium balloon. That has lots of surface area and very little capability to "hold heat".
B is wrong. The "heat holding volume" (by which I assume you mean the specific heat times the mass of the sink) does not affect the properties of the sink under steady load. It does, however, affect the sinks ability to level off peaks under varying load conditions.
And you forgot:
D) as short and wide a path as possible for the heat flux to travel from the cpu to the fins.
Sorry, but your physics need polishing.
As I said, the hollow model may be superior, but my hunch is that the explanation is not the solid core but some other change in the sink.
Yes, but irrelevant. The whole story of heat "accumulating" in the heat sink core is layman's logic and irrelevant to the problem in question.
Sorry, but no. That's simply wrong. If the surface geometry and airflow are identical in the two sinks, then the fin temperature will be exactly the same in both. What matters to the cpu temperature is the temperature gradient inside the sink when moving from the fins towards the cpu. And that gradient is the steeper the less cross-sectional area there is for the heat flux. Steep gradient -> hot cpu.
The heat transferred to the fins is equal to the heat emitted by the processor and has nothing to do with the core being hollow or solid.
Try making a heatsink out of a large aluminium balloon. That has lots of surface area and very little capability to "hold heat".
B is wrong. The "heat holding volume" (by which I assume you mean the specific heat times the mass of the sink) does not affect the properties of the sink under steady load. It does, however, affect the sinks ability to level off peaks under varying load conditions.
And you forgot:
D) as short and wide a path as possible for the heat flux to travel from the cpu to the fins.
Sorry, but your physics need polishing.
As I said, the hollow model may be superior, but my hunch is that the explanation is not the solid core but some other change in the sink.
Beats me. I can only offer guesses.
Maybe the cavity was somehow used in the tooling process; like in order to attach the sink (blank) to the machine for tooling. Thermosonic says they changed the process - perhaps in the new process the cavity was no more needed.
So we do or dont have an entire side to side comparison of both units?
i mean someone posting "49c with it, DAMN" could just be bad luck / installed bad
we need a professional comparison
waiting patiently - a shame i have a solid though, i'd be more confident without it..
- Scott
i mean someone posting "49c with it, DAMN" could just be bad luck / installed bad
we need a professional comparison
waiting patiently - a shame i have a solid though, i'd be more confident without it..
- Scott
I would consider the comparison conducted by Thermosonic rather reliable. It shows a small but consistent difference in favour of the hollow core version. They measure the thermal resistance (junction to ambient) to be 0.6 C/W for the solid core and 0.56 C/W for the hollow core. In practice, that would mean 1-2 C difference in cpu temp, depending on power consumption.
More 'solid' proof... and my findings!
G'day,
I'm yet another sucker who recently purchased a Thermoengine under the presumtion it would be the same as those reviewed.
I received mine over a week ago (before news of the 'scam' came to light), and I immediately suspected it might be solid due to the lack of any visible plug. Upon closer inspection my colleagues and I figured the sinks were extruded in a long bar and then machined into individual units. Without drilling we saw no way Thermosonic could have made the core hollow.
Not content to sit on my laurels, I popped over to our NDI lab (non destructive investigation) and asked if they could help. Unfortunately all the X-Ray gear was now interstate (damn, I really wanted an x-ray to post online) but they did have an ultrasonic depth gage which could be calibrated to aluminium. This gage (yes, it's spelt gage) will detect even a hairline fracture in metals, so a whopping great hollow core should have showed up no worries. I tested all over the top of the sink, until I was convinced it was solid right the way through.
I present to you, a world exclusive... the exact thickness of the thermoengine core: 42.49mm
[img="[URL]http://www.netspace.net.au/~jesse/images/ThermoEngine_sm.jpg[/URL]"]
This is even more irrefutible proof that the core is solid. I just wish I had published this earlier, I may have even got the scoop over Overclockers.com. I wasn't convinced that performance would be all that different to the hollow core models however, and was trying to get hold of one for testing. Now it seems I won't have to as [H]ard OCP and Overclockers.com will probably both do it.
In any case, I DID do some testing against my all copper Kanie Hedgehog, a heatsink which is undoubtably high quality and has performered very well for me. Here's what I found:
Intel PIII FCPGA 700e @ 933 1.75v
Idle:
Hedgehog - typically 3 - 4C above ambient temps
Thermoengine - typically 0 - 1C above ambient temps
100% load (using CPU stability tester and folding@home):
Hedgehog - 39C
Thermoengine - 39C
All temps taken using MBM5. Both heatsinks used identical 5000rpm 60mm YS-Tech Fans. I am currently experiementing with 80mm and 92mm fan adaptors on the Thermoengine to see if I can't increase performance even further.
Before you question my installation technique, I should mention I am extremely fastidious when mounting heatsinks and my method generally involves the following:
1) Glueing a copper shim to the FCPGA package
2) Lapping the shim and CPU together to achieve a perfectly flush (and flat) surface
3) Lapping the heatsink in question as flat and smooth as I can manage
4) Applying only the bear minimum amount of Artic Silver required to ensure full contact between cpu and heatsink
5) Fastening, removing to check, and then refastning the heatsink on the cpu.
If you're interested in how/why I lapped my CuMine and heatsinks, check out my little site: Beast's Random Thoughts. You'll find almost all my crazy mods (such as my Blue Beast case) there.
Anyway... I'm in two minds as to whether I'll try and return my Thermoengine. On one hand I feel at betrayed and misled. On the other hand, for the money I still can't really fault the performance of the solid core Thermoengine, like any heatsink it just needs to be prepared and mounted properly. It does annoy me that it was kept so hush hush, it reeks of bait and swap.
Finally, a big hello to all the fellow Aussie overclocker out there, now stuck without our regular hangout at the OCAU forums. I'm as lost as the rest of you guys. If the forums weren't disbanded, you would have had news of my findings the day I took the ultrasonic reading. Oh well.
G'day,
I'm yet another sucker who recently purchased a Thermoengine under the presumtion it would be the same as those reviewed.
I received mine over a week ago (before news of the 'scam' came to light), and I immediately suspected it might be solid due to the lack of any visible plug. Upon closer inspection my colleagues and I figured the sinks were extruded in a long bar and then machined into individual units. Without drilling we saw no way Thermosonic could have made the core hollow.
Not content to sit on my laurels, I popped over to our NDI lab (non destructive investigation) and asked if they could help. Unfortunately all the X-Ray gear was now interstate (damn, I really wanted an x-ray to post online) but they did have an ultrasonic depth gage which could be calibrated to aluminium. This gage (yes, it's spelt gage) will detect even a hairline fracture in metals, so a whopping great hollow core should have showed up no worries. I tested all over the top of the sink, until I was convinced it was solid right the way through.
I present to you, a world exclusive... the exact thickness of the thermoengine core: 42.49mm
[img="[URL]http://www.netspace.net.au/~jesse/images/ThermoEngine_sm.jpg[/URL]"]
This is even more irrefutible proof that the core is solid. I just wish I had published this earlier, I may have even got the scoop over Overclockers.com
. I wasn't convinced that performance would be all that different to the hollow core models however, and was trying to get hold of one for testing. Now it seems I won't have to as [H]ard OCP and Overclockers.com will probably both do it.In any case, I DID do some testing against my all copper Kanie Hedgehog, a heatsink which is undoubtably high quality and has performered very well for me. Here's what I found:
Intel PIII FCPGA 700e @ 933 1.75v
Idle:
Hedgehog - typically 3 - 4C above ambient temps
Thermoengine - typically 0 - 1C above ambient temps
100% load (using CPU stability tester and folding@home):
Hedgehog - 39C
Thermoengine - 39C
All temps taken using MBM5. Both heatsinks used identical 5000rpm 60mm YS-Tech Fans. I am currently experiementing with 80mm and 92mm fan adaptors on the Thermoengine to see if I can't increase performance even further.
Before you question my installation technique, I should mention I am extremely fastidious when mounting heatsinks and my method generally involves the following:
1) Glueing a copper shim to the FCPGA package
2) Lapping the shim and CPU together to achieve a perfectly flush (and flat) surface
3) Lapping the heatsink in question as flat and smooth as I can manage
4) Applying only the bear minimum amount of Artic Silver required to ensure full contact between cpu and heatsink
5) Fastening, removing to check, and then refastning the heatsink on the cpu.
If you're interested in how/why I lapped my CuMine and heatsinks, check out my little site: Beast's Random Thoughts. You'll find almost all my crazy mods (such as my Blue Beast case) there.
Anyway... I'm in two minds as to whether I'll try and return my Thermoengine. On one hand I feel at betrayed and misled. On the other hand, for the money I still can't really fault the performance of the solid core Thermoengine, like any heatsink it just needs to be prepared and mounted properly. It does annoy me that it was kept so hush hush, it reeks of bait and swap.
Finally, a big hello to all the fellow Aussie overclocker out there, now stuck without our regular hangout at the OCAU forums. I'm as lost as the rest of you guys. If the forums weren't disbanded, you would have had news of my findings the day I took the ultrasonic reading. Oh well.
What they should do is a better chart, showing rise in temperature over wattage from 30 - 80 watts, rather than test only at 30 watts. If you're even upping the voltage on you FCPGA procs above 1.75, you're definitely hitting more than the 29.312 watts that Thermosonic is testing the heatsinks at. High-end athlons put out 70+ watts.
We need a full chart.
Ok, I dont know much about what is going on but I found out last night about this. I took a look at my ThermoEngine and it has a solid core. I was told by a friend that it gets 15C hotter than a hollow core one and so he said that the solid core one is not worth the money. Im not to sure about that, but this is mine:-
I have a ABIT KT7A-RAID with a AMD T-Bird 1GHz AXIA @ 1.25GHz, also I have Juno P6 Full Tower(5 * 80mm fans). When idle the CPU is about 28C, when I start using games and programs it gets up to about 39-41C.
I dont know to much about it. So could someone tell me whether this is good or bad and is worth keeping the cooler?
I have a ABIT KT7A-RAID with a AMD T-Bird 1GHz AXIA @ 1.25GHz, also I have Juno P6 Full Tower(5 * 80mm fans). When idle the CPU is about 28C, when I start using games and programs it gets up to about 39-41C.
I dont know to much about it. So could someone tell me whether this is good or bad and is worth keeping the cooler?
If you're seeing below 45 at load, then you're doing pretty well. If it stays around there after a good hour of gaming, I imagine you're doing just fine. I'm still trying to figure out what's up with why the thermoengine performs that poorly on my Celeron...grr.
Qwegji,
Until some valid comparisons between the hollow and solid core units are available (valid = other than from measurement compressed in-socket thermistors), we will not know the performance difference, if any, between the 2 models. I think that you friend is exaggerating when he said that the solid core gets 15C hotter than the hollow core, but then again, exaggeration is commonplace on breaking news.
The question now is why Thermosonic continues to send the old hollow core units out for all reviews when they say that the new solid core units perform better.
That doesn't make sense.
Nevin
Until some valid comparisons between the hollow and solid core units are available (valid = other than from measurement compressed in-socket thermistors), we will not know the performance difference, if any, between the 2 models. I think that you friend is exaggerating when he said that the solid core gets 15C hotter than the hollow core, but then again, exaggeration is commonplace on breaking news.
The question now is why Thermosonic continues to send the old hollow core units out for all reviews when they say that the new solid core units perform better.
That doesn't make sense.
Nevin
Companies change their products all the time. During its life span, a car model undergoes a constant series of cost-cutting changes. What Thermosonic did was unfortunate, but nothing out of the ordinary.
On the other hand, intentionally spreading false measurements diguised as impartial controlled test would be out of the ordinary. Firms present data selectively and may suggest misleading interpretations, but they rarely present outright wrong numbers. That is illegal and tends to get one in trouble.
I've so far seen nothing to suggest that Thermosonic is guilty of the latter.
Besides, if they intentionally distorted the figures, why would they still show figures that confirm the allegations? Don't you think that they would have gone all the way and given us figures that show that nothing has changed?
Years ago, a friend of mine told me his theory that when a conspiracy theory and a F***-up theory fit the facts equally well, then people always believe in the conspiracy theory but the F***-up theory is the correct one. I have since verified the theory in numerous occasions.
Accordingly, I believe Thermosonic did not conspire but F***ed up.
